{ "best_metric": null, "best_model_checkpoint": null, "epoch": 14.999366420274551, "global_step": 17745, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.399280548095703, "epoch": 0.0, "learning_rate": 2.8169014084507045e-08, "loss": 10.0071, "step": 1, "task_loss": 4.891738414764404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.043680191040039, "epoch": 0.0, "learning_rate": 5.633802816901409e-08, "loss": 11.6805, "step": 2, "task_loss": 4.73251485824585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.580682754516602, "epoch": 0.0, "learning_rate": 8.450704225352114e-08, "loss": 10.5272, "step": 3, "task_loss": 4.749826431274414 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.154165267944336, "epoch": 0.0, "learning_rate": 1.1267605633802818e-07, "loss": 10.5234, "step": 4, "task_loss": 4.633692741394043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.92446517944336, "epoch": 0.0, "learning_rate": 1.4084507042253522e-07, "loss": 10.9615, "step": 5, "task_loss": 4.72067928314209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.130830764770508, "epoch": 0.01, "learning_rate": 1.6901408450704228e-07, "loss": 10.3515, "step": 6, "task_loss": 4.630175590515137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.704891204833984, "epoch": 0.01, "learning_rate": 1.971830985915493e-07, "loss": 10.8688, "step": 7, "task_loss": 4.697543621063232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.523429870605469, "epoch": 0.01, "learning_rate": 2.2535211267605636e-07, "loss": 10.8928, "step": 8, "task_loss": 4.785290241241455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.853700637817383, "epoch": 0.01, "learning_rate": 2.535211267605634e-07, "loss": 9.9249, "step": 9, "task_loss": 4.663565158843994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.901111602783203, "epoch": 0.01, "learning_rate": 2.8169014084507043e-07, "loss": 11.1932, "step": 10, "task_loss": 4.828158855438232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.091609954833984, "epoch": 0.01, "learning_rate": 3.0985915492957747e-07, "loss": 10.3821, "step": 11, "task_loss": 4.563029766082764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.824514389038086, "epoch": 0.01, "learning_rate": 3.3802816901408456e-07, "loss": 11.1491, "step": 12, "task_loss": 4.635919094085693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.270526885986328, "epoch": 0.01, "learning_rate": 3.6619718309859155e-07, "loss": 10.7159, "step": 13, "task_loss": 4.609441757202148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.251729965209961, "epoch": 0.01, "learning_rate": 3.943661971830986e-07, "loss": 10.9164, "step": 14, "task_loss": 4.690701484680176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.487654685974121, "epoch": 0.01, "learning_rate": 4.225352112676056e-07, "loss": 10.8943, "step": 15, "task_loss": 4.611202716827393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.20788860321045, "epoch": 0.01, "learning_rate": 4.507042253521127e-07, "loss": 10.8489, "step": 16, "task_loss": 4.61071252822876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.422500610351562, "epoch": 0.01, "learning_rate": 4.788732394366198e-07, "loss": 10.6796, "step": 17, "task_loss": 4.4972429275512695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.608845710754395, "epoch": 0.02, "learning_rate": 5.070422535211268e-07, "loss": 10.4477, "step": 18, "task_loss": 4.795047283172607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.164052963256836, "epoch": 0.02, "learning_rate": 5.352112676056338e-07, "loss": 10.2483, "step": 19, "task_loss": 4.690609455108643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.532797813415527, "epoch": 0.02, "learning_rate": 5.633802816901409e-07, "loss": 10.7818, "step": 20, "task_loss": 4.76283073425293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.962697982788086, "epoch": 0.02, "learning_rate": 5.915492957746479e-07, "loss": 10.9846, "step": 21, "task_loss": 4.760645866394043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.07254409790039, "epoch": 0.02, "learning_rate": 6.197183098591549e-07, "loss": 10.836, "step": 22, "task_loss": 4.693588733673096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.156700134277344, "epoch": 0.02, "learning_rate": 6.47887323943662e-07, "loss": 10.9571, "step": 23, "task_loss": 4.7592997550964355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.314519882202148, "epoch": 0.02, "learning_rate": 6.760563380281691e-07, "loss": 11.4295, "step": 24, "task_loss": 4.613511085510254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.720993995666504, "epoch": 0.02, "learning_rate": 7.042253521126761e-07, "loss": 10.7734, "step": 25, "task_loss": 4.495548725128174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.651700973510742, "epoch": 0.02, "learning_rate": 7.323943661971831e-07, "loss": 11.107, "step": 26, "task_loss": 4.585072994232178 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.273109436035156, "epoch": 0.02, "learning_rate": 7.605633802816901e-07, "loss": 10.9263, "step": 27, "task_loss": 4.6136016845703125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.071043014526367, "epoch": 0.02, "learning_rate": 7.887323943661972e-07, "loss": 11.9547, "step": 28, "task_loss": 4.565066814422607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.542444229125977, "epoch": 0.02, "learning_rate": 8.169014084507043e-07, "loss": 10.6434, "step": 29, "task_loss": 4.739510536193848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.029298782348633, "epoch": 0.03, "learning_rate": 8.450704225352112e-07, "loss": 11.1649, "step": 30, "task_loss": 4.691373348236084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.772214889526367, "epoch": 0.03, "learning_rate": 8.732394366197183e-07, "loss": 10.7238, "step": 31, "task_loss": 4.613551139831543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.930421829223633, "epoch": 0.03, "learning_rate": 9.014084507042254e-07, "loss": 10.9756, "step": 32, "task_loss": 4.706444263458252 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.574166297912598, "epoch": 0.03, "learning_rate": 9.295774647887324e-07, "loss": 10.4798, "step": 33, "task_loss": 4.529212474822998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.481444358825684, "epoch": 0.03, "learning_rate": 9.577464788732395e-07, "loss": 10.0728, "step": 34, "task_loss": 4.68502950668335 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.878408432006836, "epoch": 0.03, "learning_rate": 9.859154929577465e-07, "loss": 10.8027, "step": 35, "task_loss": 4.738052845001221 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 14.702432632446289, "epoch": 0.03, "learning_rate": 1.0140845070422536e-06, "loss": 11.71, "step": 36, "task_loss": 4.768803596496582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.106947898864746, "epoch": 0.03, "learning_rate": 1.0422535211267606e-06, "loss": 10.2682, "step": 37, "task_loss": 4.6503119468688965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.527111053466797, "epoch": 0.03, "learning_rate": 1.0704225352112677e-06, "loss": 11.6225, "step": 38, "task_loss": 4.669107437133789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.199080467224121, "epoch": 0.03, "learning_rate": 1.0985915492957747e-06, "loss": 11.5687, "step": 39, "task_loss": 4.728382587432861 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.792350769042969, "epoch": 0.03, "learning_rate": 1.1267605633802817e-06, "loss": 10.3094, "step": 40, "task_loss": 4.589993000030518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.995615005493164, "epoch": 0.03, "learning_rate": 1.1549295774647888e-06, "loss": 10.2234, "step": 41, "task_loss": 4.646106719970703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.894269943237305, "epoch": 0.04, "learning_rate": 1.1830985915492958e-06, "loss": 11.0, "step": 42, "task_loss": 4.722424507141113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.907012939453125, "epoch": 0.04, "learning_rate": 1.211267605633803e-06, "loss": 10.9857, "step": 43, "task_loss": 4.520284175872803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.693136215209961, "epoch": 0.04, "learning_rate": 1.2394366197183099e-06, "loss": 10.1582, "step": 44, "task_loss": 4.544990062713623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.440729141235352, "epoch": 0.04, "learning_rate": 1.267605633802817e-06, "loss": 11.1325, "step": 45, "task_loss": 4.650798320770264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.90224838256836, "epoch": 0.04, "learning_rate": 1.295774647887324e-06, "loss": 11.3326, "step": 46, "task_loss": 4.618316650390625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.97432804107666, "epoch": 0.04, "learning_rate": 1.323943661971831e-06, "loss": 10.0193, "step": 47, "task_loss": 4.772137641906738 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.797518730163574, "epoch": 0.04, "learning_rate": 1.3521126760563382e-06, "loss": 11.2877, "step": 48, "task_loss": 4.579623222351074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.300416946411133, "epoch": 0.04, "learning_rate": 1.380281690140845e-06, "loss": 9.8252, "step": 49, "task_loss": 4.591859340667725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.106938362121582, "epoch": 0.04, "learning_rate": 1.4084507042253521e-06, "loss": 10.9885, "step": 50, "task_loss": 4.510244846343994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.987412452697754, "epoch": 0.04, "learning_rate": 1.4366197183098591e-06, "loss": 11.1488, "step": 51, "task_loss": 4.773073196411133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.975807189941406, "epoch": 0.04, "learning_rate": 1.4647887323943662e-06, "loss": 10.6479, "step": 52, "task_loss": 4.779064178466797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.326028823852539, "epoch": 0.04, "learning_rate": 1.4929577464788734e-06, "loss": 10.9567, "step": 53, "task_loss": 4.6484785079956055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.722723007202148, "epoch": 0.05, "learning_rate": 1.5211267605633803e-06, "loss": 10.8885, "step": 54, "task_loss": 4.783504962921143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.634294509887695, "epoch": 0.05, "learning_rate": 1.5492957746478875e-06, "loss": 10.4851, "step": 55, "task_loss": 4.667494773864746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.207616806030273, "epoch": 0.05, "learning_rate": 1.5774647887323943e-06, "loss": 11.1749, "step": 56, "task_loss": 4.508828639984131 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.490966796875, "epoch": 0.05, "learning_rate": 1.6056338028169016e-06, "loss": 10.776, "step": 57, "task_loss": 4.746265411376953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.183530807495117, "epoch": 0.05, "learning_rate": 1.6338028169014086e-06, "loss": 10.5301, "step": 58, "task_loss": 4.731396675109863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.617422103881836, "epoch": 0.05, "learning_rate": 1.6619718309859155e-06, "loss": 10.9423, "step": 59, "task_loss": 4.559072971343994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.361732482910156, "epoch": 0.05, "learning_rate": 1.6901408450704225e-06, "loss": 11.2359, "step": 60, "task_loss": 4.603846549987793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.512797355651855, "epoch": 0.05, "learning_rate": 1.7183098591549297e-06, "loss": 10.1261, "step": 61, "task_loss": 4.484249114990234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.929279327392578, "epoch": 0.05, "learning_rate": 1.7464788732394366e-06, "loss": 10.4392, "step": 62, "task_loss": 4.767356872558594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.293985366821289, "epoch": 0.05, "learning_rate": 1.7746478873239436e-06, "loss": 10.6814, "step": 63, "task_loss": 4.581082820892334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.530652046203613, "epoch": 0.05, "learning_rate": 1.8028169014084509e-06, "loss": 10.5422, "step": 64, "task_loss": 4.593387603759766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.658989906311035, "epoch": 0.05, "learning_rate": 1.8309859154929579e-06, "loss": 11.4724, "step": 65, "task_loss": 4.689209938049316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.65280532836914, "epoch": 0.06, "learning_rate": 1.8591549295774647e-06, "loss": 11.0593, "step": 66, "task_loss": 4.621992588043213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.59499740600586, "epoch": 0.06, "learning_rate": 1.887323943661972e-06, "loss": 10.7843, "step": 67, "task_loss": 4.579416751861572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.106800079345703, "epoch": 0.06, "learning_rate": 1.915492957746479e-06, "loss": 10.6469, "step": 68, "task_loss": 4.856501579284668 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.152206420898438, "epoch": 0.06, "learning_rate": 1.943661971830986e-06, "loss": 11.2025, "step": 69, "task_loss": 4.75594425201416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.754813194274902, "epoch": 0.06, "learning_rate": 1.971830985915493e-06, "loss": 10.6631, "step": 70, "task_loss": 4.657159328460693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.337379455566406, "epoch": 0.06, "learning_rate": 2.0000000000000003e-06, "loss": 10.0276, "step": 71, "task_loss": 4.761862754821777 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.872325897216797, "epoch": 0.06, "learning_rate": 2.028169014084507e-06, "loss": 10.2641, "step": 72, "task_loss": 4.640522480010986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.693085670471191, "epoch": 0.06, "learning_rate": 2.056338028169014e-06, "loss": 11.5914, "step": 73, "task_loss": 4.582435607910156 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.04063606262207, "epoch": 0.06, "learning_rate": 2.0845070422535212e-06, "loss": 10.3241, "step": 74, "task_loss": 4.766111373901367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.973433494567871, "epoch": 0.06, "learning_rate": 2.112676056338028e-06, "loss": 10.669, "step": 75, "task_loss": 4.4964776039123535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.486418724060059, "epoch": 0.06, "learning_rate": 2.1408450704225353e-06, "loss": 10.7383, "step": 76, "task_loss": 4.650036334991455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.418624877929688, "epoch": 0.07, "learning_rate": 2.1690140845070426e-06, "loss": 11.2113, "step": 77, "task_loss": 4.644071578979492 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.922872543334961, "epoch": 0.07, "learning_rate": 2.1971830985915494e-06, "loss": 11.1233, "step": 78, "task_loss": 4.66641902923584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.391945838928223, "epoch": 0.07, "learning_rate": 2.2253521126760562e-06, "loss": 10.2484, "step": 79, "task_loss": 4.585971832275391 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.599638938903809, "epoch": 0.07, "learning_rate": 2.2535211267605635e-06, "loss": 11.664, "step": 80, "task_loss": 4.563695430755615 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.839431762695312, "epoch": 0.07, "learning_rate": 2.2816901408450707e-06, "loss": 10.3478, "step": 81, "task_loss": 4.771480560302734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.951419830322266, "epoch": 0.07, "learning_rate": 2.3098591549295775e-06, "loss": 11.4821, "step": 82, "task_loss": 4.619713306427002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.799234390258789, "epoch": 0.07, "learning_rate": 2.338028169014085e-06, "loss": 12.058, "step": 83, "task_loss": 4.561478137969971 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.449674606323242, "epoch": 0.07, "learning_rate": 2.3661971830985916e-06, "loss": 11.8026, "step": 84, "task_loss": 4.626176834106445 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.154557228088379, "epoch": 0.07, "learning_rate": 2.3943661971830984e-06, "loss": 11.2429, "step": 85, "task_loss": 4.613253116607666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.895186424255371, "epoch": 0.07, "learning_rate": 2.422535211267606e-06, "loss": 10.5002, "step": 86, "task_loss": 4.756536960601807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.251350402832031, "epoch": 0.07, "learning_rate": 2.450704225352113e-06, "loss": 10.5044, "step": 87, "task_loss": 4.595606327056885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.199639320373535, "epoch": 0.07, "learning_rate": 2.4788732394366198e-06, "loss": 10.6771, "step": 88, "task_loss": 4.648237705230713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.980515480041504, "epoch": 0.08, "learning_rate": 2.5070422535211266e-06, "loss": 11.7213, "step": 89, "task_loss": 4.651578903198242 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.191513061523438, "epoch": 0.08, "learning_rate": 2.535211267605634e-06, "loss": 10.491, "step": 90, "task_loss": 4.769591331481934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.882904052734375, "epoch": 0.08, "learning_rate": 2.563380281690141e-06, "loss": 10.7959, "step": 91, "task_loss": 4.662966251373291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.562097549438477, "epoch": 0.08, "learning_rate": 2.591549295774648e-06, "loss": 10.288, "step": 92, "task_loss": 4.714051246643066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.805070877075195, "epoch": 0.08, "learning_rate": 2.619718309859155e-06, "loss": 10.5045, "step": 93, "task_loss": 4.528115272521973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.064163208007812, "epoch": 0.08, "learning_rate": 2.647887323943662e-06, "loss": 11.4752, "step": 94, "task_loss": 4.673368453979492 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.234027862548828, "epoch": 0.08, "learning_rate": 2.6760563380281692e-06, "loss": 10.5994, "step": 95, "task_loss": 4.416010856628418 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.381855010986328, "epoch": 0.08, "learning_rate": 2.7042253521126765e-06, "loss": 10.8196, "step": 96, "task_loss": 4.479274749755859 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.296269416809082, "epoch": 0.08, "learning_rate": 2.7323943661971833e-06, "loss": 10.3652, "step": 97, "task_loss": 4.63809871673584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.026663780212402, "epoch": 0.08, "learning_rate": 2.76056338028169e-06, "loss": 10.8135, "step": 98, "task_loss": 4.628067493438721 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.557378768920898, "epoch": 0.08, "learning_rate": 2.7887323943661974e-06, "loss": 11.3301, "step": 99, "task_loss": 4.584196090698242 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.899571418762207, "epoch": 0.08, "learning_rate": 2.8169014084507042e-06, "loss": 10.5453, "step": 100, "task_loss": 4.562654972076416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.915513038635254, "epoch": 0.09, "learning_rate": 2.8450704225352115e-06, "loss": 10.4357, "step": 101, "task_loss": 4.610260009765625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.80290412902832, "epoch": 0.09, "learning_rate": 2.8732394366197183e-06, "loss": 10.2376, "step": 102, "task_loss": 4.529834747314453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.406658172607422, "epoch": 0.09, "learning_rate": 2.9014084507042255e-06, "loss": 10.1504, "step": 103, "task_loss": 4.560537338256836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.819093704223633, "epoch": 0.09, "learning_rate": 2.9295774647887324e-06, "loss": 10.3824, "step": 104, "task_loss": 4.6420440673828125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.025821685791016, "epoch": 0.09, "learning_rate": 2.9577464788732396e-06, "loss": 10.6693, "step": 105, "task_loss": 4.599424839019775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.691964149475098, "epoch": 0.09, "learning_rate": 2.985915492957747e-06, "loss": 11.2052, "step": 106, "task_loss": 4.664707183837891 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.457517623901367, "epoch": 0.09, "learning_rate": 3.0140845070422537e-06, "loss": 10.879, "step": 107, "task_loss": 4.648604869842529 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.926033020019531, "epoch": 0.09, "learning_rate": 3.0422535211267605e-06, "loss": 10.9723, "step": 108, "task_loss": 4.60695219039917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.213266372680664, "epoch": 0.09, "learning_rate": 3.0704225352112678e-06, "loss": 10.3091, "step": 109, "task_loss": 4.490808486938477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.98620891571045, "epoch": 0.09, "learning_rate": 3.098591549295775e-06, "loss": 11.0937, "step": 110, "task_loss": 4.5143022537231445 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.129890441894531, "epoch": 0.09, "learning_rate": 3.1267605633802823e-06, "loss": 9.941, "step": 111, "task_loss": 4.466829776763916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.640400886535645, "epoch": 0.09, "learning_rate": 3.1549295774647887e-06, "loss": 10.231, "step": 112, "task_loss": 4.725858688354492 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.465828895568848, "epoch": 0.1, "learning_rate": 3.183098591549296e-06, "loss": 11.0722, "step": 113, "task_loss": 4.593950271606445 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.677474975585938, "epoch": 0.1, "learning_rate": 3.211267605633803e-06, "loss": 11.2786, "step": 114, "task_loss": 4.545686721801758 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.140052795410156, "epoch": 0.1, "learning_rate": 3.23943661971831e-06, "loss": 10.1499, "step": 115, "task_loss": 4.513463020324707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.159743309020996, "epoch": 0.1, "learning_rate": 3.2676056338028173e-06, "loss": 10.5326, "step": 116, "task_loss": 4.595307350158691 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.306463241577148, "epoch": 0.1, "learning_rate": 3.2957746478873245e-06, "loss": 10.7601, "step": 117, "task_loss": 4.547955513000488 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.017741203308105, "epoch": 0.1, "learning_rate": 3.323943661971831e-06, "loss": 11.0774, "step": 118, "task_loss": 4.582327365875244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.738935470581055, "epoch": 0.1, "learning_rate": 3.352112676056338e-06, "loss": 11.4335, "step": 119, "task_loss": 4.462286949157715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.039283752441406, "epoch": 0.1, "learning_rate": 3.380281690140845e-06, "loss": 10.5954, "step": 120, "task_loss": 4.519924163818359 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.270971298217773, "epoch": 0.1, "learning_rate": 3.4084507042253522e-06, "loss": 10.5582, "step": 121, "task_loss": 4.577351093292236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.93690299987793, "epoch": 0.1, "learning_rate": 3.4366197183098595e-06, "loss": 10.4955, "step": 122, "task_loss": 4.511572360992432 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.33281135559082, "epoch": 0.1, "learning_rate": 3.464788732394366e-06, "loss": 10.7102, "step": 123, "task_loss": 4.467489242553711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.276394844055176, "epoch": 0.1, "learning_rate": 3.492957746478873e-06, "loss": 10.3499, "step": 124, "task_loss": 4.63872766494751 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.115531921386719, "epoch": 0.11, "learning_rate": 3.521126760563381e-06, "loss": 10.467, "step": 125, "task_loss": 4.474095344543457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.955999374389648, "epoch": 0.11, "learning_rate": 3.549295774647887e-06, "loss": 10.6859, "step": 126, "task_loss": 4.427630424499512 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.615371704101562, "epoch": 0.11, "learning_rate": 3.5774647887323945e-06, "loss": 10.9464, "step": 127, "task_loss": 4.611941814422607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.313074111938477, "epoch": 0.11, "learning_rate": 3.6056338028169017e-06, "loss": 10.0694, "step": 128, "task_loss": 4.407715797424316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.80350112915039, "epoch": 0.11, "learning_rate": 3.6338028169014085e-06, "loss": 10.1331, "step": 129, "task_loss": 4.603161334991455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.506087303161621, "epoch": 0.11, "learning_rate": 3.6619718309859158e-06, "loss": 10.7389, "step": 130, "task_loss": 4.655784606933594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.683990478515625, "epoch": 0.11, "learning_rate": 3.690140845070423e-06, "loss": 11.0823, "step": 131, "task_loss": 4.5683441162109375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.482134819030762, "epoch": 0.11, "learning_rate": 3.7183098591549294e-06, "loss": 10.9517, "step": 132, "task_loss": 4.491758823394775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.356341361999512, "epoch": 0.11, "learning_rate": 3.7464788732394367e-06, "loss": 10.4455, "step": 133, "task_loss": 4.5772528648376465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.916454315185547, "epoch": 0.11, "learning_rate": 3.774647887323944e-06, "loss": 10.0334, "step": 134, "task_loss": 4.490162372589111 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.858617782592773, "epoch": 0.11, "learning_rate": 3.8028169014084508e-06, "loss": 11.2223, "step": 135, "task_loss": 4.561485767364502 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.848136901855469, "epoch": 0.11, "learning_rate": 3.830985915492958e-06, "loss": 10.651, "step": 136, "task_loss": 4.574880123138428 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.8009033203125, "epoch": 0.12, "learning_rate": 3.859154929577465e-06, "loss": 10.6488, "step": 137, "task_loss": 4.586858749389648 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.848316192626953, "epoch": 0.12, "learning_rate": 3.887323943661972e-06, "loss": 10.1914, "step": 138, "task_loss": 4.651371955871582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.536291122436523, "epoch": 0.12, "learning_rate": 3.915492957746479e-06, "loss": 11.2229, "step": 139, "task_loss": 4.531460285186768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.782855987548828, "epoch": 0.12, "learning_rate": 3.943661971830986e-06, "loss": 11.1588, "step": 140, "task_loss": 4.547407627105713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.685376167297363, "epoch": 0.12, "learning_rate": 3.9718309859154926e-06, "loss": 9.8536, "step": 141, "task_loss": 4.637712478637695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.340768814086914, "epoch": 0.12, "learning_rate": 4.000000000000001e-06, "loss": 10.2679, "step": 142, "task_loss": 4.549135684967041 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.418891906738281, "epoch": 0.12, "learning_rate": 4.028169014084508e-06, "loss": 11.1923, "step": 143, "task_loss": 4.4391984939575195 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.523265838623047, "epoch": 0.12, "learning_rate": 4.056338028169014e-06, "loss": 11.1801, "step": 144, "task_loss": 4.631706237792969 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.578893661499023, "epoch": 0.12, "learning_rate": 4.0845070422535216e-06, "loss": 11.0093, "step": 145, "task_loss": 4.563064098358154 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.25750732421875, "epoch": 0.12, "learning_rate": 4.112676056338028e-06, "loss": 10.8373, "step": 146, "task_loss": 4.479740619659424 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.698243141174316, "epoch": 0.12, "learning_rate": 4.140845070422535e-06, "loss": 10.6484, "step": 147, "task_loss": 4.495515823364258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.675031661987305, "epoch": 0.13, "learning_rate": 4.1690140845070425e-06, "loss": 10.4258, "step": 148, "task_loss": 4.516814708709717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.315908432006836, "epoch": 0.13, "learning_rate": 4.197183098591549e-06, "loss": 10.3745, "step": 149, "task_loss": 4.388041973114014 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.79265022277832, "epoch": 0.13, "learning_rate": 4.225352112676056e-06, "loss": 11.0548, "step": 150, "task_loss": 4.583268642425537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.573726654052734, "epoch": 0.13, "learning_rate": 4.253521126760563e-06, "loss": 9.6913, "step": 151, "task_loss": 4.647573471069336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.413125991821289, "epoch": 0.13, "learning_rate": 4.281690140845071e-06, "loss": 10.7727, "step": 152, "task_loss": 4.496472358703613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.03062629699707, "epoch": 0.13, "learning_rate": 4.309859154929578e-06, "loss": 10.4948, "step": 153, "task_loss": 4.47908878326416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.084280967712402, "epoch": 0.13, "learning_rate": 4.338028169014085e-06, "loss": 10.2389, "step": 154, "task_loss": 4.515031337738037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.07840633392334, "epoch": 0.13, "learning_rate": 4.3661971830985915e-06, "loss": 10.2652, "step": 155, "task_loss": 4.4497971534729 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.853221893310547, "epoch": 0.13, "learning_rate": 4.394366197183099e-06, "loss": 10.3585, "step": 156, "task_loss": 4.353901386260986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.914134979248047, "epoch": 0.13, "learning_rate": 4.422535211267606e-06, "loss": 10.0713, "step": 157, "task_loss": 4.575733184814453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.544150352478027, "epoch": 0.13, "learning_rate": 4.4507042253521124e-06, "loss": 10.7172, "step": 158, "task_loss": 4.420012474060059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.307382583618164, "epoch": 0.13, "learning_rate": 4.47887323943662e-06, "loss": 10.8168, "step": 159, "task_loss": 4.413660526275635 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.920637130737305, "epoch": 0.14, "learning_rate": 4.507042253521127e-06, "loss": 10.5731, "step": 160, "task_loss": 4.3378825187683105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.677581787109375, "epoch": 0.14, "learning_rate": 4.535211267605634e-06, "loss": 10.5923, "step": 161, "task_loss": 4.445382595062256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.677480697631836, "epoch": 0.14, "learning_rate": 4.5633802816901414e-06, "loss": 11.2358, "step": 162, "task_loss": 4.370922565460205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.014175415039062, "epoch": 0.14, "learning_rate": 4.591549295774649e-06, "loss": 11.0791, "step": 163, "task_loss": 4.673698902130127 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.011930465698242, "epoch": 0.14, "learning_rate": 4.619718309859155e-06, "loss": 10.2401, "step": 164, "task_loss": 4.459265232086182 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.024348258972168, "epoch": 0.14, "learning_rate": 4.647887323943662e-06, "loss": 10.745, "step": 165, "task_loss": 4.377274990081787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.38964557647705, "epoch": 0.14, "learning_rate": 4.67605633802817e-06, "loss": 11.0233, "step": 166, "task_loss": 4.407256126403809 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.651541709899902, "epoch": 0.14, "learning_rate": 4.704225352112676e-06, "loss": 10.6303, "step": 167, "task_loss": 4.491535663604736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.352401733398438, "epoch": 0.14, "learning_rate": 4.732394366197183e-06, "loss": 9.8787, "step": 168, "task_loss": 4.362551689147949 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.053401947021484, "epoch": 0.14, "learning_rate": 4.7605633802816905e-06, "loss": 11.468, "step": 169, "task_loss": 4.321714401245117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.11303997039795, "epoch": 0.14, "learning_rate": 4.788732394366197e-06, "loss": 11.5697, "step": 170, "task_loss": 4.406832695007324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.285576820373535, "epoch": 0.14, "learning_rate": 4.816901408450705e-06, "loss": 10.5944, "step": 171, "task_loss": 4.349706649780273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.672544479370117, "epoch": 0.15, "learning_rate": 4.845070422535212e-06, "loss": 11.4187, "step": 172, "task_loss": 4.280782699584961 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.2306489944458, "epoch": 0.15, "learning_rate": 4.873239436619719e-06, "loss": 10.7378, "step": 173, "task_loss": 4.378617763519287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.238370895385742, "epoch": 0.15, "learning_rate": 4.901408450704226e-06, "loss": 9.673, "step": 174, "task_loss": 4.414499282836914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.979421615600586, "epoch": 0.15, "learning_rate": 4.929577464788732e-06, "loss": 11.127, "step": 175, "task_loss": 4.457372188568115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.48719596862793, "epoch": 0.15, "learning_rate": 4.9577464788732395e-06, "loss": 10.3078, "step": 176, "task_loss": 4.589346408843994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.89100170135498, "epoch": 0.15, "learning_rate": 4.985915492957747e-06, "loss": 9.451, "step": 177, "task_loss": 4.415113925933838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.30606746673584, "epoch": 0.15, "learning_rate": 5.014084507042253e-06, "loss": 10.3613, "step": 178, "task_loss": 4.371075630187988 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.752979278564453, "epoch": 0.15, "learning_rate": 5.0422535211267604e-06, "loss": 10.5584, "step": 179, "task_loss": 4.4004621505737305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.77967357635498, "epoch": 0.15, "learning_rate": 5.070422535211268e-06, "loss": 9.6219, "step": 180, "task_loss": 4.3466620445251465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.760498046875, "epoch": 0.15, "learning_rate": 5.098591549295775e-06, "loss": 10.3036, "step": 181, "task_loss": 4.416411876678467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.75898265838623, "epoch": 0.15, "learning_rate": 5.126760563380282e-06, "loss": 10.6311, "step": 182, "task_loss": 4.370694160461426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.72964096069336, "epoch": 0.15, "learning_rate": 5.1549295774647894e-06, "loss": 10.2341, "step": 183, "task_loss": 4.471965789794922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.091055870056152, "epoch": 0.16, "learning_rate": 5.183098591549296e-06, "loss": 10.0255, "step": 184, "task_loss": 4.4736785888671875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.57967758178711, "epoch": 0.16, "learning_rate": 5.211267605633803e-06, "loss": 10.3222, "step": 185, "task_loss": 4.55633544921875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.28968620300293, "epoch": 0.16, "learning_rate": 5.23943661971831e-06, "loss": 9.8487, "step": 186, "task_loss": 4.16321325302124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.754732131958008, "epoch": 0.16, "learning_rate": 5.267605633802817e-06, "loss": 10.9906, "step": 187, "task_loss": 4.37198543548584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.801636695861816, "epoch": 0.16, "learning_rate": 5.295774647887324e-06, "loss": 11.022, "step": 188, "task_loss": 4.3986430168151855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.378889083862305, "epoch": 0.16, "learning_rate": 5.323943661971831e-06, "loss": 10.438, "step": 189, "task_loss": 4.494696617126465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.736745834350586, "epoch": 0.16, "learning_rate": 5.3521126760563385e-06, "loss": 10.7332, "step": 190, "task_loss": 4.18865442276001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 13.501799583435059, "epoch": 0.16, "learning_rate": 5.380281690140846e-06, "loss": 11.8028, "step": 191, "task_loss": 4.429253578186035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.79987907409668, "epoch": 0.16, "learning_rate": 5.408450704225353e-06, "loss": 10.9764, "step": 192, "task_loss": 4.41756010055542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.542933464050293, "epoch": 0.16, "learning_rate": 5.436619718309859e-06, "loss": 10.2711, "step": 193, "task_loss": 4.3243327140808105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.842691421508789, "epoch": 0.16, "learning_rate": 5.464788732394367e-06, "loss": 10.1009, "step": 194, "task_loss": 4.55487060546875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.616206169128418, "epoch": 0.16, "learning_rate": 5.492957746478874e-06, "loss": 10.7875, "step": 195, "task_loss": 4.329843044281006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.590604782104492, "epoch": 0.17, "learning_rate": 5.52112676056338e-06, "loss": 10.3911, "step": 196, "task_loss": 4.5238938331604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.000479698181152, "epoch": 0.17, "learning_rate": 5.5492957746478875e-06, "loss": 10.4168, "step": 197, "task_loss": 4.3793134689331055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.460063934326172, "epoch": 0.17, "learning_rate": 5.577464788732395e-06, "loss": 10.1978, "step": 198, "task_loss": 4.384171485900879 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.920083999633789, "epoch": 0.17, "learning_rate": 5.605633802816901e-06, "loss": 10.3119, "step": 199, "task_loss": 4.390791893005371 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.767108917236328, "epoch": 0.17, "learning_rate": 5.6338028169014084e-06, "loss": 10.2315, "step": 200, "task_loss": 4.503992557525635 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.830652236938477, "epoch": 0.17, "learning_rate": 5.661971830985916e-06, "loss": 11.2811, "step": 201, "task_loss": 4.384411334991455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.112621307373047, "epoch": 0.17, "learning_rate": 5.690140845070423e-06, "loss": 10.383, "step": 202, "task_loss": 4.314823150634766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.200511932373047, "epoch": 0.17, "learning_rate": 5.71830985915493e-06, "loss": 10.2679, "step": 203, "task_loss": 4.3806867599487305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 12.78781509399414, "epoch": 0.17, "learning_rate": 5.746478873239437e-06, "loss": 10.8044, "step": 204, "task_loss": 4.352518558502197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.330766677856445, "epoch": 0.17, "learning_rate": 5.774647887323944e-06, "loss": 10.2156, "step": 205, "task_loss": 4.301047325134277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.77873420715332, "epoch": 0.17, "learning_rate": 5.802816901408451e-06, "loss": 9.8538, "step": 206, "task_loss": 4.104227542877197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.026676177978516, "epoch": 0.17, "learning_rate": 5.8309859154929575e-06, "loss": 9.9709, "step": 207, "task_loss": 4.47815465927124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.114535331726074, "epoch": 0.18, "learning_rate": 5.859154929577465e-06, "loss": 9.5795, "step": 208, "task_loss": 4.561560153961182 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.44635009765625, "epoch": 0.18, "learning_rate": 5.887323943661972e-06, "loss": 10.4955, "step": 209, "task_loss": 4.331540584564209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.757562637329102, "epoch": 0.18, "learning_rate": 5.915492957746479e-06, "loss": 10.5591, "step": 210, "task_loss": 4.241861343383789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.213203430175781, "epoch": 0.18, "learning_rate": 5.9436619718309865e-06, "loss": 9.7532, "step": 211, "task_loss": 4.209874629974365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.92913818359375, "epoch": 0.18, "learning_rate": 5.971830985915494e-06, "loss": 10.6379, "step": 212, "task_loss": 4.351600170135498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.87527847290039, "epoch": 0.18, "learning_rate": 6e-06, "loss": 10.2812, "step": 213, "task_loss": 4.040771961212158 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.90794563293457, "epoch": 0.18, "learning_rate": 6.028169014084507e-06, "loss": 9.5434, "step": 214, "task_loss": 4.383078575134277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.192621231079102, "epoch": 0.18, "learning_rate": 6.056338028169015e-06, "loss": 10.1422, "step": 215, "task_loss": 4.140993595123291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.495984077453613, "epoch": 0.18, "learning_rate": 6.084507042253521e-06, "loss": 10.5979, "step": 216, "task_loss": 4.143148899078369 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.57532024383545, "epoch": 0.18, "learning_rate": 6.112676056338028e-06, "loss": 9.2703, "step": 217, "task_loss": 4.275446891784668 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.774991989135742, "epoch": 0.18, "learning_rate": 6.1408450704225356e-06, "loss": 9.6062, "step": 218, "task_loss": 4.373274803161621 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.115221977233887, "epoch": 0.19, "learning_rate": 6.169014084507042e-06, "loss": 9.1257, "step": 219, "task_loss": 4.054250240325928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.839506149291992, "epoch": 0.19, "learning_rate": 6.19718309859155e-06, "loss": 9.6512, "step": 220, "task_loss": 4.007802963256836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.441226959228516, "epoch": 0.19, "learning_rate": 6.225352112676057e-06, "loss": 10.3522, "step": 221, "task_loss": 4.351891994476318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.545177459716797, "epoch": 0.19, "learning_rate": 6.2535211267605646e-06, "loss": 9.6603, "step": 222, "task_loss": 3.86029052734375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.518407821655273, "epoch": 0.19, "learning_rate": 6.28169014084507e-06, "loss": 9.3919, "step": 223, "task_loss": 4.194509983062744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.588296890258789, "epoch": 0.19, "learning_rate": 6.309859154929577e-06, "loss": 9.7471, "step": 224, "task_loss": 4.295816898345947 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.81348705291748, "epoch": 0.19, "learning_rate": 6.338028169014085e-06, "loss": 9.6351, "step": 225, "task_loss": 3.8114185333251953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.824386596679688, "epoch": 0.19, "learning_rate": 6.366197183098592e-06, "loss": 10.7292, "step": 226, "task_loss": 4.024959087371826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.342994689941406, "epoch": 0.19, "learning_rate": 6.394366197183099e-06, "loss": 10.4315, "step": 227, "task_loss": 4.115330219268799 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.08212661743164, "epoch": 0.19, "learning_rate": 6.422535211267606e-06, "loss": 10.1037, "step": 228, "task_loss": 4.2968668937683105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.851308822631836, "epoch": 0.19, "learning_rate": 6.450704225352113e-06, "loss": 9.5469, "step": 229, "task_loss": 4.182309150695801 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.134824752807617, "epoch": 0.19, "learning_rate": 6.47887323943662e-06, "loss": 9.9529, "step": 230, "task_loss": 4.4332380294799805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.498416900634766, "epoch": 0.2, "learning_rate": 6.507042253521127e-06, "loss": 10.3518, "step": 231, "task_loss": 4.068986892700195 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.684327125549316, "epoch": 0.2, "learning_rate": 6.5352112676056345e-06, "loss": 9.9438, "step": 232, "task_loss": 4.112462520599365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.226621627807617, "epoch": 0.2, "learning_rate": 6.563380281690142e-06, "loss": 10.0708, "step": 233, "task_loss": 4.249608516693115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.985512733459473, "epoch": 0.2, "learning_rate": 6.591549295774649e-06, "loss": 9.872, "step": 234, "task_loss": 3.784219741821289 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.859283447265625, "epoch": 0.2, "learning_rate": 6.6197183098591546e-06, "loss": 9.9938, "step": 235, "task_loss": 4.243819236755371 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.167388916015625, "epoch": 0.2, "learning_rate": 6.647887323943662e-06, "loss": 9.667, "step": 236, "task_loss": 4.195701599121094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.539896965026855, "epoch": 0.2, "learning_rate": 6.676056338028169e-06, "loss": 9.8673, "step": 237, "task_loss": 4.219735145568848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.817333221435547, "epoch": 0.2, "learning_rate": 6.704225352112676e-06, "loss": 9.5121, "step": 238, "task_loss": 4.044241905212402 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.902064323425293, "epoch": 0.2, "learning_rate": 6.7323943661971836e-06, "loss": 10.0664, "step": 239, "task_loss": 4.094220161437988 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.120040893554688, "epoch": 0.2, "learning_rate": 6.76056338028169e-06, "loss": 9.6749, "step": 240, "task_loss": 4.270837783813477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.744121551513672, "epoch": 0.2, "learning_rate": 6.788732394366197e-06, "loss": 9.9371, "step": 241, "task_loss": 4.072160720825195 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.062346458435059, "epoch": 0.2, "learning_rate": 6.8169014084507045e-06, "loss": 9.9015, "step": 242, "task_loss": 4.017858505249023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.848584175109863, "epoch": 0.21, "learning_rate": 6.845070422535212e-06, "loss": 9.6946, "step": 243, "task_loss": 4.02824068069458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.353474617004395, "epoch": 0.21, "learning_rate": 6.873239436619719e-06, "loss": 10.0184, "step": 244, "task_loss": 4.092763900756836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.516221046447754, "epoch": 0.21, "learning_rate": 6.901408450704226e-06, "loss": 9.4275, "step": 245, "task_loss": 3.878628730773926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.124242782592773, "epoch": 0.21, "learning_rate": 6.929577464788732e-06, "loss": 9.845, "step": 246, "task_loss": 3.8637027740478516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.710416793823242, "epoch": 0.21, "learning_rate": 6.957746478873239e-06, "loss": 10.0719, "step": 247, "task_loss": 3.8311970233917236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.422137260437012, "epoch": 0.21, "learning_rate": 6.985915492957746e-06, "loss": 9.8397, "step": 248, "task_loss": 3.901399850845337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.122580528259277, "epoch": 0.21, "learning_rate": 7.014084507042254e-06, "loss": 10.5474, "step": 249, "task_loss": 3.919623613357544 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.491177558898926, "epoch": 0.21, "learning_rate": 7.042253521126762e-06, "loss": 10.127, "step": 250, "task_loss": 4.0034356117248535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.696395874023438, "epoch": 0.21, "learning_rate": 7.070422535211269e-06, "loss": 9.1054, "step": 251, "task_loss": 3.9972379207611084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.478248596191406, "epoch": 0.21, "learning_rate": 7.098591549295774e-06, "loss": 10.1703, "step": 252, "task_loss": 4.159002304077148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.108709335327148, "epoch": 0.21, "learning_rate": 7.126760563380282e-06, "loss": 8.2715, "step": 253, "task_loss": 3.971951723098755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.508358001708984, "epoch": 0.21, "learning_rate": 7.154929577464789e-06, "loss": 9.7633, "step": 254, "task_loss": 4.063025951385498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.901662826538086, "epoch": 0.22, "learning_rate": 7.183098591549296e-06, "loss": 9.6771, "step": 255, "task_loss": 3.71547532081604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.490808486938477, "epoch": 0.22, "learning_rate": 7.211267605633803e-06, "loss": 9.8222, "step": 256, "task_loss": 4.068603992462158 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.375261306762695, "epoch": 0.22, "learning_rate": 7.239436619718311e-06, "loss": 9.2053, "step": 257, "task_loss": 3.844226837158203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.037941932678223, "epoch": 0.22, "learning_rate": 7.267605633802817e-06, "loss": 9.0197, "step": 258, "task_loss": 4.190129280090332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.490335464477539, "epoch": 0.22, "learning_rate": 7.295774647887324e-06, "loss": 8.6459, "step": 259, "task_loss": 4.026490211486816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.559481620788574, "epoch": 0.22, "learning_rate": 7.3239436619718316e-06, "loss": 9.8627, "step": 260, "task_loss": 3.555706024169922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.304780006408691, "epoch": 0.22, "learning_rate": 7.352112676056339e-06, "loss": 9.7161, "step": 261, "task_loss": 3.4398550987243652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.942546844482422, "epoch": 0.22, "learning_rate": 7.380281690140846e-06, "loss": 9.9393, "step": 262, "task_loss": 3.7509469985961914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.587044715881348, "epoch": 0.22, "learning_rate": 7.408450704225353e-06, "loss": 9.0398, "step": 263, "task_loss": 3.8700292110443115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.495553970336914, "epoch": 0.22, "learning_rate": 7.436619718309859e-06, "loss": 9.9945, "step": 264, "task_loss": 3.711794853210449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.818904876708984, "epoch": 0.22, "learning_rate": 7.464788732394366e-06, "loss": 9.3025, "step": 265, "task_loss": 3.944291591644287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.385010719299316, "epoch": 0.22, "learning_rate": 7.492957746478873e-06, "loss": 8.954, "step": 266, "task_loss": 3.5303120613098145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.85380744934082, "epoch": 0.23, "learning_rate": 7.521126760563381e-06, "loss": 9.0208, "step": 267, "task_loss": 4.139451026916504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.270426750183105, "epoch": 0.23, "learning_rate": 7.549295774647888e-06, "loss": 9.3873, "step": 268, "task_loss": 3.715198278427124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.696300506591797, "epoch": 0.23, "learning_rate": 7.577464788732394e-06, "loss": 8.6848, "step": 269, "task_loss": 3.6412222385406494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.976898193359375, "epoch": 0.23, "learning_rate": 7.6056338028169015e-06, "loss": 9.6739, "step": 270, "task_loss": 3.4873697757720947 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.14134407043457, "epoch": 0.23, "learning_rate": 7.633802816901409e-06, "loss": 8.9568, "step": 271, "task_loss": 3.926787853240967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.90835189819336, "epoch": 0.23, "learning_rate": 7.661971830985916e-06, "loss": 9.0015, "step": 272, "task_loss": 3.7651710510253906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.074337005615234, "epoch": 0.23, "learning_rate": 7.690140845070423e-06, "loss": 9.7344, "step": 273, "task_loss": 3.8010449409484863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.617801666259766, "epoch": 0.23, "learning_rate": 7.71830985915493e-06, "loss": 9.7256, "step": 274, "task_loss": 3.441756248474121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.971430778503418, "epoch": 0.23, "learning_rate": 7.746478873239436e-06, "loss": 10.0853, "step": 275, "task_loss": 3.8108794689178467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.826993942260742, "epoch": 0.23, "learning_rate": 7.774647887323943e-06, "loss": 9.1404, "step": 276, "task_loss": 3.615917682647705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 11.469215393066406, "epoch": 0.23, "learning_rate": 7.80281690140845e-06, "loss": 9.4083, "step": 277, "task_loss": 3.83221173286438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.845389366149902, "epoch": 0.23, "learning_rate": 7.830985915492958e-06, "loss": 9.3367, "step": 278, "task_loss": 3.8096020221710205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.282831192016602, "epoch": 0.24, "learning_rate": 7.859154929577465e-06, "loss": 9.475, "step": 279, "task_loss": 3.779928684234619 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.57934856414795, "epoch": 0.24, "learning_rate": 7.887323943661972e-06, "loss": 9.4847, "step": 280, "task_loss": 3.735959529876709 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.969745635986328, "epoch": 0.24, "learning_rate": 7.915492957746478e-06, "loss": 9.1712, "step": 281, "task_loss": 3.793896198272705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.756889343261719, "epoch": 0.24, "learning_rate": 7.943661971830985e-06, "loss": 9.5545, "step": 282, "task_loss": 3.868887424468994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.840197563171387, "epoch": 0.24, "learning_rate": 7.971830985915494e-06, "loss": 9.1474, "step": 283, "task_loss": 3.5532660484313965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.261518478393555, "epoch": 0.24, "learning_rate": 8.000000000000001e-06, "loss": 8.8779, "step": 284, "task_loss": 3.4109644889831543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.404696464538574, "epoch": 0.24, "learning_rate": 8.028169014084509e-06, "loss": 8.6047, "step": 285, "task_loss": 3.8750998973846436 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.01717472076416, "epoch": 0.24, "learning_rate": 8.056338028169016e-06, "loss": 8.8935, "step": 286, "task_loss": 3.4817116260528564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.608448028564453, "epoch": 0.24, "learning_rate": 8.084507042253521e-06, "loss": 9.083, "step": 287, "task_loss": 3.6901144981384277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.794292449951172, "epoch": 0.24, "learning_rate": 8.112676056338029e-06, "loss": 8.8886, "step": 288, "task_loss": 3.8568289279937744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.6219635009765625, "epoch": 0.24, "learning_rate": 8.140845070422536e-06, "loss": 8.0855, "step": 289, "task_loss": 3.8076083660125732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.150201797485352, "epoch": 0.24, "learning_rate": 8.169014084507043e-06, "loss": 8.4486, "step": 290, "task_loss": 3.328174591064453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.08449935913086, "epoch": 0.25, "learning_rate": 8.19718309859155e-06, "loss": 8.6462, "step": 291, "task_loss": 3.8236074447631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.212358474731445, "epoch": 0.25, "learning_rate": 8.225352112676056e-06, "loss": 8.6692, "step": 292, "task_loss": 3.365229606628418 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.166473388671875, "epoch": 0.25, "learning_rate": 8.253521126760563e-06, "loss": 8.1912, "step": 293, "task_loss": 3.2412304878234863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.429041862487793, "epoch": 0.25, "learning_rate": 8.28169014084507e-06, "loss": 8.6753, "step": 294, "task_loss": 3.43083119392395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.556553840637207, "epoch": 0.25, "learning_rate": 8.309859154929578e-06, "loss": 8.7377, "step": 295, "task_loss": 3.623307466506958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.763278007507324, "epoch": 0.25, "learning_rate": 8.338028169014085e-06, "loss": 8.6758, "step": 296, "task_loss": 3.8599281311035156 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.50567626953125, "epoch": 0.25, "learning_rate": 8.366197183098592e-06, "loss": 8.2804, "step": 297, "task_loss": 3.5250017642974854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.019044876098633, "epoch": 0.25, "learning_rate": 8.394366197183098e-06, "loss": 8.4272, "step": 298, "task_loss": 3.287724256515503 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.530113220214844, "epoch": 0.25, "learning_rate": 8.422535211267605e-06, "loss": 8.5335, "step": 299, "task_loss": 3.365999460220337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.690508842468262, "epoch": 0.25, "learning_rate": 8.450704225352112e-06, "loss": 8.8478, "step": 300, "task_loss": 3.702277898788452 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.846294403076172, "epoch": 0.25, "learning_rate": 8.47887323943662e-06, "loss": 8.743, "step": 301, "task_loss": 3.1246697902679443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.066999435424805, "epoch": 0.26, "learning_rate": 8.507042253521127e-06, "loss": 8.939, "step": 302, "task_loss": 3.4893457889556885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.889676094055176, "epoch": 0.26, "learning_rate": 8.535211267605634e-06, "loss": 8.1868, "step": 303, "task_loss": 3.2983694076538086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.633882522583008, "epoch": 0.26, "learning_rate": 8.563380281690141e-06, "loss": 8.5168, "step": 304, "task_loss": 3.6111299991607666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.143087387084961, "epoch": 0.26, "learning_rate": 8.591549295774648e-06, "loss": 8.9639, "step": 305, "task_loss": 3.544762134552002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.48599624633789, "epoch": 0.26, "learning_rate": 8.619718309859156e-06, "loss": 8.6602, "step": 306, "task_loss": 2.927975654602051 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.809699058532715, "epoch": 0.26, "learning_rate": 8.647887323943663e-06, "loss": 8.6622, "step": 307, "task_loss": 2.973365545272827 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.480162143707275, "epoch": 0.26, "learning_rate": 8.67605633802817e-06, "loss": 8.059, "step": 308, "task_loss": 3.3903987407684326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.841232776641846, "epoch": 0.26, "learning_rate": 8.704225352112677e-06, "loss": 8.8033, "step": 309, "task_loss": 3.6406338214874268 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.625747680664062, "epoch": 0.26, "learning_rate": 8.732394366197183e-06, "loss": 7.4533, "step": 310, "task_loss": 3.4544565677642822 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.78294849395752, "epoch": 0.26, "learning_rate": 8.76056338028169e-06, "loss": 8.392, "step": 311, "task_loss": 3.6611454486846924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.210204124450684, "epoch": 0.26, "learning_rate": 8.788732394366198e-06, "loss": 8.7675, "step": 312, "task_loss": 3.154282331466675 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.979257583618164, "epoch": 0.26, "learning_rate": 8.816901408450705e-06, "loss": 8.4982, "step": 313, "task_loss": 3.34869384765625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 10.051904678344727, "epoch": 0.27, "learning_rate": 8.845070422535212e-06, "loss": 8.2877, "step": 314, "task_loss": 3.707380771636963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.192636489868164, "epoch": 0.27, "learning_rate": 8.87323943661972e-06, "loss": 7.9243, "step": 315, "task_loss": 3.2126760482788086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.602607727050781, "epoch": 0.27, "learning_rate": 8.901408450704225e-06, "loss": 8.735, "step": 316, "task_loss": 3.2603938579559326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.971074104309082, "epoch": 0.27, "learning_rate": 8.929577464788732e-06, "loss": 8.2543, "step": 317, "task_loss": 3.4197936058044434 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.479056358337402, "epoch": 0.27, "learning_rate": 8.95774647887324e-06, "loss": 7.7128, "step": 318, "task_loss": 3.061222553253174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.430188179016113, "epoch": 0.27, "learning_rate": 8.985915492957747e-06, "loss": 8.6526, "step": 319, "task_loss": 3.7191739082336426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.039786338806152, "epoch": 0.27, "learning_rate": 9.014084507042254e-06, "loss": 8.8419, "step": 320, "task_loss": 3.383272886276245 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.464742660522461, "epoch": 0.27, "learning_rate": 9.042253521126761e-06, "loss": 7.5427, "step": 321, "task_loss": 3.3240768909454346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.464814186096191, "epoch": 0.27, "learning_rate": 9.070422535211268e-06, "loss": 8.1164, "step": 322, "task_loss": 2.848857879638672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.866616725921631, "epoch": 0.27, "learning_rate": 9.098591549295776e-06, "loss": 8.6757, "step": 323, "task_loss": 3.534390449523926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.659099578857422, "epoch": 0.27, "learning_rate": 9.126760563380283e-06, "loss": 7.6514, "step": 324, "task_loss": 2.9656589031219482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.151924133300781, "epoch": 0.27, "learning_rate": 9.15492957746479e-06, "loss": 8.3515, "step": 325, "task_loss": 2.939431667327881 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.243948459625244, "epoch": 0.28, "learning_rate": 9.183098591549297e-06, "loss": 7.703, "step": 326, "task_loss": 3.165945053100586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.604759216308594, "epoch": 0.28, "learning_rate": 9.211267605633803e-06, "loss": 8.1792, "step": 327, "task_loss": 3.1140804290771484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.991395950317383, "epoch": 0.28, "learning_rate": 9.23943661971831e-06, "loss": 7.9834, "step": 328, "task_loss": 3.0847387313842773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.874151229858398, "epoch": 0.28, "learning_rate": 9.267605633802817e-06, "loss": 8.1444, "step": 329, "task_loss": 3.2763946056365967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.521008491516113, "epoch": 0.28, "learning_rate": 9.295774647887325e-06, "loss": 7.7667, "step": 330, "task_loss": 2.8071932792663574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.044706344604492, "epoch": 0.28, "learning_rate": 9.323943661971832e-06, "loss": 7.58, "step": 331, "task_loss": 2.8021326065063477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.263099670410156, "epoch": 0.28, "learning_rate": 9.35211267605634e-06, "loss": 7.9718, "step": 332, "task_loss": 3.213937759399414 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.965255737304688, "epoch": 0.28, "learning_rate": 9.380281690140845e-06, "loss": 8.3957, "step": 333, "task_loss": 2.7993907928466797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.224508285522461, "epoch": 0.28, "learning_rate": 9.408450704225352e-06, "loss": 7.7606, "step": 334, "task_loss": 3.2044272422790527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.041767120361328, "epoch": 0.28, "learning_rate": 9.43661971830986e-06, "loss": 8.744, "step": 335, "task_loss": 3.2426023483276367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.501925468444824, "epoch": 0.28, "learning_rate": 9.464788732394366e-06, "loss": 7.8225, "step": 336, "task_loss": 3.0217390060424805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.55257511138916, "epoch": 0.28, "learning_rate": 9.492957746478874e-06, "loss": 7.49, "step": 337, "task_loss": 3.094108819961548 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.507192134857178, "epoch": 0.29, "learning_rate": 9.521126760563381e-06, "loss": 7.8533, "step": 338, "task_loss": 2.891679525375366 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.00389575958252, "epoch": 0.29, "learning_rate": 9.549295774647887e-06, "loss": 7.632, "step": 339, "task_loss": 3.2321176528930664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.377330303192139, "epoch": 0.29, "learning_rate": 9.577464788732394e-06, "loss": 7.3427, "step": 340, "task_loss": 3.0627033710479736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.172029495239258, "epoch": 0.29, "learning_rate": 9.605633802816901e-06, "loss": 7.5764, "step": 341, "task_loss": 3.1100919246673584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.69929027557373, "epoch": 0.29, "learning_rate": 9.63380281690141e-06, "loss": 7.8047, "step": 342, "task_loss": 3.3254153728485107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.6726884841918945, "epoch": 0.29, "learning_rate": 9.661971830985917e-06, "loss": 7.8201, "step": 343, "task_loss": 3.1534392833709717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.117541313171387, "epoch": 0.29, "learning_rate": 9.690140845070424e-06, "loss": 7.8886, "step": 344, "task_loss": 3.047461748123169 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.083487510681152, "epoch": 0.29, "learning_rate": 9.71830985915493e-06, "loss": 7.4447, "step": 345, "task_loss": 2.9252357482910156 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.1089887619018555, "epoch": 0.29, "learning_rate": 9.746478873239437e-06, "loss": 7.5206, "step": 346, "task_loss": 3.010157585144043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.355757713317871, "epoch": 0.29, "learning_rate": 9.774647887323945e-06, "loss": 7.6536, "step": 347, "task_loss": 2.7569308280944824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.296476364135742, "epoch": 0.29, "learning_rate": 9.802816901408452e-06, "loss": 7.2425, "step": 348, "task_loss": 3.1172635555267334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.624408721923828, "epoch": 0.29, "learning_rate": 9.830985915492959e-06, "loss": 7.3258, "step": 349, "task_loss": 2.573185920715332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.931515216827393, "epoch": 0.3, "learning_rate": 9.859154929577465e-06, "loss": 6.4304, "step": 350, "task_loss": 2.916036367416382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.645968914031982, "epoch": 0.3, "learning_rate": 9.887323943661972e-06, "loss": 7.5577, "step": 351, "task_loss": 2.5563416481018066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.758049011230469, "epoch": 0.3, "learning_rate": 9.915492957746479e-06, "loss": 7.7769, "step": 352, "task_loss": 3.1247825622558594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.702396392822266, "epoch": 0.3, "learning_rate": 9.943661971830986e-06, "loss": 7.0933, "step": 353, "task_loss": 2.4408838748931885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.078197479248047, "epoch": 0.3, "learning_rate": 9.971830985915494e-06, "loss": 7.9552, "step": 354, "task_loss": 2.6736903190612793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.677787780761719, "epoch": 0.3, "learning_rate": 1e-05, "loss": 7.3869, "step": 355, "task_loss": 3.348017454147339 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.968388557434082, "epoch": 0.3, "learning_rate": 1.0028169014084506e-05, "loss": 6.5914, "step": 356, "task_loss": 2.9288525581359863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.626275539398193, "epoch": 0.3, "learning_rate": 1.0056338028169014e-05, "loss": 7.0251, "step": 357, "task_loss": 2.736698865890503 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.667544841766357, "epoch": 0.3, "learning_rate": 1.0084507042253521e-05, "loss": 7.0727, "step": 358, "task_loss": 2.985847234725952 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.359759330749512, "epoch": 0.3, "learning_rate": 1.0112676056338028e-05, "loss": 8.3772, "step": 359, "task_loss": 2.6650466918945312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.514115810394287, "epoch": 0.3, "learning_rate": 1.0140845070422535e-05, "loss": 7.0417, "step": 360, "task_loss": 3.2823915481567383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.380797863006592, "epoch": 0.3, "learning_rate": 1.0169014084507043e-05, "loss": 6.7848, "step": 361, "task_loss": 2.8131725788116455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.297064781188965, "epoch": 0.31, "learning_rate": 1.019718309859155e-05, "loss": 7.2644, "step": 362, "task_loss": 2.299675703048706 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.568693161010742, "epoch": 0.31, "learning_rate": 1.0225352112676057e-05, "loss": 7.4765, "step": 363, "task_loss": 3.154043197631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.480041027069092, "epoch": 0.31, "learning_rate": 1.0253521126760564e-05, "loss": 6.6906, "step": 364, "task_loss": 2.8347115516662598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.940911293029785, "epoch": 0.31, "learning_rate": 1.0281690140845072e-05, "loss": 7.5154, "step": 365, "task_loss": 2.691312074661255 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.140597343444824, "epoch": 0.31, "learning_rate": 1.0309859154929579e-05, "loss": 7.5492, "step": 366, "task_loss": 2.648038625717163 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.027529716491699, "epoch": 0.31, "learning_rate": 1.0338028169014086e-05, "loss": 6.6344, "step": 367, "task_loss": 2.761312246322632 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.6353607177734375, "epoch": 0.31, "learning_rate": 1.0366197183098592e-05, "loss": 6.2926, "step": 368, "task_loss": 2.9303057193756104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.875520706176758, "epoch": 0.31, "learning_rate": 1.0394366197183099e-05, "loss": 6.2892, "step": 369, "task_loss": 2.361325979232788 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.972585201263428, "epoch": 0.31, "learning_rate": 1.0422535211267606e-05, "loss": 7.5614, "step": 370, "task_loss": 3.0235743522644043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.407526016235352, "epoch": 0.31, "learning_rate": 1.0450704225352113e-05, "loss": 7.2762, "step": 371, "task_loss": 2.4804024696350098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.310657501220703, "epoch": 0.31, "learning_rate": 1.047887323943662e-05, "loss": 7.3853, "step": 372, "task_loss": 1.8529287576675415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.052430629730225, "epoch": 0.32, "learning_rate": 1.0507042253521126e-05, "loss": 7.0679, "step": 373, "task_loss": 2.546995162963867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 9.159801483154297, "epoch": 0.32, "learning_rate": 1.0535211267605633e-05, "loss": 7.0588, "step": 374, "task_loss": 2.5175955295562744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.318206787109375, "epoch": 0.32, "learning_rate": 1.056338028169014e-05, "loss": 7.3538, "step": 375, "task_loss": 2.3775901794433594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.162003993988037, "epoch": 0.32, "learning_rate": 1.0591549295774648e-05, "loss": 6.4381, "step": 376, "task_loss": 1.9681452512741089 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.141857147216797, "epoch": 0.32, "learning_rate": 1.0619718309859155e-05, "loss": 7.1329, "step": 377, "task_loss": 2.357860803604126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.626873970031738, "epoch": 0.32, "learning_rate": 1.0647887323943662e-05, "loss": 6.979, "step": 378, "task_loss": 2.879138469696045 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.316118240356445, "epoch": 0.32, "learning_rate": 1.067605633802817e-05, "loss": 7.3342, "step": 379, "task_loss": 2.0899975299835205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.689858436584473, "epoch": 0.32, "learning_rate": 1.0704225352112677e-05, "loss": 6.5487, "step": 380, "task_loss": 2.7448863983154297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.519338130950928, "epoch": 0.32, "learning_rate": 1.0732394366197184e-05, "loss": 6.9483, "step": 381, "task_loss": 2.334686756134033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.875733852386475, "epoch": 0.32, "learning_rate": 1.0760563380281691e-05, "loss": 6.4557, "step": 382, "task_loss": 2.491464614868164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.825099945068359, "epoch": 0.32, "learning_rate": 1.0788732394366199e-05, "loss": 6.8655, "step": 383, "task_loss": 3.117288589477539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.160243034362793, "epoch": 0.32, "learning_rate": 1.0816901408450706e-05, "loss": 6.6322, "step": 384, "task_loss": 1.9919565916061401 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.787468910217285, "epoch": 0.33, "learning_rate": 1.0845070422535212e-05, "loss": 6.7109, "step": 385, "task_loss": 2.7341253757476807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.57350492477417, "epoch": 0.33, "learning_rate": 1.0873239436619719e-05, "loss": 6.5287, "step": 386, "task_loss": 2.167433500289917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.9347100257873535, "epoch": 0.33, "learning_rate": 1.0901408450704226e-05, "loss": 6.2481, "step": 387, "task_loss": 2.6592373847961426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.932929992675781, "epoch": 0.33, "learning_rate": 1.0929577464788733e-05, "loss": 5.7928, "step": 388, "task_loss": 2.261199951171875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.426111221313477, "epoch": 0.33, "learning_rate": 1.095774647887324e-05, "loss": 6.5075, "step": 389, "task_loss": 2.9510440826416016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.829806327819824, "epoch": 0.33, "learning_rate": 1.0985915492957748e-05, "loss": 6.3899, "step": 390, "task_loss": 2.6163406372070312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.5053582191467285, "epoch": 0.33, "learning_rate": 1.1014084507042253e-05, "loss": 6.8194, "step": 391, "task_loss": 2.7447805404663086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.601184844970703, "epoch": 0.33, "learning_rate": 1.104225352112676e-05, "loss": 6.6808, "step": 392, "task_loss": 2.228890895843506 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.306127548217773, "epoch": 0.33, "learning_rate": 1.1070422535211268e-05, "loss": 5.984, "step": 393, "task_loss": 2.4058828353881836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.898892402648926, "epoch": 0.33, "learning_rate": 1.1098591549295775e-05, "loss": 5.8397, "step": 394, "task_loss": 2.6670429706573486 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.8198561668396, "epoch": 0.33, "learning_rate": 1.1126760563380282e-05, "loss": 6.9361, "step": 395, "task_loss": 2.5320956707000732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.62319278717041, "epoch": 0.33, "learning_rate": 1.115492957746479e-05, "loss": 6.3765, "step": 396, "task_loss": 2.3790512084960938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.039251804351807, "epoch": 0.34, "learning_rate": 1.1183098591549295e-05, "loss": 6.6642, "step": 397, "task_loss": 2.634848117828369 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.362253189086914, "epoch": 0.34, "learning_rate": 1.1211267605633802e-05, "loss": 6.0084, "step": 398, "task_loss": 2.47275447845459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.417665958404541, "epoch": 0.34, "learning_rate": 1.123943661971831e-05, "loss": 6.6793, "step": 399, "task_loss": 1.8105794191360474 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.832834720611572, "epoch": 0.34, "learning_rate": 1.1267605633802817e-05, "loss": 6.757, "step": 400, "task_loss": 2.3159563541412354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.8216962814331055, "epoch": 0.34, "learning_rate": 1.1295774647887324e-05, "loss": 5.7547, "step": 401, "task_loss": 2.656388282775879 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 8.020660400390625, "epoch": 0.34, "learning_rate": 1.1323943661971831e-05, "loss": 6.4662, "step": 402, "task_loss": 3.0721511840820312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.687561988830566, "epoch": 0.34, "learning_rate": 1.1352112676056339e-05, "loss": 5.9544, "step": 403, "task_loss": 1.8456882238388062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.287472724914551, "epoch": 0.34, "learning_rate": 1.1380281690140846e-05, "loss": 5.9388, "step": 404, "task_loss": 2.1965091228485107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.423454761505127, "epoch": 0.34, "learning_rate": 1.1408450704225353e-05, "loss": 6.4311, "step": 405, "task_loss": 2.3210530281066895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.702147483825684, "epoch": 0.34, "learning_rate": 1.143661971830986e-05, "loss": 5.9905, "step": 406, "task_loss": 1.5442860126495361 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.609010696411133, "epoch": 0.34, "learning_rate": 1.1464788732394368e-05, "loss": 5.9327, "step": 407, "task_loss": 1.617924690246582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.825512886047363, "epoch": 0.34, "learning_rate": 1.1492957746478873e-05, "loss": 6.0024, "step": 408, "task_loss": 2.636655807495117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.43419075012207, "epoch": 0.35, "learning_rate": 1.152112676056338e-05, "loss": 6.2661, "step": 409, "task_loss": 1.966934084892273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.335294723510742, "epoch": 0.35, "learning_rate": 1.1549295774647888e-05, "loss": 6.4056, "step": 410, "task_loss": 1.8101224899291992 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.057373046875, "epoch": 0.35, "learning_rate": 1.1577464788732395e-05, "loss": 6.5344, "step": 411, "task_loss": 2.5225822925567627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.744077682495117, "epoch": 0.35, "learning_rate": 1.1605633802816902e-05, "loss": 6.4567, "step": 412, "task_loss": 1.5260555744171143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.335652828216553, "epoch": 0.35, "learning_rate": 1.163380281690141e-05, "loss": 6.0228, "step": 413, "task_loss": 2.2052457332611084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.340767860412598, "epoch": 0.35, "learning_rate": 1.1661971830985915e-05, "loss": 5.6168, "step": 414, "task_loss": 2.4265236854553223 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.463897705078125, "epoch": 0.35, "learning_rate": 1.1690140845070422e-05, "loss": 5.8554, "step": 415, "task_loss": 1.9620351791381836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.8635125160217285, "epoch": 0.35, "learning_rate": 1.171830985915493e-05, "loss": 5.7437, "step": 416, "task_loss": 1.9623647928237915 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.634193420410156, "epoch": 0.35, "learning_rate": 1.1746478873239437e-05, "loss": 6.4659, "step": 417, "task_loss": 2.11674427986145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.007228851318359, "epoch": 0.35, "learning_rate": 1.1774647887323944e-05, "loss": 5.7283, "step": 418, "task_loss": 2.2864298820495605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.314881801605225, "epoch": 0.35, "learning_rate": 1.1802816901408451e-05, "loss": 5.6146, "step": 419, "task_loss": 2.134979009628296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.317595958709717, "epoch": 0.35, "learning_rate": 1.1830985915492958e-05, "loss": 5.4815, "step": 420, "task_loss": 2.7087595462799072 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.101804733276367, "epoch": 0.36, "learning_rate": 1.1859154929577466e-05, "loss": 5.2587, "step": 421, "task_loss": 1.9626675844192505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.282543659210205, "epoch": 0.36, "learning_rate": 1.1887323943661973e-05, "loss": 5.8673, "step": 422, "task_loss": 1.7465267181396484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.894416332244873, "epoch": 0.36, "learning_rate": 1.191549295774648e-05, "loss": 5.2359, "step": 423, "task_loss": 2.1099681854248047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.275361061096191, "epoch": 0.36, "learning_rate": 1.1943661971830987e-05, "loss": 6.1923, "step": 424, "task_loss": 2.41560435295105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 7.7496747970581055, "epoch": 0.36, "learning_rate": 1.1971830985915493e-05, "loss": 5.7994, "step": 425, "task_loss": 1.7629340887069702 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.424802780151367, "epoch": 0.36, "learning_rate": 1.2e-05, "loss": 6.0565, "step": 426, "task_loss": 3.1678268909454346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.775101184844971, "epoch": 0.36, "learning_rate": 1.2028169014084508e-05, "loss": 5.437, "step": 427, "task_loss": 2.5030019283294678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.5718865394592285, "epoch": 0.36, "learning_rate": 1.2056338028169015e-05, "loss": 5.3078, "step": 428, "task_loss": 2.5066275596618652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.6953606605529785, "epoch": 0.36, "learning_rate": 1.2084507042253522e-05, "loss": 5.607, "step": 429, "task_loss": 1.7345818281173706 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.635362148284912, "epoch": 0.36, "learning_rate": 1.211267605633803e-05, "loss": 5.476, "step": 430, "task_loss": 2.7133171558380127 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.189793586730957, "epoch": 0.36, "learning_rate": 1.2140845070422535e-05, "loss": 5.2298, "step": 431, "task_loss": 2.466742515563965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.4330878257751465, "epoch": 0.36, "learning_rate": 1.2169014084507042e-05, "loss": 6.2072, "step": 432, "task_loss": 2.3112761974334717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.968086242675781, "epoch": 0.37, "learning_rate": 1.219718309859155e-05, "loss": 5.5535, "step": 433, "task_loss": 2.097911834716797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.492676734924316, "epoch": 0.37, "learning_rate": 1.2225352112676057e-05, "loss": 5.5296, "step": 434, "task_loss": 2.252678394317627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.1550397872924805, "epoch": 0.37, "learning_rate": 1.2253521126760564e-05, "loss": 5.559, "step": 435, "task_loss": 1.138802170753479 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.325154781341553, "epoch": 0.37, "learning_rate": 1.2281690140845071e-05, "loss": 5.007, "step": 436, "task_loss": 2.09116530418396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.556341171264648, "epoch": 0.37, "learning_rate": 1.2309859154929577e-05, "loss": 5.3766, "step": 437, "task_loss": 2.468707323074341 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.4613447189331055, "epoch": 0.37, "learning_rate": 1.2338028169014084e-05, "loss": 5.1892, "step": 438, "task_loss": 1.9713608026504517 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.277791976928711, "epoch": 0.37, "learning_rate": 1.2366197183098593e-05, "loss": 4.9623, "step": 439, "task_loss": 2.349102735519409 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.458353042602539, "epoch": 0.37, "learning_rate": 1.23943661971831e-05, "loss": 5.1038, "step": 440, "task_loss": 1.8657505512237549 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.963713645935059, "epoch": 0.37, "learning_rate": 1.2422535211267607e-05, "loss": 4.8747, "step": 441, "task_loss": 1.6702121496200562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.383641242980957, "epoch": 0.37, "learning_rate": 1.2450704225352115e-05, "loss": 5.2964, "step": 442, "task_loss": 1.7730185985565186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.641238689422607, "epoch": 0.37, "learning_rate": 1.247887323943662e-05, "loss": 5.4705, "step": 443, "task_loss": 2.4815330505371094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.389098167419434, "epoch": 0.38, "learning_rate": 1.2507042253521129e-05, "loss": 5.1952, "step": 444, "task_loss": 2.0686488151550293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.659499168395996, "epoch": 0.38, "learning_rate": 1.2535211267605635e-05, "loss": 5.1908, "step": 445, "task_loss": 1.7178361415863037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.29627799987793, "epoch": 0.38, "learning_rate": 1.256338028169014e-05, "loss": 4.7346, "step": 446, "task_loss": 1.6672016382217407 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.6648406982421875, "epoch": 0.38, "learning_rate": 1.259154929577465e-05, "loss": 5.2578, "step": 447, "task_loss": 1.8567981719970703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.176135540008545, "epoch": 0.38, "learning_rate": 1.2619718309859155e-05, "loss": 4.7599, "step": 448, "task_loss": 1.5530396699905396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.645479202270508, "epoch": 0.38, "learning_rate": 1.2647887323943664e-05, "loss": 4.2562, "step": 449, "task_loss": 2.2281363010406494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.526419639587402, "epoch": 0.38, "learning_rate": 1.267605633802817e-05, "loss": 5.2562, "step": 450, "task_loss": 1.6496117115020752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.975046157836914, "epoch": 0.38, "learning_rate": 1.2704225352112675e-05, "loss": 5.1088, "step": 451, "task_loss": 1.6631581783294678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.691431999206543, "epoch": 0.38, "learning_rate": 1.2732394366197184e-05, "loss": 5.079, "step": 452, "task_loss": 1.981769323348999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.153226852416992, "epoch": 0.38, "learning_rate": 1.276056338028169e-05, "loss": 5.4982, "step": 453, "task_loss": 1.962629795074463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.165278911590576, "epoch": 0.38, "learning_rate": 1.2788732394366198e-05, "loss": 5.4447, "step": 454, "task_loss": 2.170316219329834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.76801061630249, "epoch": 0.38, "learning_rate": 1.2816901408450704e-05, "loss": 5.0138, "step": 455, "task_loss": 2.1612648963928223 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.014156341552734, "epoch": 0.39, "learning_rate": 1.2845070422535213e-05, "loss": 4.9387, "step": 456, "task_loss": 1.503078818321228 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.167777061462402, "epoch": 0.39, "learning_rate": 1.2873239436619718e-05, "loss": 5.6637, "step": 457, "task_loss": 2.17938232421875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.8170061111450195, "epoch": 0.39, "learning_rate": 1.2901408450704226e-05, "loss": 4.6709, "step": 458, "task_loss": 2.34438419342041 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.80134916305542, "epoch": 0.39, "learning_rate": 1.2929577464788733e-05, "loss": 4.1809, "step": 459, "task_loss": 1.4715136289596558 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.870413303375244, "epoch": 0.39, "learning_rate": 1.295774647887324e-05, "loss": 4.46, "step": 460, "task_loss": 2.2589077949523926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.976487636566162, "epoch": 0.39, "learning_rate": 1.2985915492957749e-05, "loss": 4.8763, "step": 461, "task_loss": 1.99483323097229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.734901428222656, "epoch": 0.39, "learning_rate": 1.3014084507042255e-05, "loss": 4.557, "step": 462, "task_loss": 1.6392884254455566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.892328262329102, "epoch": 0.39, "learning_rate": 1.304225352112676e-05, "loss": 4.7721, "step": 463, "task_loss": 1.7658814191818237 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.592401504516602, "epoch": 0.39, "learning_rate": 1.3070422535211269e-05, "loss": 4.7685, "step": 464, "task_loss": 1.7762761116027832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.010746955871582, "epoch": 0.39, "learning_rate": 1.3098591549295775e-05, "loss": 4.5594, "step": 465, "task_loss": 1.809626579284668 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.708662033081055, "epoch": 0.39, "learning_rate": 1.3126760563380284e-05, "loss": 4.883, "step": 466, "task_loss": 2.0015037059783936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.425746917724609, "epoch": 0.39, "learning_rate": 1.3154929577464789e-05, "loss": 4.921, "step": 467, "task_loss": 1.4553717374801636 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.030062675476074, "epoch": 0.4, "learning_rate": 1.3183098591549298e-05, "loss": 4.8975, "step": 468, "task_loss": 1.9168205261230469 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.054338455200195, "epoch": 0.4, "learning_rate": 1.3211267605633804e-05, "loss": 4.7026, "step": 469, "task_loss": 1.5936261415481567 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.021197319030762, "epoch": 0.4, "learning_rate": 1.3239436619718309e-05, "loss": 4.5822, "step": 470, "task_loss": 2.4459683895111084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.034005641937256, "epoch": 0.4, "learning_rate": 1.3267605633802818e-05, "loss": 4.861, "step": 471, "task_loss": 1.8167698383331299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.40986967086792, "epoch": 0.4, "learning_rate": 1.3295774647887324e-05, "loss": 4.624, "step": 472, "task_loss": 1.7690730094909668 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.652115821838379, "epoch": 0.4, "learning_rate": 1.3323943661971833e-05, "loss": 4.9469, "step": 473, "task_loss": 1.5728529691696167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.479743480682373, "epoch": 0.4, "learning_rate": 1.3352112676056338e-05, "loss": 4.2529, "step": 474, "task_loss": 1.422099232673645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.027270317077637, "epoch": 0.4, "learning_rate": 1.3380281690140845e-05, "loss": 4.9862, "step": 475, "task_loss": 1.2742600440979004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.533626556396484, "epoch": 0.4, "learning_rate": 1.3408450704225353e-05, "loss": 4.8484, "step": 476, "task_loss": 2.153365135192871 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.344356536865234, "epoch": 0.4, "learning_rate": 1.343661971830986e-05, "loss": 4.7074, "step": 477, "task_loss": 1.7212892770767212 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.367631435394287, "epoch": 0.4, "learning_rate": 1.3464788732394367e-05, "loss": 4.5214, "step": 478, "task_loss": 2.682474136352539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.6466073989868164, "epoch": 0.4, "learning_rate": 1.3492957746478874e-05, "loss": 3.966, "step": 479, "task_loss": 2.036750078201294 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 6.014623641967773, "epoch": 0.41, "learning_rate": 1.352112676056338e-05, "loss": 4.7547, "step": 480, "task_loss": 2.1078786849975586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.491680145263672, "epoch": 0.41, "learning_rate": 1.3549295774647889e-05, "loss": 4.8602, "step": 481, "task_loss": 2.774101495742798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.545058727264404, "epoch": 0.41, "learning_rate": 1.3577464788732394e-05, "loss": 4.4193, "step": 482, "task_loss": 1.5826956033706665 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.66225528717041, "epoch": 0.41, "learning_rate": 1.3605633802816903e-05, "loss": 4.6291, "step": 483, "task_loss": 1.8062714338302612 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.48625373840332, "epoch": 0.41, "learning_rate": 1.3633802816901409e-05, "loss": 4.543, "step": 484, "task_loss": 1.469786286354065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.348448753356934, "epoch": 0.41, "learning_rate": 1.3661971830985918e-05, "loss": 4.6631, "step": 485, "task_loss": 2.100823163986206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.386277675628662, "epoch": 0.41, "learning_rate": 1.3690140845070423e-05, "loss": 3.9344, "step": 486, "task_loss": 1.5757336616516113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.583735466003418, "epoch": 0.41, "learning_rate": 1.3718309859154929e-05, "loss": 4.505, "step": 487, "task_loss": 1.696644902229309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.987791061401367, "epoch": 0.41, "learning_rate": 1.3746478873239438e-05, "loss": 4.5897, "step": 488, "task_loss": 1.454193115234375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.7178471088409424, "epoch": 0.41, "learning_rate": 1.3774647887323943e-05, "loss": 4.5046, "step": 489, "task_loss": 1.6560149192810059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.322027206420898, "epoch": 0.41, "learning_rate": 1.3802816901408452e-05, "loss": 4.6219, "step": 490, "task_loss": 1.3628454208374023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.011401176452637, "epoch": 0.41, "learning_rate": 1.3830985915492958e-05, "loss": 4.4868, "step": 491, "task_loss": 1.4787832498550415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.560376167297363, "epoch": 0.42, "learning_rate": 1.3859154929577464e-05, "loss": 4.3755, "step": 492, "task_loss": 1.821246862411499 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.015873432159424, "epoch": 0.42, "learning_rate": 1.3887323943661972e-05, "loss": 4.059, "step": 493, "task_loss": 2.150442600250244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.43079948425293, "epoch": 0.42, "learning_rate": 1.3915492957746478e-05, "loss": 4.4465, "step": 494, "task_loss": 1.63032066822052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.230546951293945, "epoch": 0.42, "learning_rate": 1.3943661971830987e-05, "loss": 3.9347, "step": 495, "task_loss": 2.1113345623016357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.571636199951172, "epoch": 0.42, "learning_rate": 1.3971830985915493e-05, "loss": 4.1927, "step": 496, "task_loss": 2.2503652572631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.891207695007324, "epoch": 0.42, "learning_rate": 1.4000000000000001e-05, "loss": 4.5598, "step": 497, "task_loss": 1.677118182182312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.944802284240723, "epoch": 0.42, "learning_rate": 1.4028169014084509e-05, "loss": 4.5944, "step": 498, "task_loss": 1.3673174381256104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.3295037746429443, "epoch": 0.42, "learning_rate": 1.4056338028169014e-05, "loss": 3.6315, "step": 499, "task_loss": 1.2290605306625366 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.096893787384033, "epoch": 0.42, "learning_rate": 1.4084507042253523e-05, "loss": 3.8787, "step": 500, "task_loss": 1.007515549659729 }, { "epoch": 0.42, "eval_accuracy": 0.7163168316831683, "eval_loss": 3.9970788955688477, "eval_runtime": 209.9035, "eval_samples_per_second": 120.293, "eval_steps_per_second": 0.943, "step": 500 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.6166276931762695, "epoch": 0.42, "learning_rate": 1.4112676056338029e-05, "loss": 4.179, "step": 501, "task_loss": 2.1177053451538086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.112758159637451, "epoch": 0.42, "learning_rate": 1.4140845070422538e-05, "loss": 4.122, "step": 502, "task_loss": 1.1636425256729126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.054601192474365, "epoch": 0.42, "learning_rate": 1.4169014084507043e-05, "loss": 4.0992, "step": 503, "task_loss": 1.3891127109527588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.384589195251465, "epoch": 0.43, "learning_rate": 1.4197183098591549e-05, "loss": 3.9424, "step": 504, "task_loss": 1.8737651109695435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.473025321960449, "epoch": 0.43, "learning_rate": 1.4225352112676058e-05, "loss": 3.8042, "step": 505, "task_loss": 1.904947280883789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.354865550994873, "epoch": 0.43, "learning_rate": 1.4253521126760563e-05, "loss": 3.3452, "step": 506, "task_loss": 1.4950652122497559 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.3768842220306396, "epoch": 0.43, "learning_rate": 1.4281690140845072e-05, "loss": 3.8154, "step": 507, "task_loss": 0.8713154792785645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.34188175201416, "epoch": 0.43, "learning_rate": 1.4309859154929578e-05, "loss": 3.9933, "step": 508, "task_loss": 0.8645720481872559 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.059744834899902, "epoch": 0.43, "learning_rate": 1.4338028169014083e-05, "loss": 4.1655, "step": 509, "task_loss": 1.5869700908660889 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.584723949432373, "epoch": 0.43, "learning_rate": 1.4366197183098592e-05, "loss": 3.9234, "step": 510, "task_loss": 1.9012998342514038 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.2749104499816895, "epoch": 0.43, "learning_rate": 1.4394366197183098e-05, "loss": 3.7552, "step": 511, "task_loss": 1.674513816833496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.051058769226074, "epoch": 0.43, "learning_rate": 1.4422535211267607e-05, "loss": 3.6376, "step": 512, "task_loss": 1.4297312498092651 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.7422566413879395, "epoch": 0.43, "learning_rate": 1.4450704225352112e-05, "loss": 4.6421, "step": 513, "task_loss": 1.3113099336624146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.4815096855163574, "epoch": 0.43, "learning_rate": 1.4478873239436621e-05, "loss": 3.451, "step": 514, "task_loss": 1.6122208833694458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.7584569454193115, "epoch": 0.44, "learning_rate": 1.4507042253521127e-05, "loss": 3.8569, "step": 515, "task_loss": 1.7796612977981567 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.8626341819763184, "epoch": 0.44, "learning_rate": 1.4535211267605634e-05, "loss": 3.2844, "step": 516, "task_loss": 1.6421499252319336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.093626499176025, "epoch": 0.44, "learning_rate": 1.4563380281690141e-05, "loss": 3.6682, "step": 517, "task_loss": 2.0482263565063477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.046904563903809, "epoch": 0.44, "learning_rate": 1.4591549295774649e-05, "loss": 3.4497, "step": 518, "task_loss": 1.434609293937683 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.737701416015625, "epoch": 0.44, "learning_rate": 1.4619718309859156e-05, "loss": 3.4914, "step": 519, "task_loss": 1.9826374053955078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.628640174865723, "epoch": 0.44, "learning_rate": 1.4647887323943663e-05, "loss": 4.0874, "step": 520, "task_loss": 1.3740766048431396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.27382755279541, "epoch": 0.44, "learning_rate": 1.4676056338028169e-05, "loss": 3.758, "step": 521, "task_loss": 1.5452888011932373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.164515495300293, "epoch": 0.44, "learning_rate": 1.4704225352112678e-05, "loss": 3.7604, "step": 522, "task_loss": 1.3976974487304688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.492462635040283, "epoch": 0.44, "learning_rate": 1.4732394366197183e-05, "loss": 3.3159, "step": 523, "task_loss": 1.4963033199310303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.437184810638428, "epoch": 0.44, "learning_rate": 1.4760563380281692e-05, "loss": 3.3655, "step": 524, "task_loss": 1.2165933847427368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.5983920097351074, "epoch": 0.44, "learning_rate": 1.4788732394366198e-05, "loss": 3.506, "step": 525, "task_loss": 1.4860986471176147 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.6944451332092285, "epoch": 0.44, "learning_rate": 1.4816901408450707e-05, "loss": 3.731, "step": 526, "task_loss": 1.3080295324325562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.71939754486084, "epoch": 0.45, "learning_rate": 1.4845070422535212e-05, "loss": 3.3072, "step": 527, "task_loss": 1.7853186130523682 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.128744125366211, "epoch": 0.45, "learning_rate": 1.4873239436619718e-05, "loss": 3.5789, "step": 528, "task_loss": 1.944583773612976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.9824602603912354, "epoch": 0.45, "learning_rate": 1.4901408450704227e-05, "loss": 3.5889, "step": 529, "task_loss": 1.6610890626907349 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.114217758178711, "epoch": 0.45, "learning_rate": 1.4929577464788732e-05, "loss": 3.7145, "step": 530, "task_loss": 1.4136425256729126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.364049911499023, "epoch": 0.45, "learning_rate": 1.4957746478873241e-05, "loss": 3.7412, "step": 531, "task_loss": 1.7477610111236572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.4147512912750244, "epoch": 0.45, "learning_rate": 1.4985915492957747e-05, "loss": 3.2816, "step": 532, "task_loss": 1.879472017288208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.9051833152770996, "epoch": 0.45, "learning_rate": 1.5014084507042252e-05, "loss": 3.7114, "step": 533, "task_loss": 1.655503511428833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.426325798034668, "epoch": 0.45, "learning_rate": 1.5042253521126761e-05, "loss": 3.4835, "step": 534, "task_loss": 1.6233099699020386 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.682565689086914, "epoch": 0.45, "learning_rate": 1.5070422535211269e-05, "loss": 3.7549, "step": 535, "task_loss": 1.7456638813018799 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.483036518096924, "epoch": 0.45, "learning_rate": 1.5098591549295776e-05, "loss": 3.2642, "step": 536, "task_loss": 1.8548330068588257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.704169273376465, "epoch": 0.45, "learning_rate": 1.5126760563380283e-05, "loss": 3.2959, "step": 537, "task_loss": 1.31657075881958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 5.36446475982666, "epoch": 0.45, "learning_rate": 1.5154929577464789e-05, "loss": 4.3327, "step": 538, "task_loss": 2.14127779006958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.7185497283935547, "epoch": 0.46, "learning_rate": 1.5183098591549298e-05, "loss": 3.2373, "step": 539, "task_loss": 1.4561219215393066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.815772294998169, "epoch": 0.46, "learning_rate": 1.5211267605633803e-05, "loss": 3.4627, "step": 540, "task_loss": 1.257128119468689 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.9190382957458496, "epoch": 0.46, "learning_rate": 1.5239436619718312e-05, "loss": 3.5863, "step": 541, "task_loss": 1.5523756742477417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.9574642181396484, "epoch": 0.46, "learning_rate": 1.5267605633802818e-05, "loss": 3.336, "step": 542, "task_loss": 1.0817890167236328 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.695988178253174, "epoch": 0.46, "learning_rate": 1.5295774647887325e-05, "loss": 3.1959, "step": 543, "task_loss": 1.3092560768127441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.3063454627990723, "epoch": 0.46, "learning_rate": 1.5323943661971832e-05, "loss": 3.4275, "step": 544, "task_loss": 1.324326753616333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.673339605331421, "epoch": 0.46, "learning_rate": 1.535211267605634e-05, "loss": 3.4777, "step": 545, "task_loss": 1.5655810832977295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.6682281494140625, "epoch": 0.46, "learning_rate": 1.5380281690140847e-05, "loss": 3.4981, "step": 546, "task_loss": 0.8035436272621155 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.274836540222168, "epoch": 0.46, "learning_rate": 1.5408450704225354e-05, "loss": 3.3545, "step": 547, "task_loss": 1.5687724351882935 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.037229537963867, "epoch": 0.46, "learning_rate": 1.543661971830986e-05, "loss": 3.5097, "step": 548, "task_loss": 1.4525964260101318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.924689769744873, "epoch": 0.46, "learning_rate": 1.546478873239437e-05, "loss": 3.1174, "step": 549, "task_loss": 1.979385495185852 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.198172569274902, "epoch": 0.46, "learning_rate": 1.5492957746478872e-05, "loss": 3.5753, "step": 550, "task_loss": 1.8374199867248535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.2524702548980713, "epoch": 0.47, "learning_rate": 1.5521126760563383e-05, "loss": 2.9681, "step": 551, "task_loss": 1.1076743602752686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.386861562728882, "epoch": 0.47, "learning_rate": 1.5549295774647887e-05, "loss": 2.972, "step": 552, "task_loss": 1.6688483953475952 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.7125134468078613, "epoch": 0.47, "learning_rate": 1.5577464788732397e-05, "loss": 3.2117, "step": 553, "task_loss": 1.5057034492492676 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.5738308429718018, "epoch": 0.47, "learning_rate": 1.56056338028169e-05, "loss": 3.135, "step": 554, "task_loss": 1.0897103548049927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.906282901763916, "epoch": 0.47, "learning_rate": 1.5633802816901412e-05, "loss": 3.2921, "step": 555, "task_loss": 1.3482651710510254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.7278647422790527, "epoch": 0.47, "learning_rate": 1.5661971830985916e-05, "loss": 3.1928, "step": 556, "task_loss": 1.2032959461212158 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.076094627380371, "epoch": 0.47, "learning_rate": 1.5690140845070423e-05, "loss": 2.8717, "step": 557, "task_loss": 1.9845432043075562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.6442253589630127, "epoch": 0.47, "learning_rate": 1.571830985915493e-05, "loss": 3.124, "step": 558, "task_loss": 1.4958018064498901 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.571322202682495, "epoch": 0.47, "learning_rate": 1.5746478873239437e-05, "loss": 3.5713, "step": 559, "task_loss": 0.9030741453170776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.1229770183563232, "epoch": 0.47, "learning_rate": 1.5774647887323945e-05, "loss": 3.1903, "step": 560, "task_loss": 1.8160704374313354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.0462615489959717, "epoch": 0.47, "learning_rate": 1.5802816901408452e-05, "loss": 2.9044, "step": 561, "task_loss": 0.9207674264907837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.789585590362549, "epoch": 0.47, "learning_rate": 1.5830985915492956e-05, "loss": 3.3481, "step": 562, "task_loss": 1.4596476554870605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.9587173461914062, "epoch": 0.48, "learning_rate": 1.5859154929577466e-05, "loss": 3.1575, "step": 563, "task_loss": 1.487863302230835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.6891732215881348, "epoch": 0.48, "learning_rate": 1.588732394366197e-05, "loss": 2.8709, "step": 564, "task_loss": 1.8249479532241821 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.65269136428833, "epoch": 0.48, "learning_rate": 1.591549295774648e-05, "loss": 3.187, "step": 565, "task_loss": 1.4860100746154785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.382808208465576, "epoch": 0.48, "learning_rate": 1.5943661971830988e-05, "loss": 3.0772, "step": 566, "task_loss": 1.0431417226791382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.926156759262085, "epoch": 0.48, "learning_rate": 1.5971830985915492e-05, "loss": 3.4196, "step": 567, "task_loss": 1.9890713691711426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.0714714527130127, "epoch": 0.48, "learning_rate": 1.6000000000000003e-05, "loss": 2.9432, "step": 568, "task_loss": 0.8034501075744629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.356598138809204, "epoch": 0.48, "learning_rate": 1.6028169014084507e-05, "loss": 3.4319, "step": 569, "task_loss": 1.7381786108016968 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.9565725326538086, "epoch": 0.48, "learning_rate": 1.6056338028169017e-05, "loss": 2.9744, "step": 570, "task_loss": 1.279245376586914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6690900325775146, "epoch": 0.48, "learning_rate": 1.608450704225352e-05, "loss": 2.9787, "step": 571, "task_loss": 0.9212945103645325 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.928852081298828, "epoch": 0.48, "learning_rate": 1.611267605633803e-05, "loss": 2.6834, "step": 572, "task_loss": 1.183809757232666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.5316152572631836, "epoch": 0.48, "learning_rate": 1.6140845070422536e-05, "loss": 3.3059, "step": 573, "task_loss": 1.6690236330032349 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.863335371017456, "epoch": 0.48, "learning_rate": 1.6169014084507043e-05, "loss": 3.1535, "step": 574, "task_loss": 1.5338009595870972 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.614313840866089, "epoch": 0.49, "learning_rate": 1.619718309859155e-05, "loss": 3.2997, "step": 575, "task_loss": 1.6282873153686523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.4861743450164795, "epoch": 0.49, "learning_rate": 1.6225352112676057e-05, "loss": 3.2612, "step": 576, "task_loss": 1.2477585077285767 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.2992608547210693, "epoch": 0.49, "learning_rate": 1.6253521126760565e-05, "loss": 2.9295, "step": 577, "task_loss": 1.4416483640670776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.021012783050537, "epoch": 0.49, "learning_rate": 1.6281690140845072e-05, "loss": 3.2099, "step": 578, "task_loss": 1.879157543182373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.3894143104553223, "epoch": 0.49, "learning_rate": 1.6309859154929576e-05, "loss": 2.751, "step": 579, "task_loss": 1.8970246315002441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.664910078048706, "epoch": 0.49, "learning_rate": 1.6338028169014086e-05, "loss": 3.3005, "step": 580, "task_loss": 1.495977520942688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.033362865447998, "epoch": 0.49, "learning_rate": 1.636619718309859e-05, "loss": 3.0385, "step": 581, "task_loss": 1.0157878398895264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.566267728805542, "epoch": 0.49, "learning_rate": 1.63943661971831e-05, "loss": 2.9337, "step": 582, "task_loss": 1.6010215282440186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8523974418640137, "epoch": 0.49, "learning_rate": 1.6422535211267605e-05, "loss": 3.0873, "step": 583, "task_loss": 0.8597267866134644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 4.0205159187316895, "epoch": 0.49, "learning_rate": 1.6450704225352112e-05, "loss": 3.4188, "step": 584, "task_loss": 1.4471458196640015 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8184781074523926, "epoch": 0.49, "learning_rate": 1.647887323943662e-05, "loss": 3.0886, "step": 585, "task_loss": 1.6967302560806274 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.3995189666748047, "epoch": 0.5, "learning_rate": 1.6507042253521126e-05, "loss": 2.7489, "step": 586, "task_loss": 0.6230946183204651 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.598870277404785, "epoch": 0.5, "learning_rate": 1.6535211267605634e-05, "loss": 2.5881, "step": 587, "task_loss": 1.3241996765136719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3678393363952637, "epoch": 0.5, "learning_rate": 1.656338028169014e-05, "loss": 3.1029, "step": 588, "task_loss": 2.0536115169525146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.995225429534912, "epoch": 0.5, "learning_rate": 1.659154929577465e-05, "loss": 3.1701, "step": 589, "task_loss": 1.3865033388137817 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6743292808532715, "epoch": 0.5, "learning_rate": 1.6619718309859155e-05, "loss": 2.7423, "step": 590, "task_loss": 1.6739667654037476 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.7314577102661133, "epoch": 0.5, "learning_rate": 1.6647887323943663e-05, "loss": 2.5873, "step": 591, "task_loss": 1.1124565601348877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.859431266784668, "epoch": 0.5, "learning_rate": 1.667605633802817e-05, "loss": 3.1013, "step": 592, "task_loss": 1.1225887537002563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.116793632507324, "epoch": 0.5, "learning_rate": 1.6704225352112677e-05, "loss": 2.9514, "step": 593, "task_loss": 1.5750210285186768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6408908367156982, "epoch": 0.5, "learning_rate": 1.6732394366197184e-05, "loss": 2.5462, "step": 594, "task_loss": 0.9937089681625366 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.0290231704711914, "epoch": 0.5, "learning_rate": 1.676056338028169e-05, "loss": 2.4723, "step": 595, "task_loss": 0.9639025926589966 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.731489419937134, "epoch": 0.5, "learning_rate": 1.6788732394366195e-05, "loss": 2.5895, "step": 596, "task_loss": 1.720457673072815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.76826810836792, "epoch": 0.5, "learning_rate": 1.6816901408450706e-05, "loss": 2.68, "step": 597, "task_loss": 0.8113469481468201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.48366641998291, "epoch": 0.51, "learning_rate": 1.684507042253521e-05, "loss": 3.0298, "step": 598, "task_loss": 2.4095141887664795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.1510400772094727, "epoch": 0.51, "learning_rate": 1.687323943661972e-05, "loss": 2.4897, "step": 599, "task_loss": 1.5691567659378052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9961298704147339, "epoch": 0.51, "learning_rate": 1.6901408450704224e-05, "loss": 2.6353, "step": 600, "task_loss": 1.1834359169006348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6826374530792236, "epoch": 0.51, "learning_rate": 1.6929577464788735e-05, "loss": 3.2572, "step": 601, "task_loss": 1.7965114116668701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.676723003387451, "epoch": 0.51, "learning_rate": 1.695774647887324e-05, "loss": 2.8414, "step": 602, "task_loss": 1.653803825378418 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8152341842651367, "epoch": 0.51, "learning_rate": 1.6985915492957746e-05, "loss": 2.422, "step": 603, "task_loss": 1.2122632265090942 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6073555946350098, "epoch": 0.51, "learning_rate": 1.7014084507042253e-05, "loss": 2.5811, "step": 604, "task_loss": 1.5323851108551025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.943655252456665, "epoch": 0.51, "learning_rate": 1.704225352112676e-05, "loss": 2.5063, "step": 605, "task_loss": 1.8971638679504395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.3647165298461914, "epoch": 0.51, "learning_rate": 1.7070422535211268e-05, "loss": 2.367, "step": 606, "task_loss": 1.0362067222595215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.826585292816162, "epoch": 0.51, "learning_rate": 1.7098591549295775e-05, "loss": 3.1291, "step": 607, "task_loss": 1.4940276145935059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8632750511169434, "epoch": 0.51, "learning_rate": 1.7126760563380282e-05, "loss": 2.776, "step": 608, "task_loss": 1.203714370727539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.323424816131592, "epoch": 0.51, "learning_rate": 1.715492957746479e-05, "loss": 2.8013, "step": 609, "task_loss": 1.3272455930709839 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.242588996887207, "epoch": 0.52, "learning_rate": 1.7183098591549297e-05, "loss": 2.6089, "step": 610, "task_loss": 1.1587406396865845 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3100194931030273, "epoch": 0.52, "learning_rate": 1.7211267605633804e-05, "loss": 2.2678, "step": 611, "task_loss": 1.2684319019317627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3272950649261475, "epoch": 0.52, "learning_rate": 1.723943661971831e-05, "loss": 2.5438, "step": 612, "task_loss": 1.0473456382751465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5483145713806152, "epoch": 0.52, "learning_rate": 1.7267605633802815e-05, "loss": 2.6998, "step": 613, "task_loss": 1.6970114707946777 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0766031742095947, "epoch": 0.52, "learning_rate": 1.7295774647887326e-05, "loss": 2.8675, "step": 614, "task_loss": 1.0120477676391602 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6893787384033203, "epoch": 0.52, "learning_rate": 1.732394366197183e-05, "loss": 2.8008, "step": 615, "task_loss": 1.934200406074524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.888680934906006, "epoch": 0.52, "learning_rate": 1.735211267605634e-05, "loss": 2.8948, "step": 616, "task_loss": 1.9365462064743042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.160521984100342, "epoch": 0.52, "learning_rate": 1.7380281690140844e-05, "loss": 2.3905, "step": 617, "task_loss": 2.0256783962249756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4986763000488281, "epoch": 0.52, "learning_rate": 1.7408450704225355e-05, "loss": 2.1184, "step": 618, "task_loss": 0.9664560556411743 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.4248886108398438, "epoch": 0.52, "learning_rate": 1.743661971830986e-05, "loss": 2.8243, "step": 619, "task_loss": 1.353227972984314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.7675399780273438, "epoch": 0.52, "learning_rate": 1.7464788732394366e-05, "loss": 2.4334, "step": 620, "task_loss": 1.3117276430130005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.7830591201782227, "epoch": 0.52, "learning_rate": 1.7492957746478873e-05, "loss": 2.7845, "step": 621, "task_loss": 1.2904573678970337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8092398643493652, "epoch": 0.53, "learning_rate": 1.752112676056338e-05, "loss": 2.3692, "step": 622, "task_loss": 1.949466586112976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.4860198497772217, "epoch": 0.53, "learning_rate": 1.7549295774647888e-05, "loss": 2.102, "step": 623, "task_loss": 1.0180561542510986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.2898504734039307, "epoch": 0.53, "learning_rate": 1.7577464788732395e-05, "loss": 2.4718, "step": 624, "task_loss": 1.2684690952301025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5214316844940186, "epoch": 0.53, "learning_rate": 1.7605633802816902e-05, "loss": 2.8561, "step": 625, "task_loss": 1.3951313495635986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.829968214035034, "epoch": 0.53, "learning_rate": 1.763380281690141e-05, "loss": 3.08, "step": 626, "task_loss": 1.5317842960357666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5899343490600586, "epoch": 0.53, "learning_rate": 1.7661971830985917e-05, "loss": 2.4464, "step": 627, "task_loss": 1.0027706623077393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6261229515075684, "epoch": 0.53, "learning_rate": 1.7690140845070424e-05, "loss": 2.4661, "step": 628, "task_loss": 1.2766175270080566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8844032287597656, "epoch": 0.53, "learning_rate": 1.771830985915493e-05, "loss": 2.2292, "step": 629, "task_loss": 1.637429118156433 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.1943817138671875, "epoch": 0.53, "learning_rate": 1.774647887323944e-05, "loss": 2.2258, "step": 630, "task_loss": 1.7106339931488037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.150585651397705, "epoch": 0.53, "learning_rate": 1.7774647887323946e-05, "loss": 2.6845, "step": 631, "task_loss": 0.8921009302139282 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5127146244049072, "epoch": 0.53, "learning_rate": 1.780281690140845e-05, "loss": 2.2368, "step": 632, "task_loss": 1.7971388101577759 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.811774969100952, "epoch": 0.53, "learning_rate": 1.783098591549296e-05, "loss": 2.4531, "step": 633, "task_loss": 1.3252023458480835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.536064624786377, "epoch": 0.54, "learning_rate": 1.7859154929577464e-05, "loss": 2.8178, "step": 634, "task_loss": 1.3023265600204468 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0318524837493896, "epoch": 0.54, "learning_rate": 1.7887323943661975e-05, "loss": 2.067, "step": 635, "task_loss": 0.6948379278182983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.328010320663452, "epoch": 0.54, "learning_rate": 1.791549295774648e-05, "loss": 2.4967, "step": 636, "task_loss": 1.8350281715393066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.099194288253784, "epoch": 0.54, "learning_rate": 1.7943661971830986e-05, "loss": 2.3365, "step": 637, "task_loss": 0.8786344528198242 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.998077630996704, "epoch": 0.54, "learning_rate": 1.7971830985915493e-05, "loss": 2.284, "step": 638, "task_loss": 1.84226393699646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6775035858154297, "epoch": 0.54, "learning_rate": 1.8e-05, "loss": 2.7364, "step": 639, "task_loss": 0.5221678614616394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.422546148300171, "epoch": 0.54, "learning_rate": 1.8028169014084508e-05, "loss": 2.756, "step": 640, "task_loss": 1.4518271684646606 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6454066038131714, "epoch": 0.54, "learning_rate": 1.8056338028169015e-05, "loss": 2.2618, "step": 641, "task_loss": 0.955149233341217 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.470529079437256, "epoch": 0.54, "learning_rate": 1.8084507042253522e-05, "loss": 2.5944, "step": 642, "task_loss": 1.8390703201293945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8257336616516113, "epoch": 0.54, "learning_rate": 1.811267605633803e-05, "loss": 2.8414, "step": 643, "task_loss": 1.3773953914642334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.182842254638672, "epoch": 0.54, "learning_rate": 1.8140845070422537e-05, "loss": 2.4882, "step": 644, "task_loss": 0.8225191235542297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.537513256072998, "epoch": 0.54, "learning_rate": 1.8169014084507044e-05, "loss": 2.4207, "step": 645, "task_loss": 0.9990071058273315 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.363530158996582, "epoch": 0.55, "learning_rate": 1.819718309859155e-05, "loss": 2.0541, "step": 646, "task_loss": 0.8603134155273438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.501936674118042, "epoch": 0.55, "learning_rate": 1.822535211267606e-05, "loss": 1.9278, "step": 647, "task_loss": 0.9532979130744934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.9918746948242188, "epoch": 0.55, "learning_rate": 1.8253521126760566e-05, "loss": 2.5251, "step": 648, "task_loss": 0.838292121887207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5707316398620605, "epoch": 0.55, "learning_rate": 1.828169014084507e-05, "loss": 2.0127, "step": 649, "task_loss": 0.9503273367881775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.8221964836120605, "epoch": 0.55, "learning_rate": 1.830985915492958e-05, "loss": 2.136, "step": 650, "task_loss": 1.5597987174987793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6967906951904297, "epoch": 0.55, "learning_rate": 1.8338028169014084e-05, "loss": 2.611, "step": 651, "task_loss": 1.574896216392517 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.370274543762207, "epoch": 0.55, "learning_rate": 1.8366197183098595e-05, "loss": 1.8281, "step": 652, "task_loss": 1.0679010152816772 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.859806537628174, "epoch": 0.55, "learning_rate": 1.83943661971831e-05, "loss": 2.1059, "step": 653, "task_loss": 0.8147316575050354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3680334091186523, "epoch": 0.55, "learning_rate": 1.8422535211267606e-05, "loss": 1.9351, "step": 654, "task_loss": 1.0542118549346924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.725268602371216, "epoch": 0.55, "learning_rate": 1.8450704225352113e-05, "loss": 2.2329, "step": 655, "task_loss": 0.8940112590789795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.849514961242676, "epoch": 0.55, "learning_rate": 1.847887323943662e-05, "loss": 2.2484, "step": 656, "task_loss": 1.4287128448486328 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0484797954559326, "epoch": 0.56, "learning_rate": 1.8507042253521128e-05, "loss": 2.0038, "step": 657, "task_loss": 0.8824728727340698 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5875442028045654, "epoch": 0.56, "learning_rate": 1.8535211267605635e-05, "loss": 2.5317, "step": 658, "task_loss": 0.922415018081665 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0197553634643555, "epoch": 0.56, "learning_rate": 1.8563380281690142e-05, "loss": 2.1768, "step": 659, "task_loss": 1.2689388990402222 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.419966220855713, "epoch": 0.56, "learning_rate": 1.859154929577465e-05, "loss": 2.612, "step": 660, "task_loss": 0.615218997001648 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.7805733680725098, "epoch": 0.56, "learning_rate": 1.8619718309859157e-05, "loss": 2.4964, "step": 661, "task_loss": 1.0263174772262573 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6156039237976074, "epoch": 0.56, "learning_rate": 1.8647887323943664e-05, "loss": 2.2224, "step": 662, "task_loss": 2.1494550704956055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0571534633636475, "epoch": 0.56, "learning_rate": 1.867605633802817e-05, "loss": 2.1002, "step": 663, "task_loss": 0.680842399597168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.403670072555542, "epoch": 0.56, "learning_rate": 1.870422535211268e-05, "loss": 2.4136, "step": 664, "task_loss": 1.095179557800293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.924905776977539, "epoch": 0.56, "learning_rate": 1.8732394366197186e-05, "loss": 2.158, "step": 665, "task_loss": 1.8632867336273193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5188846588134766, "epoch": 0.56, "learning_rate": 1.876056338028169e-05, "loss": 2.0325, "step": 666, "task_loss": 0.9695818424224854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.2819626331329346, "epoch": 0.56, "learning_rate": 1.87887323943662e-05, "loss": 2.0652, "step": 667, "task_loss": 1.4170125722885132 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5591800212860107, "epoch": 0.56, "learning_rate": 1.8816901408450704e-05, "loss": 2.2081, "step": 668, "task_loss": 2.0005176067352295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.368297576904297, "epoch": 0.57, "learning_rate": 1.8845070422535215e-05, "loss": 2.1216, "step": 669, "task_loss": 2.006211042404175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5429563522338867, "epoch": 0.57, "learning_rate": 1.887323943661972e-05, "loss": 2.426, "step": 670, "task_loss": 1.5953952074050903 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6170248985290527, "epoch": 0.57, "learning_rate": 1.8901408450704226e-05, "loss": 2.6653, "step": 671, "task_loss": 1.6802127361297607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9930808544158936, "epoch": 0.57, "learning_rate": 1.8929577464788733e-05, "loss": 1.9811, "step": 672, "task_loss": 0.7351468801498413 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.2782530784606934, "epoch": 0.57, "learning_rate": 1.895774647887324e-05, "loss": 2.0612, "step": 673, "task_loss": 1.965089201927185 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7626447677612305, "epoch": 0.57, "learning_rate": 1.8985915492957747e-05, "loss": 2.0099, "step": 674, "task_loss": 0.8866486549377441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.910172939300537, "epoch": 0.57, "learning_rate": 1.9014084507042255e-05, "loss": 2.41, "step": 675, "task_loss": 1.8536746501922607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.503964424133301, "epoch": 0.57, "learning_rate": 1.9042253521126762e-05, "loss": 2.3273, "step": 676, "task_loss": 1.0988515615463257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.383261203765869, "epoch": 0.57, "learning_rate": 1.907042253521127e-05, "loss": 2.1475, "step": 677, "task_loss": 0.9046470522880554 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9749959707260132, "epoch": 0.57, "learning_rate": 1.9098591549295773e-05, "loss": 1.9294, "step": 678, "task_loss": 0.6373335123062134 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.980058193206787, "epoch": 0.57, "learning_rate": 1.9126760563380284e-05, "loss": 1.9847, "step": 679, "task_loss": 1.4337003231048584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.343250274658203, "epoch": 0.57, "learning_rate": 1.9154929577464788e-05, "loss": 2.0189, "step": 680, "task_loss": 1.4706759452819824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.831944227218628, "epoch": 0.58, "learning_rate": 1.9183098591549298e-05, "loss": 2.2685, "step": 681, "task_loss": 1.04743492603302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.026303291320801, "epoch": 0.58, "learning_rate": 1.9211267605633802e-05, "loss": 2.3114, "step": 682, "task_loss": 0.8355689644813538 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.091902732849121, "epoch": 0.58, "learning_rate": 1.923943661971831e-05, "loss": 2.1291, "step": 683, "task_loss": 1.4049513339996338 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8017466068267822, "epoch": 0.58, "learning_rate": 1.926760563380282e-05, "loss": 1.8943, "step": 684, "task_loss": 1.2781094312667847 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5578627586364746, "epoch": 0.58, "learning_rate": 1.9295774647887324e-05, "loss": 1.9125, "step": 685, "task_loss": 1.3223499059677124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 3.2198116779327393, "epoch": 0.58, "learning_rate": 1.9323943661971834e-05, "loss": 2.3383, "step": 686, "task_loss": 0.9937374591827393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.399663209915161, "epoch": 0.58, "learning_rate": 1.9352112676056338e-05, "loss": 1.9243, "step": 687, "task_loss": 1.9996898174285889 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9192755222320557, "epoch": 0.58, "learning_rate": 1.938028169014085e-05, "loss": 1.8341, "step": 688, "task_loss": 1.3862757682800293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9410579204559326, "epoch": 0.58, "learning_rate": 1.9408450704225353e-05, "loss": 1.8819, "step": 689, "task_loss": 0.7355526685714722 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8633229732513428, "epoch": 0.58, "learning_rate": 1.943661971830986e-05, "loss": 2.0529, "step": 690, "task_loss": 0.5528593063354492 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5752959251403809, "epoch": 0.58, "learning_rate": 1.9464788732394367e-05, "loss": 1.8852, "step": 691, "task_loss": 0.5402342677116394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7484369277954102, "epoch": 0.58, "learning_rate": 1.9492957746478875e-05, "loss": 2.0509, "step": 692, "task_loss": 1.5105679035186768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7993571758270264, "epoch": 0.59, "learning_rate": 1.9521126760563382e-05, "loss": 1.9581, "step": 693, "task_loss": 1.7804815769195557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.999260425567627, "epoch": 0.59, "learning_rate": 1.954929577464789e-05, "loss": 2.0536, "step": 694, "task_loss": 1.663329839706421 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.990776777267456, "epoch": 0.59, "learning_rate": 1.9577464788732393e-05, "loss": 1.9634, "step": 695, "task_loss": 1.4139103889465332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9558966159820557, "epoch": 0.59, "learning_rate": 1.9605633802816904e-05, "loss": 2.0181, "step": 696, "task_loss": 0.8625527620315552 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8952395915985107, "epoch": 0.59, "learning_rate": 1.9633802816901407e-05, "loss": 1.6983, "step": 697, "task_loss": 1.0652843713760376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.475353717803955, "epoch": 0.59, "learning_rate": 1.9661971830985918e-05, "loss": 1.8722, "step": 698, "task_loss": 0.6927515864372253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8836183547973633, "epoch": 0.59, "learning_rate": 1.9690140845070422e-05, "loss": 1.9832, "step": 699, "task_loss": 0.802126407623291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.2115800380706787, "epoch": 0.59, "learning_rate": 1.971830985915493e-05, "loss": 2.042, "step": 700, "task_loss": 1.1001917123794556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9545130729675293, "epoch": 0.59, "learning_rate": 1.9746478873239436e-05, "loss": 2.067, "step": 701, "task_loss": 0.977561354637146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3654390573501587, "epoch": 0.59, "learning_rate": 1.9774647887323944e-05, "loss": 1.7511, "step": 702, "task_loss": 0.9540075063705444 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9628114700317383, "epoch": 0.59, "learning_rate": 1.980281690140845e-05, "loss": 2.1381, "step": 703, "task_loss": 1.2848026752471924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.4854636192321777, "epoch": 0.59, "learning_rate": 1.9830985915492958e-05, "loss": 2.2574, "step": 704, "task_loss": 1.850175380706787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.014627456665039, "epoch": 0.6, "learning_rate": 1.9859154929577465e-05, "loss": 2.0802, "step": 705, "task_loss": 2.0518641471862793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7706359624862671, "epoch": 0.6, "learning_rate": 1.9887323943661973e-05, "loss": 1.5677, "step": 706, "task_loss": 0.29504331946372986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.775712490081787, "epoch": 0.6, "learning_rate": 1.991549295774648e-05, "loss": 1.9932, "step": 707, "task_loss": 1.2281757593154907 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9662954807281494, "epoch": 0.6, "learning_rate": 1.9943661971830987e-05, "loss": 1.9671, "step": 708, "task_loss": 1.2022144794464111 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.312777042388916, "epoch": 0.6, "learning_rate": 1.9971830985915494e-05, "loss": 2.4283, "step": 709, "task_loss": 0.8498018980026245 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2055606842041016, "epoch": 0.6, "learning_rate": 2e-05, "loss": 2.0791, "step": 710, "task_loss": 1.1448895931243896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.685192584991455, "epoch": 0.6, "learning_rate": 2.002816901408451e-05, "loss": 2.0959, "step": 711, "task_loss": 1.0867475271224976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.2475881576538086, "epoch": 0.6, "learning_rate": 2.0056338028169013e-05, "loss": 2.2815, "step": 712, "task_loss": 1.3839083909988403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7731499671936035, "epoch": 0.6, "learning_rate": 2.0084507042253523e-05, "loss": 1.5799, "step": 713, "task_loss": 0.9018091559410095 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.297883987426758, "epoch": 0.6, "learning_rate": 2.0112676056338027e-05, "loss": 2.5042, "step": 714, "task_loss": 0.8733706474304199 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.743404507637024, "epoch": 0.6, "learning_rate": 2.0140845070422538e-05, "loss": 1.6826, "step": 715, "task_loss": 0.8548939824104309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9079421758651733, "epoch": 0.6, "learning_rate": 2.0169014084507042e-05, "loss": 1.535, "step": 716, "task_loss": 1.4101980924606323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8086555004119873, "epoch": 0.61, "learning_rate": 2.019718309859155e-05, "loss": 1.8027, "step": 717, "task_loss": 1.5646262168884277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6543513536453247, "epoch": 0.61, "learning_rate": 2.0225352112676056e-05, "loss": 1.8164, "step": 718, "task_loss": 0.8958156108856201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.315153121948242, "epoch": 0.61, "learning_rate": 2.0253521126760563e-05, "loss": 2.0155, "step": 719, "task_loss": 1.241468906402588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5242083072662354, "epoch": 0.61, "learning_rate": 2.028169014084507e-05, "loss": 2.0854, "step": 720, "task_loss": 1.1362740993499756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0871706008911133, "epoch": 0.61, "learning_rate": 2.0309859154929578e-05, "loss": 1.6224, "step": 721, "task_loss": 2.1623034477233887 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.692465305328369, "epoch": 0.61, "learning_rate": 2.0338028169014085e-05, "loss": 1.8191, "step": 722, "task_loss": 1.259090781211853 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6953015327453613, "epoch": 0.61, "learning_rate": 2.0366197183098592e-05, "loss": 2.0948, "step": 723, "task_loss": 0.7310623526573181 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.957451343536377, "epoch": 0.61, "learning_rate": 2.03943661971831e-05, "loss": 1.8688, "step": 724, "task_loss": 1.6863561868667603 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3317997455596924, "epoch": 0.61, "learning_rate": 2.0422535211267607e-05, "loss": 1.9568, "step": 725, "task_loss": 1.995743989944458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5580278635025024, "epoch": 0.61, "learning_rate": 2.0450704225352114e-05, "loss": 1.9575, "step": 726, "task_loss": 0.8538435697555542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7160851955413818, "epoch": 0.61, "learning_rate": 2.047887323943662e-05, "loss": 1.583, "step": 727, "task_loss": 1.0750534534454346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2943708896636963, "epoch": 0.61, "learning_rate": 2.050704225352113e-05, "loss": 1.5277, "step": 728, "task_loss": 0.5922237634658813 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.468818187713623, "epoch": 0.62, "learning_rate": 2.0535211267605633e-05, "loss": 1.9426, "step": 729, "task_loss": 1.345767855644226 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8292747735977173, "epoch": 0.62, "learning_rate": 2.0563380281690143e-05, "loss": 1.9934, "step": 730, "task_loss": 1.1075605154037476 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3624355792999268, "epoch": 0.62, "learning_rate": 2.0591549295774647e-05, "loss": 1.627, "step": 731, "task_loss": 0.9934952259063721 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1796648502349854, "epoch": 0.62, "learning_rate": 2.0619718309859158e-05, "loss": 1.7358, "step": 732, "task_loss": 0.8831654787063599 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.629849672317505, "epoch": 0.62, "learning_rate": 2.064788732394366e-05, "loss": 1.8058, "step": 733, "task_loss": 1.9990426301956177 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8622606992721558, "epoch": 0.62, "learning_rate": 2.0676056338028172e-05, "loss": 2.002, "step": 734, "task_loss": 0.9263083934783936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3185925483703613, "epoch": 0.62, "learning_rate": 2.0704225352112676e-05, "loss": 1.8303, "step": 735, "task_loss": 0.8083608150482178 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.036029100418091, "epoch": 0.62, "learning_rate": 2.0732394366197183e-05, "loss": 1.8265, "step": 736, "task_loss": 1.1688401699066162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9127273559570312, "epoch": 0.62, "learning_rate": 2.076056338028169e-05, "loss": 1.298, "step": 737, "task_loss": 1.0243377685546875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.4397921562194824, "epoch": 0.62, "learning_rate": 2.0788732394366198e-05, "loss": 2.1223, "step": 738, "task_loss": 0.6109216213226318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9202370643615723, "epoch": 0.62, "learning_rate": 2.0816901408450705e-05, "loss": 2.017, "step": 739, "task_loss": 1.091659426689148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4359852075576782, "epoch": 0.63, "learning_rate": 2.0845070422535212e-05, "loss": 1.3062, "step": 740, "task_loss": 0.8874736428260803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.1018269062042236, "epoch": 0.63, "learning_rate": 2.087323943661972e-05, "loss": 1.7736, "step": 741, "task_loss": 0.49894410371780396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0226125717163086, "epoch": 0.63, "learning_rate": 2.0901408450704227e-05, "loss": 1.6794, "step": 742, "task_loss": 1.6569796800613403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8361985683441162, "epoch": 0.63, "learning_rate": 2.0929577464788734e-05, "loss": 1.7061, "step": 743, "task_loss": 1.769020438194275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.152873992919922, "epoch": 0.63, "learning_rate": 2.095774647887324e-05, "loss": 1.7257, "step": 744, "task_loss": 2.2114181518554688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6249363422393799, "epoch": 0.63, "learning_rate": 2.098591549295775e-05, "loss": 1.6025, "step": 745, "task_loss": 0.6853619813919067 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.416859745979309, "epoch": 0.63, "learning_rate": 2.1014084507042252e-05, "loss": 1.7207, "step": 746, "task_loss": 1.701546311378479 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3313231468200684, "epoch": 0.63, "learning_rate": 2.1042253521126763e-05, "loss": 1.6129, "step": 747, "task_loss": 0.5461305975914001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8804666996002197, "epoch": 0.63, "learning_rate": 2.1070422535211267e-05, "loss": 1.8436, "step": 748, "task_loss": 1.6233875751495361 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2577314376831055, "epoch": 0.63, "learning_rate": 2.1098591549295778e-05, "loss": 1.498, "step": 749, "task_loss": 0.6826940774917603 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.1208529472351074, "epoch": 0.63, "learning_rate": 2.112676056338028e-05, "loss": 1.8787, "step": 750, "task_loss": 1.098710060119629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8421151638031006, "epoch": 0.63, "learning_rate": 2.1154929577464792e-05, "loss": 1.6578, "step": 751, "task_loss": 1.295739769935608 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5617358684539795, "epoch": 0.64, "learning_rate": 2.1183098591549296e-05, "loss": 1.745, "step": 752, "task_loss": 1.1658220291137695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7045674324035645, "epoch": 0.64, "learning_rate": 2.1211267605633803e-05, "loss": 1.6162, "step": 753, "task_loss": 1.4460628032684326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.668870449066162, "epoch": 0.64, "learning_rate": 2.123943661971831e-05, "loss": 1.4231, "step": 754, "task_loss": 0.8613659143447876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.307202696800232, "epoch": 0.64, "learning_rate": 2.1267605633802818e-05, "loss": 1.2699, "step": 755, "task_loss": 0.8277620673179626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.238795280456543, "epoch": 0.64, "learning_rate": 2.1295774647887325e-05, "loss": 1.4335, "step": 756, "task_loss": 0.8372811675071716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.566551685333252, "epoch": 0.64, "learning_rate": 2.1323943661971832e-05, "loss": 1.9548, "step": 757, "task_loss": 1.6693578958511353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4045202732086182, "epoch": 0.64, "learning_rate": 2.135211267605634e-05, "loss": 1.5148, "step": 758, "task_loss": 0.499186635017395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4801965951919556, "epoch": 0.64, "learning_rate": 2.1380281690140847e-05, "loss": 1.3672, "step": 759, "task_loss": 0.266851007938385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.640653610229492, "epoch": 0.64, "learning_rate": 2.1408450704225354e-05, "loss": 1.8671, "step": 760, "task_loss": 1.0046682357788086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.330306053161621, "epoch": 0.64, "learning_rate": 2.143661971830986e-05, "loss": 1.7277, "step": 761, "task_loss": 0.8563199043273926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6619977951049805, "epoch": 0.64, "learning_rate": 2.146478873239437e-05, "loss": 1.3763, "step": 762, "task_loss": 1.4969401359558105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7756943702697754, "epoch": 0.64, "learning_rate": 2.1492957746478876e-05, "loss": 1.6836, "step": 763, "task_loss": 0.5833191871643066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5706725120544434, "epoch": 0.65, "learning_rate": 2.1521126760563383e-05, "loss": 1.5924, "step": 764, "task_loss": 1.0970302820205688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5795881748199463, "epoch": 0.65, "learning_rate": 2.1549295774647887e-05, "loss": 1.9887, "step": 765, "task_loss": 0.8463152050971985 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7526946067810059, "epoch": 0.65, "learning_rate": 2.1577464788732397e-05, "loss": 1.5982, "step": 766, "task_loss": 1.1121677160263062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3968513011932373, "epoch": 0.65, "learning_rate": 2.16056338028169e-05, "loss": 1.7718, "step": 767, "task_loss": 1.260400414466858 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0161058902740479, "epoch": 0.65, "learning_rate": 2.1633802816901412e-05, "loss": 1.6668, "step": 768, "task_loss": 0.6748580932617188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2036612033843994, "epoch": 0.65, "learning_rate": 2.1661971830985916e-05, "loss": 1.2715, "step": 769, "task_loss": 1.628238320350647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.245849847793579, "epoch": 0.65, "learning_rate": 2.1690140845070423e-05, "loss": 1.5291, "step": 770, "task_loss": 1.7865175008773804 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.867519736289978, "epoch": 0.65, "learning_rate": 2.171830985915493e-05, "loss": 1.7824, "step": 771, "task_loss": 0.876433253288269 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.55698823928833, "epoch": 0.65, "learning_rate": 2.1746478873239438e-05, "loss": 1.4318, "step": 772, "task_loss": 1.11099374294281 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5812749862670898, "epoch": 0.65, "learning_rate": 2.1774647887323945e-05, "loss": 1.8844, "step": 773, "task_loss": 0.7156081199645996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6603411436080933, "epoch": 0.65, "learning_rate": 2.1802816901408452e-05, "loss": 1.4696, "step": 774, "task_loss": 1.7654619216918945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4288454055786133, "epoch": 0.65, "learning_rate": 2.1830985915492956e-05, "loss": 1.331, "step": 775, "task_loss": 0.7231191992759705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.2420077323913574, "epoch": 0.66, "learning_rate": 2.1859154929577467e-05, "loss": 1.7619, "step": 776, "task_loss": 1.47881281375885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2709290981292725, "epoch": 0.66, "learning_rate": 2.188732394366197e-05, "loss": 1.4859, "step": 777, "task_loss": 0.27060213685035706 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9081125259399414, "epoch": 0.66, "learning_rate": 2.191549295774648e-05, "loss": 1.3351, "step": 778, "task_loss": 1.5613409280776978 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.337864875793457, "epoch": 0.66, "learning_rate": 2.1943661971830985e-05, "loss": 1.5537, "step": 779, "task_loss": 1.3064930438995361 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3826532363891602, "epoch": 0.66, "learning_rate": 2.1971830985915496e-05, "loss": 1.6065, "step": 780, "task_loss": 0.5846068859100342 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2934510707855225, "epoch": 0.66, "learning_rate": 2.2000000000000003e-05, "loss": 1.5178, "step": 781, "task_loss": 0.848359227180481 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3813867568969727, "epoch": 0.66, "learning_rate": 2.2028169014084507e-05, "loss": 1.5789, "step": 782, "task_loss": 1.7373712062835693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.390201449394226, "epoch": 0.66, "learning_rate": 2.2056338028169017e-05, "loss": 1.2485, "step": 783, "task_loss": 0.4696228504180908 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.4078705310821533, "epoch": 0.66, "learning_rate": 2.208450704225352e-05, "loss": 1.6105, "step": 784, "task_loss": 0.9727026224136353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.847015142440796, "epoch": 0.66, "learning_rate": 2.2112676056338032e-05, "loss": 1.3837, "step": 785, "task_loss": 1.2509067058563232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4191980361938477, "epoch": 0.66, "learning_rate": 2.2140845070422536e-05, "loss": 1.5264, "step": 786, "task_loss": 0.4853179454803467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9770253896713257, "epoch": 0.66, "learning_rate": 2.2169014084507043e-05, "loss": 1.4127, "step": 787, "task_loss": 0.5723888874053955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4197728633880615, "epoch": 0.67, "learning_rate": 2.219718309859155e-05, "loss": 1.6484, "step": 788, "task_loss": 0.8161102533340454 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8536734580993652, "epoch": 0.67, "learning_rate": 2.2225352112676057e-05, "loss": 1.4611, "step": 789, "task_loss": 1.018897294998169 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3663787841796875, "epoch": 0.67, "learning_rate": 2.2253521126760565e-05, "loss": 1.5101, "step": 790, "task_loss": 1.1776089668273926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7453126907348633, "epoch": 0.67, "learning_rate": 2.2281690140845072e-05, "loss": 1.7365, "step": 791, "task_loss": 0.530036449432373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.253589391708374, "epoch": 0.67, "learning_rate": 2.230985915492958e-05, "loss": 1.5796, "step": 792, "task_loss": 1.0725141763687134 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3179899454116821, "epoch": 0.67, "learning_rate": 2.2338028169014086e-05, "loss": 1.4805, "step": 793, "task_loss": 0.626964807510376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0312771797180176, "epoch": 0.67, "learning_rate": 2.236619718309859e-05, "loss": 1.4731, "step": 794, "task_loss": 1.5911962985992432 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.5356333255767822, "epoch": 0.67, "learning_rate": 2.23943661971831e-05, "loss": 1.6017, "step": 795, "task_loss": 1.243147373199463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7455284595489502, "epoch": 0.67, "learning_rate": 2.2422535211267605e-05, "loss": 1.5366, "step": 796, "task_loss": 0.9062787294387817 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.098499059677124, "epoch": 0.67, "learning_rate": 2.2450704225352115e-05, "loss": 1.6216, "step": 797, "task_loss": 1.412013292312622 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.6238954067230225, "epoch": 0.67, "learning_rate": 2.247887323943662e-05, "loss": 1.7185, "step": 798, "task_loss": 2.6273739337921143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.001409649848938, "epoch": 0.67, "learning_rate": 2.2507042253521127e-05, "loss": 1.5183, "step": 799, "task_loss": 0.5990715026855469 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3451969623565674, "epoch": 0.68, "learning_rate": 2.2535211267605634e-05, "loss": 1.6107, "step": 800, "task_loss": 0.7545828819274902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9218698740005493, "epoch": 0.68, "learning_rate": 2.256338028169014e-05, "loss": 1.4031, "step": 801, "task_loss": 0.35016340017318726 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3398871421813965, "epoch": 0.68, "learning_rate": 2.2591549295774648e-05, "loss": 1.718, "step": 802, "task_loss": 1.450812578201294 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5056934356689453, "epoch": 0.68, "learning_rate": 2.2619718309859156e-05, "loss": 1.7053, "step": 803, "task_loss": 1.2730746269226074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5979478359222412, "epoch": 0.68, "learning_rate": 2.2647887323943663e-05, "loss": 1.2744, "step": 804, "task_loss": 1.5955917835235596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.657163143157959, "epoch": 0.68, "learning_rate": 2.267605633802817e-05, "loss": 1.4331, "step": 805, "task_loss": 0.8205367922782898 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.1600639820098877, "epoch": 0.68, "learning_rate": 2.2704225352112677e-05, "loss": 1.6574, "step": 806, "task_loss": 1.425527572631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.569136381149292, "epoch": 0.68, "learning_rate": 2.2732394366197185e-05, "loss": 1.4025, "step": 807, "task_loss": 0.8261222243309021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6623165607452393, "epoch": 0.68, "learning_rate": 2.2760563380281692e-05, "loss": 1.6266, "step": 808, "task_loss": 1.5448304414749146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.3062362670898438, "epoch": 0.68, "learning_rate": 2.27887323943662e-05, "loss": 1.5418, "step": 809, "task_loss": 1.2962450981140137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8435287475585938, "epoch": 0.68, "learning_rate": 2.2816901408450706e-05, "loss": 1.6467, "step": 810, "task_loss": 1.716566562652588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9329716563224792, "epoch": 0.69, "learning_rate": 2.284507042253521e-05, "loss": 1.1829, "step": 811, "task_loss": 0.2455202341079712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6946394443511963, "epoch": 0.69, "learning_rate": 2.287323943661972e-05, "loss": 1.4817, "step": 812, "task_loss": 0.7265291213989258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.35438072681427, "epoch": 0.69, "learning_rate": 2.2901408450704225e-05, "loss": 1.4303, "step": 813, "task_loss": 0.32950541377067566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5364727973937988, "epoch": 0.69, "learning_rate": 2.2929577464788735e-05, "loss": 1.3677, "step": 814, "task_loss": 1.0794832706451416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4814695119857788, "epoch": 0.69, "learning_rate": 2.295774647887324e-05, "loss": 1.6548, "step": 815, "task_loss": 0.6691979169845581 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7386112809181213, "epoch": 0.69, "learning_rate": 2.2985915492957746e-05, "loss": 1.2315, "step": 816, "task_loss": 0.45963701605796814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9637730717658997, "epoch": 0.69, "learning_rate": 2.3014084507042254e-05, "loss": 1.2276, "step": 817, "task_loss": 0.9293928742408752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0221800804138184, "epoch": 0.69, "learning_rate": 2.304225352112676e-05, "loss": 1.7697, "step": 818, "task_loss": 0.9548535943031311 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6787159442901611, "epoch": 0.69, "learning_rate": 2.3070422535211268e-05, "loss": 1.3045, "step": 819, "task_loss": 1.167372703552246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.068692684173584, "epoch": 0.69, "learning_rate": 2.3098591549295775e-05, "loss": 1.7123, "step": 820, "task_loss": 1.0823863744735718 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.105148434638977, "epoch": 0.69, "learning_rate": 2.3126760563380283e-05, "loss": 1.3427, "step": 821, "task_loss": 0.870733380317688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7189647555351257, "epoch": 0.69, "learning_rate": 2.315492957746479e-05, "loss": 1.2501, "step": 822, "task_loss": 0.9020413160324097 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1557564735412598, "epoch": 0.7, "learning_rate": 2.3183098591549297e-05, "loss": 1.2237, "step": 823, "task_loss": 0.7066966891288757 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9354766607284546, "epoch": 0.7, "learning_rate": 2.3211267605633804e-05, "loss": 1.5398, "step": 824, "task_loss": 1.0245102643966675 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2383958101272583, "epoch": 0.7, "learning_rate": 2.323943661971831e-05, "loss": 1.1386, "step": 825, "task_loss": 0.4949999153614044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0761008262634277, "epoch": 0.7, "learning_rate": 2.326760563380282e-05, "loss": 1.2574, "step": 826, "task_loss": 0.7356624603271484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6209033727645874, "epoch": 0.7, "learning_rate": 2.3295774647887326e-05, "loss": 1.4434, "step": 827, "task_loss": 1.3619142770767212 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9905048608779907, "epoch": 0.7, "learning_rate": 2.332394366197183e-05, "loss": 1.4605, "step": 828, "task_loss": 0.5368533134460449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3130991458892822, "epoch": 0.7, "learning_rate": 2.335211267605634e-05, "loss": 1.4246, "step": 829, "task_loss": 0.8089421987533569 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6364295482635498, "epoch": 0.7, "learning_rate": 2.3380281690140845e-05, "loss": 1.4156, "step": 830, "task_loss": 1.0288071632385254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4404677152633667, "epoch": 0.7, "learning_rate": 2.3408450704225355e-05, "loss": 1.3808, "step": 831, "task_loss": 1.5474152565002441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3175069093704224, "epoch": 0.7, "learning_rate": 2.343661971830986e-05, "loss": 1.2124, "step": 832, "task_loss": 1.4574071168899536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1590447425842285, "epoch": 0.7, "learning_rate": 2.3464788732394366e-05, "loss": 1.5516, "step": 833, "task_loss": 2.39730167388916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.757013201713562, "epoch": 0.7, "learning_rate": 2.3492957746478874e-05, "loss": 1.4326, "step": 834, "task_loss": 1.2907358407974243 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5434492826461792, "epoch": 0.71, "learning_rate": 2.352112676056338e-05, "loss": 1.4185, "step": 835, "task_loss": 1.3717619180679321 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6866276264190674, "epoch": 0.71, "learning_rate": 2.3549295774647888e-05, "loss": 1.3792, "step": 836, "task_loss": 1.511733889579773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9707732796669006, "epoch": 0.71, "learning_rate": 2.3577464788732395e-05, "loss": 1.5399, "step": 837, "task_loss": 0.6395320296287537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3975028991699219, "epoch": 0.71, "learning_rate": 2.3605633802816902e-05, "loss": 1.5803, "step": 838, "task_loss": 1.253529667854309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.651561975479126, "epoch": 0.71, "learning_rate": 2.363380281690141e-05, "loss": 1.5676, "step": 839, "task_loss": 1.01663076877594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0138477087020874, "epoch": 0.71, "learning_rate": 2.3661971830985917e-05, "loss": 1.1716, "step": 840, "task_loss": 1.3856985569000244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5163414478302002, "epoch": 0.71, "learning_rate": 2.3690140845070424e-05, "loss": 1.7227, "step": 841, "task_loss": 1.0217088460922241 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.1797614097595215, "epoch": 0.71, "learning_rate": 2.371830985915493e-05, "loss": 1.5388, "step": 842, "task_loss": 1.0634114742279053 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.705564558506012, "epoch": 0.71, "learning_rate": 2.374647887323944e-05, "loss": 1.2685, "step": 843, "task_loss": 0.4624873697757721 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4101500511169434, "epoch": 0.71, "learning_rate": 2.3774647887323946e-05, "loss": 1.4635, "step": 844, "task_loss": 0.45105597376823425 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.998557686805725, "epoch": 0.71, "learning_rate": 2.380281690140845e-05, "loss": 1.3836, "step": 845, "task_loss": 0.7943804264068604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.304697871208191, "epoch": 0.71, "learning_rate": 2.383098591549296e-05, "loss": 1.3003, "step": 846, "task_loss": 0.6109516024589539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4514689445495605, "epoch": 0.72, "learning_rate": 2.3859154929577464e-05, "loss": 1.264, "step": 847, "task_loss": 1.6925387382507324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3824222087860107, "epoch": 0.72, "learning_rate": 2.3887323943661975e-05, "loss": 1.4993, "step": 848, "task_loss": 1.5330089330673218 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.925555944442749, "epoch": 0.72, "learning_rate": 2.391549295774648e-05, "loss": 1.3338, "step": 849, "task_loss": 0.9188722968101501 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9074881076812744, "epoch": 0.72, "learning_rate": 2.3943661971830986e-05, "loss": 1.1664, "step": 850, "task_loss": 1.4816126823425293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3937077522277832, "epoch": 0.72, "learning_rate": 2.3971830985915493e-05, "loss": 1.5354, "step": 851, "task_loss": 1.1468114852905273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.408233642578125, "epoch": 0.72, "learning_rate": 2.4e-05, "loss": 1.2242, "step": 852, "task_loss": 0.8182432651519775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5206315517425537, "epoch": 0.72, "learning_rate": 2.4028169014084508e-05, "loss": 1.5255, "step": 853, "task_loss": 0.5622049570083618 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6485689878463745, "epoch": 0.72, "learning_rate": 2.4056338028169015e-05, "loss": 1.1447, "step": 854, "task_loss": 0.689480721950531 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5888588428497314, "epoch": 0.72, "learning_rate": 2.4084507042253522e-05, "loss": 1.4836, "step": 855, "task_loss": 1.5076277256011963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5022221803665161, "epoch": 0.72, "learning_rate": 2.411267605633803e-05, "loss": 1.3735, "step": 856, "task_loss": 0.5148055553436279 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.292418122291565, "epoch": 0.72, "learning_rate": 2.4140845070422537e-05, "loss": 1.2752, "step": 857, "task_loss": 0.5999670624732971 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3422482013702393, "epoch": 0.72, "learning_rate": 2.4169014084507044e-05, "loss": 1.4404, "step": 858, "task_loss": 1.231388807296753 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3558201789855957, "epoch": 0.73, "learning_rate": 2.419718309859155e-05, "loss": 1.384, "step": 859, "task_loss": 1.6106853485107422 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7219918966293335, "epoch": 0.73, "learning_rate": 2.422535211267606e-05, "loss": 1.1247, "step": 860, "task_loss": 0.6331835985183716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.9095728397369385, "epoch": 0.73, "learning_rate": 2.4253521126760566e-05, "loss": 1.5826, "step": 861, "task_loss": 0.570243239402771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8224810361862183, "epoch": 0.73, "learning_rate": 2.428169014084507e-05, "loss": 1.1857, "step": 862, "task_loss": 1.2648526430130005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6798423528671265, "epoch": 0.73, "learning_rate": 2.430985915492958e-05, "loss": 1.2541, "step": 863, "task_loss": 1.7378650903701782 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4397226572036743, "epoch": 0.73, "learning_rate": 2.4338028169014084e-05, "loss": 1.1948, "step": 864, "task_loss": 0.5610358715057373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0516607761383057, "epoch": 0.73, "learning_rate": 2.4366197183098595e-05, "loss": 1.1383, "step": 865, "task_loss": 1.3350591659545898 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7584878206253052, "epoch": 0.73, "learning_rate": 2.43943661971831e-05, "loss": 1.0159, "step": 866, "task_loss": 0.38145995140075684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9113258123397827, "epoch": 0.73, "learning_rate": 2.442253521126761e-05, "loss": 1.0331, "step": 867, "task_loss": 1.5229811668395996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2570816278457642, "epoch": 0.73, "learning_rate": 2.4450704225352113e-05, "loss": 1.5074, "step": 868, "task_loss": 0.469443678855896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.004331588745117, "epoch": 0.73, "learning_rate": 2.447887323943662e-05, "loss": 1.5952, "step": 869, "task_loss": 0.614041805267334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.62593674659729, "epoch": 0.73, "learning_rate": 2.4507042253521128e-05, "loss": 1.5487, "step": 870, "task_loss": 1.087019920349121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.335922122001648, "epoch": 0.74, "learning_rate": 2.4535211267605635e-05, "loss": 1.159, "step": 871, "task_loss": 1.958196759223938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0230236053466797, "epoch": 0.74, "learning_rate": 2.4563380281690142e-05, "loss": 1.3702, "step": 872, "task_loss": 1.614214301109314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4235347509384155, "epoch": 0.74, "learning_rate": 2.459154929577465e-05, "loss": 1.2183, "step": 873, "task_loss": 0.7941752672195435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.139530897140503, "epoch": 0.74, "learning_rate": 2.4619718309859153e-05, "loss": 1.1795, "step": 874, "task_loss": 0.6531791687011719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9489515423774719, "epoch": 0.74, "learning_rate": 2.4647887323943664e-05, "loss": 1.0445, "step": 875, "task_loss": 0.571410596370697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5026335716247559, "epoch": 0.74, "learning_rate": 2.4676056338028168e-05, "loss": 1.2577, "step": 876, "task_loss": 0.8769711256027222 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.63930344581604, "epoch": 0.74, "learning_rate": 2.470422535211268e-05, "loss": 1.5209, "step": 877, "task_loss": 1.435776948928833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3810970783233643, "epoch": 0.74, "learning_rate": 2.4732394366197186e-05, "loss": 1.3897, "step": 878, "task_loss": 1.0480715036392212 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4205354452133179, "epoch": 0.74, "learning_rate": 2.476056338028169e-05, "loss": 1.443, "step": 879, "task_loss": 0.888705313205719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5702440738677979, "epoch": 0.74, "learning_rate": 2.47887323943662e-05, "loss": 1.5236, "step": 880, "task_loss": 1.5710365772247314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9568125009536743, "epoch": 0.74, "learning_rate": 2.4816901408450704e-05, "loss": 0.9789, "step": 881, "task_loss": 0.311002641916275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.064399003982544, "epoch": 0.75, "learning_rate": 2.4845070422535215e-05, "loss": 1.4095, "step": 882, "task_loss": 0.4654320478439331 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.463811993598938, "epoch": 0.75, "learning_rate": 2.487323943661972e-05, "loss": 1.2228, "step": 883, "task_loss": 1.0349777936935425 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.330260992050171, "epoch": 0.75, "learning_rate": 2.490140845070423e-05, "loss": 1.3176, "step": 884, "task_loss": 0.3665597140789032 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3046895265579224, "epoch": 0.75, "learning_rate": 2.4929577464788733e-05, "loss": 1.1077, "step": 885, "task_loss": 1.8112586736679077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1157701015472412, "epoch": 0.75, "learning_rate": 2.495774647887324e-05, "loss": 1.1772, "step": 886, "task_loss": 0.8165785670280457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1765626668930054, "epoch": 0.75, "learning_rate": 2.4985915492957748e-05, "loss": 1.2277, "step": 887, "task_loss": 0.959071934223175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.39210844039917, "epoch": 0.75, "learning_rate": 2.5014084507042258e-05, "loss": 1.1557, "step": 888, "task_loss": 0.6303380131721497 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5112745761871338, "epoch": 0.75, "learning_rate": 2.5042253521126762e-05, "loss": 1.2236, "step": 889, "task_loss": 0.8227658271789551 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.458530306816101, "epoch": 0.75, "learning_rate": 2.507042253521127e-05, "loss": 1.3341, "step": 890, "task_loss": 0.6040553450584412 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.35792076587677, "epoch": 0.75, "learning_rate": 2.5098591549295773e-05, "loss": 1.3678, "step": 891, "task_loss": 1.818406343460083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9112638831138611, "epoch": 0.75, "learning_rate": 2.512676056338028e-05, "loss": 0.9689, "step": 892, "task_loss": 1.2221039533615112 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8488353490829468, "epoch": 0.75, "learning_rate": 2.515492957746479e-05, "loss": 1.3685, "step": 893, "task_loss": 0.5399192571640015 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1704118251800537, "epoch": 0.76, "learning_rate": 2.51830985915493e-05, "loss": 1.3697, "step": 894, "task_loss": 1.6677066087722778 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3526849746704102, "epoch": 0.76, "learning_rate": 2.5211267605633802e-05, "loss": 1.236, "step": 895, "task_loss": 0.17719915509223938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6340664625167847, "epoch": 0.76, "learning_rate": 2.523943661971831e-05, "loss": 1.4097, "step": 896, "task_loss": 0.949924886226654 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.204627513885498, "epoch": 0.76, "learning_rate": 2.5267605633802817e-05, "loss": 1.2763, "step": 897, "task_loss": 1.313928246498108 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.073024868965149, "epoch": 0.76, "learning_rate": 2.5295774647887327e-05, "loss": 1.2395, "step": 898, "task_loss": 0.3629971146583557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0387213230133057, "epoch": 0.76, "learning_rate": 2.5323943661971835e-05, "loss": 1.2244, "step": 899, "task_loss": 1.6787906885147095 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0529553890228271, "epoch": 0.76, "learning_rate": 2.535211267605634e-05, "loss": 0.9889, "step": 900, "task_loss": 0.47505414485931396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0840511322021484, "epoch": 0.76, "learning_rate": 2.5380281690140846e-05, "loss": 1.2137, "step": 901, "task_loss": 1.234485149383545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3794633150100708, "epoch": 0.76, "learning_rate": 2.540845070422535e-05, "loss": 1.1226, "step": 902, "task_loss": 1.2335163354873657 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5171444416046143, "epoch": 0.76, "learning_rate": 2.5436619718309864e-05, "loss": 1.335, "step": 903, "task_loss": 1.3146016597747803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0462312698364258, "epoch": 0.76, "learning_rate": 2.5464788732394367e-05, "loss": 1.2322, "step": 904, "task_loss": 1.0603787899017334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6236159801483154, "epoch": 0.76, "learning_rate": 2.5492957746478875e-05, "loss": 1.1144, "step": 905, "task_loss": 1.0404926538467407 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4294679164886475, "epoch": 0.77, "learning_rate": 2.552112676056338e-05, "loss": 1.175, "step": 906, "task_loss": 1.8538323640823364 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2322165966033936, "epoch": 0.77, "learning_rate": 2.5549295774647893e-05, "loss": 1.3587, "step": 907, "task_loss": 0.8721309304237366 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.403357744216919, "epoch": 0.77, "learning_rate": 2.5577464788732396e-05, "loss": 1.1331, "step": 908, "task_loss": 1.9969061613082886 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1900050640106201, "epoch": 0.77, "learning_rate": 2.5605633802816904e-05, "loss": 1.2502, "step": 909, "task_loss": 0.9544848799705505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2048723697662354, "epoch": 0.77, "learning_rate": 2.5633802816901408e-05, "loss": 1.3201, "step": 910, "task_loss": 0.8094683885574341 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2010557651519775, "epoch": 0.77, "learning_rate": 2.5661971830985915e-05, "loss": 1.4245, "step": 911, "task_loss": 1.0904158353805542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4120368957519531, "epoch": 0.77, "learning_rate": 2.5690140845070425e-05, "loss": 1.4895, "step": 912, "task_loss": 0.9537250399589539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.439354419708252, "epoch": 0.77, "learning_rate": 2.5718309859154933e-05, "loss": 1.259, "step": 913, "task_loss": 1.6700514554977417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0164827108383179, "epoch": 0.77, "learning_rate": 2.5746478873239437e-05, "loss": 1.1379, "step": 914, "task_loss": 1.445932149887085 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1824073791503906, "epoch": 0.77, "learning_rate": 2.5774647887323944e-05, "loss": 0.9786, "step": 915, "task_loss": 1.0062981843948364 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6738599538803101, "epoch": 0.77, "learning_rate": 2.580281690140845e-05, "loss": 1.0245, "step": 916, "task_loss": 0.7041023969650269 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3265666961669922, "epoch": 0.77, "learning_rate": 2.583098591549296e-05, "loss": 1.0805, "step": 917, "task_loss": 1.8456239700317383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.449730634689331, "epoch": 0.78, "learning_rate": 2.5859154929577466e-05, "loss": 1.2291, "step": 918, "task_loss": 1.549756407737732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8219128847122192, "epoch": 0.78, "learning_rate": 2.5887323943661973e-05, "loss": 1.0396, "step": 919, "task_loss": 0.4250470697879791 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5171246528625488, "epoch": 0.78, "learning_rate": 2.591549295774648e-05, "loss": 1.1841, "step": 920, "task_loss": 1.1608574390411377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4038668870925903, "epoch": 0.78, "learning_rate": 2.5943661971830984e-05, "loss": 1.3027, "step": 921, "task_loss": 1.2467811107635498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9380587339401245, "epoch": 0.78, "learning_rate": 2.5971830985915498e-05, "loss": 1.0248, "step": 922, "task_loss": 0.18046057224273682 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1595680713653564, "epoch": 0.78, "learning_rate": 2.6000000000000002e-05, "loss": 1.0287, "step": 923, "task_loss": 1.7187696695327759 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9955464601516724, "epoch": 0.78, "learning_rate": 2.602816901408451e-05, "loss": 1.3994, "step": 924, "task_loss": 1.614405870437622 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2703027725219727, "epoch": 0.78, "learning_rate": 2.6056338028169013e-05, "loss": 1.722, "step": 925, "task_loss": 1.4040608406066895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.737863302230835, "epoch": 0.78, "learning_rate": 2.608450704225352e-05, "loss": 0.9071, "step": 926, "task_loss": 0.6614607572555542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.769914150238037, "epoch": 0.78, "learning_rate": 2.611267605633803e-05, "loss": 1.1594, "step": 927, "task_loss": 0.699809730052948 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1456923484802246, "epoch": 0.78, "learning_rate": 2.6140845070422538e-05, "loss": 1.2817, "step": 928, "task_loss": 0.499350905418396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2274281978607178, "epoch": 0.78, "learning_rate": 2.6169014084507042e-05, "loss": 1.1073, "step": 929, "task_loss": 0.5901204943656921 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9093064069747925, "epoch": 0.79, "learning_rate": 2.619718309859155e-05, "loss": 1.099, "step": 930, "task_loss": 1.5155633687973022 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8206389546394348, "epoch": 0.79, "learning_rate": 2.6225352112676056e-05, "loss": 1.0998, "step": 931, "task_loss": 0.8410032987594604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5730984210968018, "epoch": 0.79, "learning_rate": 2.6253521126760567e-05, "loss": 0.8708, "step": 932, "task_loss": 0.24780869483947754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5470257997512817, "epoch": 0.79, "learning_rate": 2.628169014084507e-05, "loss": 1.297, "step": 933, "task_loss": 1.8907488584518433 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.844429075717926, "epoch": 0.79, "learning_rate": 2.6309859154929578e-05, "loss": 0.9608, "step": 934, "task_loss": 1.0763490200042725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6260032653808594, "epoch": 0.79, "learning_rate": 2.6338028169014085e-05, "loss": 1.3468, "step": 935, "task_loss": 0.9275493025779724 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0840373039245605, "epoch": 0.79, "learning_rate": 2.6366197183098596e-05, "loss": 1.0192, "step": 936, "task_loss": 0.4445825517177582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1052316427230835, "epoch": 0.79, "learning_rate": 2.63943661971831e-05, "loss": 1.2424, "step": 937, "task_loss": 0.6064824461936951 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9820965528488159, "epoch": 0.79, "learning_rate": 2.6422535211267607e-05, "loss": 1.0387, "step": 938, "task_loss": 0.9221377372741699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1637839078903198, "epoch": 0.79, "learning_rate": 2.6450704225352114e-05, "loss": 1.0728, "step": 939, "task_loss": 0.44556406140327454 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.01059091091156, "epoch": 0.79, "learning_rate": 2.6478873239436618e-05, "loss": 1.171, "step": 940, "task_loss": 1.3838677406311035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7012059092521667, "epoch": 0.79, "learning_rate": 2.650704225352113e-05, "loss": 0.9615, "step": 941, "task_loss": 0.6670460104942322 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.413044810295105, "epoch": 0.8, "learning_rate": 2.6535211267605636e-05, "loss": 1.1837, "step": 942, "task_loss": 1.0936989784240723 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9680198431015015, "epoch": 0.8, "learning_rate": 2.6563380281690143e-05, "loss": 1.1932, "step": 943, "task_loss": 0.25028887391090393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7011686563491821, "epoch": 0.8, "learning_rate": 2.6591549295774647e-05, "loss": 1.1243, "step": 944, "task_loss": 1.056843876838684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0838037729263306, "epoch": 0.8, "learning_rate": 2.6619718309859155e-05, "loss": 1.2691, "step": 945, "task_loss": 0.9585646390914917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.091071605682373, "epoch": 0.8, "learning_rate": 2.6647887323943665e-05, "loss": 0.9392, "step": 946, "task_loss": 0.7580676674842834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1333634853363037, "epoch": 0.8, "learning_rate": 2.6676056338028172e-05, "loss": 1.1695, "step": 947, "task_loss": 0.9213865399360657 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8171094059944153, "epoch": 0.8, "learning_rate": 2.6704225352112676e-05, "loss": 1.1155, "step": 948, "task_loss": 1.906986951828003 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0958267450332642, "epoch": 0.8, "learning_rate": 2.6732394366197184e-05, "loss": 1.364, "step": 949, "task_loss": 0.7035043239593506 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.392202377319336, "epoch": 0.8, "learning_rate": 2.676056338028169e-05, "loss": 1.0783, "step": 950, "task_loss": 0.3435916006565094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0802960395812988, "epoch": 0.8, "learning_rate": 2.67887323943662e-05, "loss": 0.95, "step": 951, "task_loss": 0.7352214455604553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0303822755813599, "epoch": 0.8, "learning_rate": 2.6816901408450705e-05, "loss": 1.2519, "step": 952, "task_loss": 1.150357723236084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5441871881484985, "epoch": 0.81, "learning_rate": 2.6845070422535213e-05, "loss": 0.9342, "step": 953, "task_loss": 0.5027093291282654 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3063628673553467, "epoch": 0.81, "learning_rate": 2.687323943661972e-05, "loss": 1.1551, "step": 954, "task_loss": 1.8270442485809326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1574537754058838, "epoch": 0.81, "learning_rate": 2.6901408450704224e-05, "loss": 0.9894, "step": 955, "task_loss": 0.6725310683250427 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1681973934173584, "epoch": 0.81, "learning_rate": 2.6929577464788734e-05, "loss": 1.1553, "step": 956, "task_loss": 0.4665925204753876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4041029214859009, "epoch": 0.81, "learning_rate": 2.695774647887324e-05, "loss": 1.1768, "step": 957, "task_loss": 1.3812992572784424 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9715734124183655, "epoch": 0.81, "learning_rate": 2.698591549295775e-05, "loss": 0.9517, "step": 958, "task_loss": 1.1768220663070679 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9792338609695435, "epoch": 0.81, "learning_rate": 2.7014084507042253e-05, "loss": 1.1221, "step": 959, "task_loss": 0.8877928853034973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1193790435791016, "epoch": 0.81, "learning_rate": 2.704225352112676e-05, "loss": 1.3041, "step": 960, "task_loss": 1.2529122829437256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5852965116500854, "epoch": 0.81, "learning_rate": 2.707042253521127e-05, "loss": 1.2136, "step": 961, "task_loss": 0.9345572590827942 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5727506875991821, "epoch": 0.81, "learning_rate": 2.7098591549295778e-05, "loss": 1.2202, "step": 962, "task_loss": 1.128395438194275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.845315158367157, "epoch": 0.81, "learning_rate": 2.712676056338028e-05, "loss": 1.2349, "step": 963, "task_loss": 1.1872464418411255 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5244837999343872, "epoch": 0.81, "learning_rate": 2.715492957746479e-05, "loss": 1.2224, "step": 964, "task_loss": 0.4301453232765198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5381828546524048, "epoch": 0.82, "learning_rate": 2.71830985915493e-05, "loss": 1.1361, "step": 965, "task_loss": 1.2516359090805054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1090247631072998, "epoch": 0.82, "learning_rate": 2.7211267605633807e-05, "loss": 1.0238, "step": 966, "task_loss": 1.949077844619751 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2344815731048584, "epoch": 0.82, "learning_rate": 2.723943661971831e-05, "loss": 1.3702, "step": 967, "task_loss": 1.4908798933029175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4238150119781494, "epoch": 0.82, "learning_rate": 2.7267605633802818e-05, "loss": 1.1434, "step": 968, "task_loss": 0.5671489834785461 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2178049087524414, "epoch": 0.82, "learning_rate": 2.7295774647887322e-05, "loss": 1.1111, "step": 969, "task_loss": 1.7470403909683228 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8715372681617737, "epoch": 0.82, "learning_rate": 2.7323943661971836e-05, "loss": 0.8615, "step": 970, "task_loss": 0.8336560726165771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.0274977684020996, "epoch": 0.82, "learning_rate": 2.735211267605634e-05, "loss": 1.5157, "step": 971, "task_loss": 1.3173967599868774 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8729153275489807, "epoch": 0.82, "learning_rate": 2.7380281690140847e-05, "loss": 0.8673, "step": 972, "task_loss": 0.4300232231616974 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.345395803451538, "epoch": 0.82, "learning_rate": 2.7408450704225354e-05, "loss": 1.1339, "step": 973, "task_loss": 0.7665396332740784 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.247422695159912, "epoch": 0.82, "learning_rate": 2.7436619718309858e-05, "loss": 1.1425, "step": 974, "task_loss": 0.6175227165222168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8730621337890625, "epoch": 0.82, "learning_rate": 2.746478873239437e-05, "loss": 0.9373, "step": 975, "task_loss": 0.47332218289375305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.169206976890564, "epoch": 0.82, "learning_rate": 2.7492957746478876e-05, "loss": 1.2858, "step": 976, "task_loss": 0.697471559047699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1268572807312012, "epoch": 0.83, "learning_rate": 2.7521126760563383e-05, "loss": 1.1447, "step": 977, "task_loss": 1.4639220237731934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9856197834014893, "epoch": 0.83, "learning_rate": 2.7549295774647887e-05, "loss": 0.8558, "step": 978, "task_loss": 1.0203348398208618 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1539313793182373, "epoch": 0.83, "learning_rate": 2.7577464788732394e-05, "loss": 1.076, "step": 979, "task_loss": 0.5245310068130493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3737823963165283, "epoch": 0.83, "learning_rate": 2.7605633802816905e-05, "loss": 1.1323, "step": 980, "task_loss": 1.3838545083999634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2848079204559326, "epoch": 0.83, "learning_rate": 2.7633802816901412e-05, "loss": 1.2474, "step": 981, "task_loss": 1.6033955812454224 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0237921476364136, "epoch": 0.83, "learning_rate": 2.7661971830985916e-05, "loss": 1.0589, "step": 982, "task_loss": 0.4422343671321869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1978263854980469, "epoch": 0.83, "learning_rate": 2.7690140845070423e-05, "loss": 1.1205, "step": 983, "task_loss": 1.0541304349899292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7401732206344604, "epoch": 0.83, "learning_rate": 2.7718309859154927e-05, "loss": 1.0931, "step": 984, "task_loss": 0.9673908948898315 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8845313787460327, "epoch": 0.83, "learning_rate": 2.774647887323944e-05, "loss": 0.9096, "step": 985, "task_loss": 1.4991430044174194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0815949440002441, "epoch": 0.83, "learning_rate": 2.7774647887323945e-05, "loss": 1.1399, "step": 986, "task_loss": 1.4996142387390137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.417172908782959, "epoch": 0.83, "learning_rate": 2.7802816901408452e-05, "loss": 1.1624, "step": 987, "task_loss": 1.3256914615631104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7606383562088013, "epoch": 0.83, "learning_rate": 2.7830985915492956e-05, "loss": 1.2739, "step": 988, "task_loss": 1.547696590423584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1385152339935303, "epoch": 0.84, "learning_rate": 2.7859154929577463e-05, "loss": 0.9969, "step": 989, "task_loss": 1.6184109449386597 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3226277828216553, "epoch": 0.84, "learning_rate": 2.7887323943661974e-05, "loss": 1.0825, "step": 990, "task_loss": 1.6336266994476318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9355644583702087, "epoch": 0.84, "learning_rate": 2.791549295774648e-05, "loss": 1.1211, "step": 991, "task_loss": 1.1465133428573608 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7223010063171387, "epoch": 0.84, "learning_rate": 2.7943661971830985e-05, "loss": 0.7458, "step": 992, "task_loss": 1.2874919176101685 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6090562343597412, "epoch": 0.84, "learning_rate": 2.7971830985915492e-05, "loss": 1.1627, "step": 993, "task_loss": 0.9001317024230957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3117129802703857, "epoch": 0.84, "learning_rate": 2.8000000000000003e-05, "loss": 1.1973, "step": 994, "task_loss": 1.7932319641113281 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7790877819061279, "epoch": 0.84, "learning_rate": 2.802816901408451e-05, "loss": 0.914, "step": 995, "task_loss": 0.7152183651924133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7137200832366943, "epoch": 0.84, "learning_rate": 2.8056338028169017e-05, "loss": 0.992, "step": 996, "task_loss": 0.7075108885765076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0675511360168457, "epoch": 0.84, "learning_rate": 2.808450704225352e-05, "loss": 1.0371, "step": 997, "task_loss": 1.4038482904434204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0240932703018188, "epoch": 0.84, "learning_rate": 2.811267605633803e-05, "loss": 1.0094, "step": 998, "task_loss": 1.630947232246399 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6119319200515747, "epoch": 0.84, "learning_rate": 2.814084507042254e-05, "loss": 1.1756, "step": 999, "task_loss": 1.3929489850997925 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8482211232185364, "epoch": 0.84, "learning_rate": 2.8169014084507046e-05, "loss": 0.8429, "step": 1000, "task_loss": 0.8347824811935425 }, { "epoch": 0.84, "eval_accuracy": 0.8677623762376238, "eval_loss": 0.6449630856513977, "eval_runtime": 207.4512, "eval_samples_per_second": 121.715, "eval_steps_per_second": 0.954, "step": 1000 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7593427300453186, "epoch": 0.85, "learning_rate": 2.819718309859155e-05, "loss": 0.9513, "step": 1001, "task_loss": 0.4575325548648834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49188166856765747, "epoch": 0.85, "learning_rate": 2.8225352112676058e-05, "loss": 0.8658, "step": 1002, "task_loss": 0.5453650951385498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8746909499168396, "epoch": 0.85, "learning_rate": 2.825352112676056e-05, "loss": 0.8855, "step": 1003, "task_loss": 0.9244720935821533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0637762546539307, "epoch": 0.85, "learning_rate": 2.8281690140845075e-05, "loss": 1.0707, "step": 1004, "task_loss": 0.7581093311309814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.020636558532715, "epoch": 0.85, "learning_rate": 2.830985915492958e-05, "loss": 1.2159, "step": 1005, "task_loss": 2.534658193588257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.202458381652832, "epoch": 0.85, "learning_rate": 2.8338028169014087e-05, "loss": 0.9624, "step": 1006, "task_loss": 0.2074671983718872 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0573878288269043, "epoch": 0.85, "learning_rate": 2.836619718309859e-05, "loss": 0.9975, "step": 1007, "task_loss": 0.7101489901542664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2075992822647095, "epoch": 0.85, "learning_rate": 2.8394366197183098e-05, "loss": 1.0933, "step": 1008, "task_loss": 1.172258973121643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6240395307540894, "epoch": 0.85, "learning_rate": 2.842253521126761e-05, "loss": 1.2332, "step": 1009, "task_loss": 1.7856272459030151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9935423731803894, "epoch": 0.85, "learning_rate": 2.8450704225352116e-05, "loss": 1.0819, "step": 1010, "task_loss": 2.0076398849487305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2395308017730713, "epoch": 0.85, "learning_rate": 2.847887323943662e-05, "loss": 1.2706, "step": 1011, "task_loss": 1.4533475637435913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0621111392974854, "epoch": 0.85, "learning_rate": 2.8507042253521127e-05, "loss": 1.1096, "step": 1012, "task_loss": 0.4310009777545929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0929715633392334, "epoch": 0.86, "learning_rate": 2.8535211267605634e-05, "loss": 0.835, "step": 1013, "task_loss": 0.4668806493282318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2925875186920166, "epoch": 0.86, "learning_rate": 2.8563380281690145e-05, "loss": 1.0327, "step": 1014, "task_loss": 0.9138253927230835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8743935227394104, "epoch": 0.86, "learning_rate": 2.859154929577465e-05, "loss": 0.9421, "step": 1015, "task_loss": 1.1862988471984863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9086697101593018, "epoch": 0.86, "learning_rate": 2.8619718309859156e-05, "loss": 1.0868, "step": 1016, "task_loss": 1.0382267236709595 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.911444365978241, "epoch": 0.86, "learning_rate": 2.8647887323943663e-05, "loss": 1.078, "step": 1017, "task_loss": 0.49846822023391724 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8153744339942932, "epoch": 0.86, "learning_rate": 2.8676056338028167e-05, "loss": 0.9259, "step": 1018, "task_loss": 1.1252528429031372 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.631105661392212, "epoch": 0.86, "learning_rate": 2.870422535211268e-05, "loss": 1.2694, "step": 1019, "task_loss": 1.381534457206726 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.004510521888733, "epoch": 0.86, "learning_rate": 2.8732394366197185e-05, "loss": 0.803, "step": 1020, "task_loss": 1.3844407796859741 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8453183174133301, "epoch": 0.86, "learning_rate": 2.8760563380281692e-05, "loss": 1.3293, "step": 1021, "task_loss": 0.15933802723884583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6557779312133789, "epoch": 0.86, "learning_rate": 2.8788732394366196e-05, "loss": 0.9872, "step": 1022, "task_loss": 0.1295677125453949 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9116742014884949, "epoch": 0.86, "learning_rate": 2.881690140845071e-05, "loss": 0.851, "step": 1023, "task_loss": 1.0979938507080078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.042966365814209, "epoch": 0.87, "learning_rate": 2.8845070422535214e-05, "loss": 1.0934, "step": 1024, "task_loss": 0.2566753327846527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9466080069541931, "epoch": 0.87, "learning_rate": 2.887323943661972e-05, "loss": 0.8985, "step": 1025, "task_loss": 0.8659387826919556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6263465881347656, "epoch": 0.87, "learning_rate": 2.8901408450704225e-05, "loss": 1.0743, "step": 1026, "task_loss": 0.05690251290798187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9204387068748474, "epoch": 0.87, "learning_rate": 2.8929577464788732e-05, "loss": 1.0821, "step": 1027, "task_loss": 0.5582957863807678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0032994747161865, "epoch": 0.87, "learning_rate": 2.8957746478873243e-05, "loss": 1.254, "step": 1028, "task_loss": 1.248001217842102 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9106415510177612, "epoch": 0.87, "learning_rate": 2.898591549295775e-05, "loss": 0.9998, "step": 1029, "task_loss": 0.993424117565155 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3746155500411987, "epoch": 0.87, "learning_rate": 2.9014084507042254e-05, "loss": 1.1408, "step": 1030, "task_loss": 1.5024198293685913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6858294010162354, "epoch": 0.87, "learning_rate": 2.904225352112676e-05, "loss": 1.2943, "step": 1031, "task_loss": 0.9169073104858398 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3341131210327148, "epoch": 0.87, "learning_rate": 2.9070422535211268e-05, "loss": 1.0637, "step": 1032, "task_loss": 0.44080108404159546 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9551115036010742, "epoch": 0.87, "learning_rate": 2.909859154929578e-05, "loss": 1.1499, "step": 1033, "task_loss": 0.5235313773155212 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3333308696746826, "epoch": 0.87, "learning_rate": 2.9126760563380283e-05, "loss": 1.2162, "step": 1034, "task_loss": 1.71952486038208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5737409591674805, "epoch": 0.87, "learning_rate": 2.915492957746479e-05, "loss": 0.8572, "step": 1035, "task_loss": 1.019910216331482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6717537641525269, "epoch": 0.88, "learning_rate": 2.9183098591549297e-05, "loss": 1.3113, "step": 1036, "task_loss": 1.209499716758728 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7280973196029663, "epoch": 0.88, "learning_rate": 2.92112676056338e-05, "loss": 0.9859, "step": 1037, "task_loss": 1.6349462270736694 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0677001476287842, "epoch": 0.88, "learning_rate": 2.9239436619718312e-05, "loss": 1.063, "step": 1038, "task_loss": 0.8625321984291077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9991448521614075, "epoch": 0.88, "learning_rate": 2.926760563380282e-05, "loss": 0.9983, "step": 1039, "task_loss": 0.5490456223487854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1246428489685059, "epoch": 0.88, "learning_rate": 2.9295774647887326e-05, "loss": 1.1354, "step": 1040, "task_loss": 0.7833432555198669 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1602286100387573, "epoch": 0.88, "learning_rate": 2.932394366197183e-05, "loss": 1.4413, "step": 1041, "task_loss": 0.9289639592170715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9450091123580933, "epoch": 0.88, "learning_rate": 2.9352112676056337e-05, "loss": 0.8627, "step": 1042, "task_loss": 1.438068151473999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8193690180778503, "epoch": 0.88, "learning_rate": 2.9380281690140848e-05, "loss": 0.9869, "step": 1043, "task_loss": 0.5756095051765442 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6654974222183228, "epoch": 0.88, "learning_rate": 2.9408450704225355e-05, "loss": 1.0044, "step": 1044, "task_loss": 0.5758503675460815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6927673816680908, "epoch": 0.88, "learning_rate": 2.943661971830986e-05, "loss": 1.2175, "step": 1045, "task_loss": 0.7513061761856079 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.785949468612671, "epoch": 0.88, "learning_rate": 2.9464788732394366e-05, "loss": 1.2773, "step": 1046, "task_loss": 1.4937093257904053 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7818616628646851, "epoch": 0.88, "learning_rate": 2.9492957746478874e-05, "loss": 1.2269, "step": 1047, "task_loss": 0.361010879278183 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7914441823959351, "epoch": 0.89, "learning_rate": 2.9521126760563384e-05, "loss": 0.9013, "step": 1048, "task_loss": 0.6949283480644226 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8369139432907104, "epoch": 0.89, "learning_rate": 2.9549295774647888e-05, "loss": 1.1611, "step": 1049, "task_loss": 1.055216908454895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8113487958908081, "epoch": 0.89, "learning_rate": 2.9577464788732395e-05, "loss": 1.1456, "step": 1050, "task_loss": 0.7723767757415771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.665715217590332, "epoch": 0.89, "learning_rate": 2.9605633802816903e-05, "loss": 0.9223, "step": 1051, "task_loss": 0.6353974342346191 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.204966425895691, "epoch": 0.89, "learning_rate": 2.9633802816901413e-05, "loss": 1.1879, "step": 1052, "task_loss": 0.645239531993866 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4019713401794434, "epoch": 0.89, "learning_rate": 2.9661971830985917e-05, "loss": 1.0436, "step": 1053, "task_loss": 1.1748777627944946 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1226509809494019, "epoch": 0.89, "learning_rate": 2.9690140845070424e-05, "loss": 1.1557, "step": 1054, "task_loss": 1.283124566078186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6956405639648438, "epoch": 0.89, "learning_rate": 2.971830985915493e-05, "loss": 1.0039, "step": 1055, "task_loss": 0.36284139752388 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4302655458450317, "epoch": 0.89, "learning_rate": 2.9746478873239436e-05, "loss": 1.1894, "step": 1056, "task_loss": 1.2908812761306763 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.957003116607666, "epoch": 0.89, "learning_rate": 2.9774647887323946e-05, "loss": 1.362, "step": 1057, "task_loss": 1.5658016204833984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2597755193710327, "epoch": 0.89, "learning_rate": 2.9802816901408453e-05, "loss": 1.04, "step": 1058, "task_loss": 1.8159717321395874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5611103177070618, "epoch": 0.89, "learning_rate": 2.983098591549296e-05, "loss": 0.9077, "step": 1059, "task_loss": 0.3236370086669922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2771944999694824, "epoch": 0.9, "learning_rate": 2.9859154929577465e-05, "loss": 0.98, "step": 1060, "task_loss": 1.5605072975158691 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9398289918899536, "epoch": 0.9, "learning_rate": 2.9887323943661972e-05, "loss": 0.9912, "step": 1061, "task_loss": 0.6549097299575806 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0897243022918701, "epoch": 0.9, "learning_rate": 2.9915492957746482e-05, "loss": 1.0381, "step": 1062, "task_loss": 0.9691208004951477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7724114656448364, "epoch": 0.9, "learning_rate": 2.994366197183099e-05, "loss": 0.8311, "step": 1063, "task_loss": 0.6471152901649475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 2.101308822631836, "epoch": 0.9, "learning_rate": 2.9971830985915494e-05, "loss": 1.2605, "step": 1064, "task_loss": 2.2569496631622314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.406369686126709, "epoch": 0.9, "learning_rate": 3e-05, "loss": 1.0822, "step": 1065, "task_loss": 0.9523471593856812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.537786602973938, "epoch": 0.9, "learning_rate": 3.0028169014084505e-05, "loss": 1.2433, "step": 1066, "task_loss": 0.7768912315368652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6844545602798462, "epoch": 0.9, "learning_rate": 3.005633802816902e-05, "loss": 0.9955, "step": 1067, "task_loss": 0.5066354274749756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.145485281944275, "epoch": 0.9, "learning_rate": 3.0084507042253523e-05, "loss": 0.8444, "step": 1068, "task_loss": 1.4101670980453491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5063529014587402, "epoch": 0.9, "learning_rate": 3.011267605633803e-05, "loss": 0.7031, "step": 1069, "task_loss": 0.9908825755119324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0153183937072754, "epoch": 0.9, "learning_rate": 3.0140845070422537e-05, "loss": 1.0138, "step": 1070, "task_loss": 0.9253076910972595 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9802341461181641, "epoch": 0.9, "learning_rate": 3.016901408450704e-05, "loss": 0.9153, "step": 1071, "task_loss": 0.9065817594528198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.978416919708252, "epoch": 0.91, "learning_rate": 3.019718309859155e-05, "loss": 0.9961, "step": 1072, "task_loss": 1.2014515399932861 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8035662174224854, "epoch": 0.91, "learning_rate": 3.022535211267606e-05, "loss": 0.8993, "step": 1073, "task_loss": 0.6417739391326904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3236968517303467, "epoch": 0.91, "learning_rate": 3.0253521126760566e-05, "loss": 0.8114, "step": 1074, "task_loss": 1.0527417659759521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.07509183883667, "epoch": 0.91, "learning_rate": 3.028169014084507e-05, "loss": 1.0265, "step": 1075, "task_loss": 0.5519587397575378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8223657608032227, "epoch": 0.91, "learning_rate": 3.0309859154929577e-05, "loss": 1.1736, "step": 1076, "task_loss": 0.5959159731864929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1561861038208008, "epoch": 0.91, "learning_rate": 3.0338028169014088e-05, "loss": 1.0308, "step": 1077, "task_loss": 1.2010157108306885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7439294457435608, "epoch": 0.91, "learning_rate": 3.0366197183098595e-05, "loss": 0.823, "step": 1078, "task_loss": 0.8983598947525024 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8068455457687378, "epoch": 0.91, "learning_rate": 3.03943661971831e-05, "loss": 0.9542, "step": 1079, "task_loss": 0.7052931785583496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8812681436538696, "epoch": 0.91, "learning_rate": 3.0422535211267606e-05, "loss": 0.975, "step": 1080, "task_loss": 0.9838396906852722 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1792051792144775, "epoch": 0.91, "learning_rate": 3.0450704225352117e-05, "loss": 1.1423, "step": 1081, "task_loss": 1.2914183139801025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.252800464630127, "epoch": 0.91, "learning_rate": 3.0478873239436624e-05, "loss": 1.129, "step": 1082, "task_loss": 1.3391733169555664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5945141315460205, "epoch": 0.91, "learning_rate": 3.0507042253521128e-05, "loss": 0.8742, "step": 1083, "task_loss": 0.3900918662548065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9994857311248779, "epoch": 0.92, "learning_rate": 3.0535211267605635e-05, "loss": 1.1446, "step": 1084, "task_loss": 1.0022528171539307 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7893470525741577, "epoch": 0.92, "learning_rate": 3.056338028169014e-05, "loss": 1.0212, "step": 1085, "task_loss": 0.8449904322624207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9564645290374756, "epoch": 0.92, "learning_rate": 3.059154929577465e-05, "loss": 0.919, "step": 1086, "task_loss": 1.0955801010131836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.306018590927124, "epoch": 0.92, "learning_rate": 3.061971830985916e-05, "loss": 0.9375, "step": 1087, "task_loss": 1.036637783050537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0175807476043701, "epoch": 0.92, "learning_rate": 3.0647887323943664e-05, "loss": 1.0052, "step": 1088, "task_loss": 0.572571337223053 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8331770896911621, "epoch": 0.92, "learning_rate": 3.067605633802817e-05, "loss": 0.8684, "step": 1089, "task_loss": 0.695405125617981 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3837506771087646, "epoch": 0.92, "learning_rate": 3.070422535211268e-05, "loss": 0.8805, "step": 1090, "task_loss": 1.0553563833236694 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7742348909378052, "epoch": 0.92, "learning_rate": 3.0732394366197186e-05, "loss": 1.0389, "step": 1091, "task_loss": 1.9606428146362305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.024019479751587, "epoch": 0.92, "learning_rate": 3.076056338028169e-05, "loss": 0.9524, "step": 1092, "task_loss": 0.9821241497993469 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4914461374282837, "epoch": 0.92, "learning_rate": 3.07887323943662e-05, "loss": 1.1549, "step": 1093, "task_loss": 1.3320156335830688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.344794750213623, "epoch": 0.92, "learning_rate": 3.081690140845071e-05, "loss": 1.0364, "step": 1094, "task_loss": 0.6141003966331482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9001776576042175, "epoch": 0.93, "learning_rate": 3.084507042253521e-05, "loss": 0.7858, "step": 1095, "task_loss": 1.1340445280075073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.884434700012207, "epoch": 0.93, "learning_rate": 3.087323943661972e-05, "loss": 0.8789, "step": 1096, "task_loss": 0.3262295126914978 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0878591537475586, "epoch": 0.93, "learning_rate": 3.090140845070423e-05, "loss": 0.9974, "step": 1097, "task_loss": 1.0964699983596802 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1420729160308838, "epoch": 0.93, "learning_rate": 3.092957746478874e-05, "loss": 0.8832, "step": 1098, "task_loss": 1.0143002271652222 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1730316877365112, "epoch": 0.93, "learning_rate": 3.095774647887324e-05, "loss": 0.9229, "step": 1099, "task_loss": 1.746788740158081 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9279778599739075, "epoch": 0.93, "learning_rate": 3.0985915492957744e-05, "loss": 1.0695, "step": 1100, "task_loss": 0.5339331030845642 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.910100519657135, "epoch": 0.93, "learning_rate": 3.101408450704226e-05, "loss": 1.0544, "step": 1101, "task_loss": 0.3926548659801483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3873209953308105, "epoch": 0.93, "learning_rate": 3.1042253521126766e-05, "loss": 1.0664, "step": 1102, "task_loss": 1.0662553310394287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5623382329940796, "epoch": 0.93, "learning_rate": 3.1070422535211266e-05, "loss": 0.7573, "step": 1103, "task_loss": 0.1082405224442482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3028755187988281, "epoch": 0.93, "learning_rate": 3.109859154929577e-05, "loss": 1.0647, "step": 1104, "task_loss": 1.4407647848129272 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1238313913345337, "epoch": 0.93, "learning_rate": 3.112676056338028e-05, "loss": 1.0754, "step": 1105, "task_loss": 1.7303612232208252 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3905248641967773, "epoch": 0.93, "learning_rate": 3.1154929577464795e-05, "loss": 1.0738, "step": 1106, "task_loss": 1.236681342124939 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9548593759536743, "epoch": 0.94, "learning_rate": 3.1183098591549295e-05, "loss": 1.0944, "step": 1107, "task_loss": 0.5106108784675598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2685399055480957, "epoch": 0.94, "learning_rate": 3.12112676056338e-05, "loss": 0.9959, "step": 1108, "task_loss": 1.2585952281951904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7153716087341309, "epoch": 0.94, "learning_rate": 3.123943661971831e-05, "loss": 0.9119, "step": 1109, "task_loss": 0.6129963994026184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8258810043334961, "epoch": 0.94, "learning_rate": 3.1267605633802824e-05, "loss": 0.7206, "step": 1110, "task_loss": 0.7236630916595459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3312703371047974, "epoch": 0.94, "learning_rate": 3.1295774647887324e-05, "loss": 1.1497, "step": 1111, "task_loss": 0.7409040927886963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.823762059211731, "epoch": 0.94, "learning_rate": 3.132394366197183e-05, "loss": 0.9306, "step": 1112, "task_loss": 1.0455822944641113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9825214743614197, "epoch": 0.94, "learning_rate": 3.135211267605634e-05, "loss": 0.9138, "step": 1113, "task_loss": 0.5640580058097839 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9581623077392578, "epoch": 0.94, "learning_rate": 3.1380281690140846e-05, "loss": 1.1692, "step": 1114, "task_loss": 0.7524353265762329 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5234683752059937, "epoch": 0.94, "learning_rate": 3.140845070422535e-05, "loss": 0.7571, "step": 1115, "task_loss": 0.11696486920118332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5853378772735596, "epoch": 0.94, "learning_rate": 3.143661971830986e-05, "loss": 0.6705, "step": 1116, "task_loss": 0.640122652053833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.18428635597229, "epoch": 0.94, "learning_rate": 3.146478873239437e-05, "loss": 1.0405, "step": 1117, "task_loss": 0.6611528992652893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0418531894683838, "epoch": 0.94, "learning_rate": 3.1492957746478875e-05, "loss": 0.7887, "step": 1118, "task_loss": 0.7625175714492798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8644493818283081, "epoch": 0.95, "learning_rate": 3.152112676056338e-05, "loss": 0.9952, "step": 1119, "task_loss": 0.4301513731479645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.648482084274292, "epoch": 0.95, "learning_rate": 3.154929577464789e-05, "loss": 0.7437, "step": 1120, "task_loss": 0.30940693616867065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.437239170074463, "epoch": 0.95, "learning_rate": 3.1577464788732397e-05, "loss": 0.9665, "step": 1121, "task_loss": 1.1366058588027954 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7556180953979492, "epoch": 0.95, "learning_rate": 3.1605633802816904e-05, "loss": 0.9736, "step": 1122, "task_loss": 0.2390281856060028 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2528012990951538, "epoch": 0.95, "learning_rate": 3.163380281690141e-05, "loss": 0.7878, "step": 1123, "task_loss": 0.8233896493911743 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0268454551696777, "epoch": 0.95, "learning_rate": 3.166197183098591e-05, "loss": 0.8988, "step": 1124, "task_loss": 0.8265514373779297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.048393964767456, "epoch": 0.95, "learning_rate": 3.1690140845070426e-05, "loss": 0.8121, "step": 1125, "task_loss": 0.9901604056358337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9509321451187134, "epoch": 0.95, "learning_rate": 3.171830985915493e-05, "loss": 0.9908, "step": 1126, "task_loss": 1.1167559623718262 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9042361378669739, "epoch": 0.95, "learning_rate": 3.174647887323944e-05, "loss": 0.8663, "step": 1127, "task_loss": 1.4829646348953247 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.884629487991333, "epoch": 0.95, "learning_rate": 3.177464788732394e-05, "loss": 0.9981, "step": 1128, "task_loss": 1.0387400388717651 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1022435426712036, "epoch": 0.95, "learning_rate": 3.180281690140845e-05, "loss": 0.8365, "step": 1129, "task_loss": 1.5091971158981323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9959489107131958, "epoch": 0.95, "learning_rate": 3.183098591549296e-05, "loss": 0.8935, "step": 1130, "task_loss": 1.772108793258667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.42530715465545654, "epoch": 0.96, "learning_rate": 3.185915492957747e-05, "loss": 0.6951, "step": 1131, "task_loss": 0.23121410608291626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9370519518852234, "epoch": 0.96, "learning_rate": 3.1887323943661976e-05, "loss": 1.0142, "step": 1132, "task_loss": 0.4105663597583771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5788878202438354, "epoch": 0.96, "learning_rate": 3.191549295774648e-05, "loss": 0.892, "step": 1133, "task_loss": 1.008068323135376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8654824495315552, "epoch": 0.96, "learning_rate": 3.1943661971830984e-05, "loss": 0.9695, "step": 1134, "task_loss": 0.7124269008636475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4408470392227173, "epoch": 0.96, "learning_rate": 3.19718309859155e-05, "loss": 1.1518, "step": 1135, "task_loss": 1.7371697425842285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9579079151153564, "epoch": 0.96, "learning_rate": 3.2000000000000005e-05, "loss": 0.9182, "step": 1136, "task_loss": 1.2651770114898682 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.836429238319397, "epoch": 0.96, "learning_rate": 3.2028169014084506e-05, "loss": 1.0088, "step": 1137, "task_loss": 0.8547056913375854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1275967359542847, "epoch": 0.96, "learning_rate": 3.205633802816901e-05, "loss": 0.8439, "step": 1138, "task_loss": 0.6997576355934143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6634745597839355, "epoch": 0.96, "learning_rate": 3.208450704225353e-05, "loss": 1.1415, "step": 1139, "task_loss": 1.0569509267807007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5985784530639648, "epoch": 0.96, "learning_rate": 3.2112676056338034e-05, "loss": 1.089, "step": 1140, "task_loss": 0.6864728927612305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.367414951324463, "epoch": 0.96, "learning_rate": 3.2140845070422535e-05, "loss": 1.0582, "step": 1141, "task_loss": 1.4149954319000244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8596222400665283, "epoch": 0.96, "learning_rate": 3.216901408450704e-05, "loss": 0.9804, "step": 1142, "task_loss": 1.4745179414749146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6966142058372498, "epoch": 0.97, "learning_rate": 3.219718309859155e-05, "loss": 0.9161, "step": 1143, "task_loss": 0.5558992028236389 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2815407514572144, "epoch": 0.97, "learning_rate": 3.222535211267606e-05, "loss": 0.8613, "step": 1144, "task_loss": 1.2148690223693848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.602104663848877, "epoch": 0.97, "learning_rate": 3.2253521126760564e-05, "loss": 0.7798, "step": 1145, "task_loss": 0.31097087264060974 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8310656547546387, "epoch": 0.97, "learning_rate": 3.228169014084507e-05, "loss": 1.1022, "step": 1146, "task_loss": 1.8354239463806152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6525941491127014, "epoch": 0.97, "learning_rate": 3.230985915492958e-05, "loss": 0.8232, "step": 1147, "task_loss": 0.32751402258872986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8042995929718018, "epoch": 0.97, "learning_rate": 3.2338028169014086e-05, "loss": 1.0499, "step": 1148, "task_loss": 0.7061102390289307 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0960584878921509, "epoch": 0.97, "learning_rate": 3.236619718309859e-05, "loss": 0.9368, "step": 1149, "task_loss": 1.6864874362945557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1719467639923096, "epoch": 0.97, "learning_rate": 3.23943661971831e-05, "loss": 0.8746, "step": 1150, "task_loss": 1.4344993829727173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.796451210975647, "epoch": 0.97, "learning_rate": 3.242253521126761e-05, "loss": 0.7614, "step": 1151, "task_loss": 1.12985360622406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.139125108718872, "epoch": 0.97, "learning_rate": 3.2450704225352115e-05, "loss": 0.874, "step": 1152, "task_loss": 0.6622596979141235 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.674725353717804, "epoch": 0.97, "learning_rate": 3.247887323943662e-05, "loss": 0.9157, "step": 1153, "task_loss": 0.6688202619552612 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7879860401153564, "epoch": 0.97, "learning_rate": 3.250704225352113e-05, "loss": 0.96, "step": 1154, "task_loss": 0.6181442737579346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.820242166519165, "epoch": 0.98, "learning_rate": 3.2535211267605636e-05, "loss": 0.8521, "step": 1155, "task_loss": 1.5479239225387573 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5802704095840454, "epoch": 0.98, "learning_rate": 3.2563380281690144e-05, "loss": 0.789, "step": 1156, "task_loss": 0.37033742666244507 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6025118827819824, "epoch": 0.98, "learning_rate": 3.259154929577465e-05, "loss": 0.7076, "step": 1157, "task_loss": 0.669731616973877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8194455504417419, "epoch": 0.98, "learning_rate": 3.261971830985915e-05, "loss": 0.8474, "step": 1158, "task_loss": 0.482734739780426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9048824906349182, "epoch": 0.98, "learning_rate": 3.2647887323943665e-05, "loss": 0.9553, "step": 1159, "task_loss": 0.6738177537918091 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7274172306060791, "epoch": 0.98, "learning_rate": 3.267605633802817e-05, "loss": 0.8595, "step": 1160, "task_loss": 1.0057735443115234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8213383555412292, "epoch": 0.98, "learning_rate": 3.270422535211268e-05, "loss": 0.824, "step": 1161, "task_loss": 0.577415406703949 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2786495685577393, "epoch": 0.98, "learning_rate": 3.273239436619718e-05, "loss": 1.0391, "step": 1162, "task_loss": 1.044956088066101 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.160908579826355, "epoch": 0.98, "learning_rate": 3.276056338028169e-05, "loss": 1.1073, "step": 1163, "task_loss": 1.2352685928344727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8824045658111572, "epoch": 0.98, "learning_rate": 3.27887323943662e-05, "loss": 0.8744, "step": 1164, "task_loss": 1.1704274415969849 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7752103209495544, "epoch": 0.98, "learning_rate": 3.281690140845071e-05, "loss": 0.7299, "step": 1165, "task_loss": 0.6679526567459106 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.196090579032898, "epoch": 0.99, "learning_rate": 3.284507042253521e-05, "loss": 0.9072, "step": 1166, "task_loss": 0.7240828275680542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7776911854743958, "epoch": 0.99, "learning_rate": 3.2873239436619717e-05, "loss": 0.8524, "step": 1167, "task_loss": 0.49345192313194275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9924798011779785, "epoch": 0.99, "learning_rate": 3.2901408450704224e-05, "loss": 0.9036, "step": 1168, "task_loss": 1.4417037963867188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4229249954223633, "epoch": 0.99, "learning_rate": 3.292957746478874e-05, "loss": 1.0278, "step": 1169, "task_loss": 0.9556698799133301 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3738927841186523, "epoch": 0.99, "learning_rate": 3.295774647887324e-05, "loss": 0.7143, "step": 1170, "task_loss": 0.3922346532344818 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7932173013687134, "epoch": 0.99, "learning_rate": 3.2985915492957746e-05, "loss": 0.7577, "step": 1171, "task_loss": 0.18679168820381165 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4612720012664795, "epoch": 0.99, "learning_rate": 3.301408450704225e-05, "loss": 0.9204, "step": 1172, "task_loss": 1.474944829940796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.51720130443573, "epoch": 0.99, "learning_rate": 3.304225352112677e-05, "loss": 0.9028, "step": 1173, "task_loss": 0.18524540960788727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1258552074432373, "epoch": 0.99, "learning_rate": 3.307042253521127e-05, "loss": 1.0582, "step": 1174, "task_loss": 1.2323304414749146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4558299779891968, "epoch": 0.99, "learning_rate": 3.3098591549295775e-05, "loss": 1.2906, "step": 1175, "task_loss": 1.439950942993164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1983811855316162, "epoch": 0.99, "learning_rate": 3.312676056338028e-05, "loss": 0.9156, "step": 1176, "task_loss": 0.5439980626106262 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7068243622779846, "epoch": 0.99, "learning_rate": 3.315492957746479e-05, "loss": 0.6989, "step": 1177, "task_loss": 0.9930229783058167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.673795461654663, "epoch": 1.0, "learning_rate": 3.31830985915493e-05, "loss": 1.2927, "step": 1178, "task_loss": 1.4402387142181396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6024789214134216, "epoch": 1.0, "learning_rate": 3.3211267605633804e-05, "loss": 0.849, "step": 1179, "task_loss": 1.2530150413513184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7807836532592773, "epoch": 1.0, "learning_rate": 3.323943661971831e-05, "loss": 0.7354, "step": 1180, "task_loss": 0.610946774482727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8082504868507385, "epoch": 1.0, "learning_rate": 3.326760563380282e-05, "loss": 0.9797, "step": 1181, "task_loss": 1.8083199262619019 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1054185628890991, "epoch": 1.0, "learning_rate": 3.3295774647887325e-05, "loss": 0.9461, "step": 1182, "task_loss": 0.5020819306373596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9442867040634155, "epoch": 1.0, "learning_rate": 3.332394366197183e-05, "loss": 0.906, "step": 1183, "task_loss": 0.7626156806945801 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2918623685836792, "epoch": 1.0, "learning_rate": 3.335211267605634e-05, "loss": 1.534, "step": 1184, "task_loss": 0.4270663857460022 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8555040955543518, "epoch": 1.0, "learning_rate": 3.338028169014085e-05, "loss": 0.9071, "step": 1185, "task_loss": 0.5815106630325317 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8014392256736755, "epoch": 1.0, "learning_rate": 3.3408450704225354e-05, "loss": 0.7853, "step": 1186, "task_loss": 2.051740884780884 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9240382313728333, "epoch": 1.0, "learning_rate": 3.343661971830986e-05, "loss": 0.939, "step": 1187, "task_loss": 1.082109808921814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8004231452941895, "epoch": 1.0, "learning_rate": 3.346478873239437e-05, "loss": 0.9881, "step": 1188, "task_loss": 1.063913345336914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1934356689453125, "epoch": 1.01, "learning_rate": 3.3492957746478876e-05, "loss": 0.8043, "step": 1189, "task_loss": 0.8237214088439941 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9928021430969238, "epoch": 1.01, "learning_rate": 3.352112676056338e-05, "loss": 0.8008, "step": 1190, "task_loss": 0.4950225353240967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0945161581039429, "epoch": 1.01, "learning_rate": 3.354929577464789e-05, "loss": 0.9191, "step": 1191, "task_loss": 1.2156010866165161 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7962568998336792, "epoch": 1.01, "learning_rate": 3.357746478873239e-05, "loss": 0.7123, "step": 1192, "task_loss": 0.6859135627746582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8586177229881287, "epoch": 1.01, "learning_rate": 3.3605633802816905e-05, "loss": 0.7631, "step": 1193, "task_loss": 0.5781033635139465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9123694896697998, "epoch": 1.01, "learning_rate": 3.363380281690141e-05, "loss": 1.0254, "step": 1194, "task_loss": 1.0316059589385986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6609846353530884, "epoch": 1.01, "learning_rate": 3.366197183098592e-05, "loss": 0.8448, "step": 1195, "task_loss": 0.7049547433853149 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.805208146572113, "epoch": 1.01, "learning_rate": 3.369014084507042e-05, "loss": 0.8183, "step": 1196, "task_loss": 0.15728560090065002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4085863828659058, "epoch": 1.01, "learning_rate": 3.371830985915493e-05, "loss": 0.8684, "step": 1197, "task_loss": 1.0215092897415161 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5439054369926453, "epoch": 1.01, "learning_rate": 3.374647887323944e-05, "loss": 0.6728, "step": 1198, "task_loss": 0.8839086890220642 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5644635558128357, "epoch": 1.01, "learning_rate": 3.377464788732395e-05, "loss": 0.5193, "step": 1199, "task_loss": 0.5074184536933899 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.3341379165649414, "epoch": 1.01, "learning_rate": 3.380281690140845e-05, "loss": 0.9964, "step": 1200, "task_loss": 0.659958004951477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8198841214179993, "epoch": 1.02, "learning_rate": 3.3830985915492956e-05, "loss": 0.8481, "step": 1201, "task_loss": 0.4957406222820282 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7336146831512451, "epoch": 1.02, "learning_rate": 3.385915492957747e-05, "loss": 0.8227, "step": 1202, "task_loss": 0.5420545935630798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6994868516921997, "epoch": 1.02, "learning_rate": 3.388732394366198e-05, "loss": 0.7848, "step": 1203, "task_loss": 0.6088904142379761 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8471839427947998, "epoch": 1.02, "learning_rate": 3.391549295774648e-05, "loss": 0.822, "step": 1204, "task_loss": 1.008898138999939 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1837654113769531, "epoch": 1.02, "learning_rate": 3.3943661971830985e-05, "loss": 0.8906, "step": 1205, "task_loss": 1.1946046352386475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.6041507720947266, "epoch": 1.02, "learning_rate": 3.397183098591549e-05, "loss": 0.9566, "step": 1206, "task_loss": 1.425720453262329 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8012924194335938, "epoch": 1.02, "learning_rate": 3.4000000000000007e-05, "loss": 1.0368, "step": 1207, "task_loss": 0.8884502053260803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7972298860549927, "epoch": 1.02, "learning_rate": 3.402816901408451e-05, "loss": 0.9405, "step": 1208, "task_loss": 1.327467679977417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7542397975921631, "epoch": 1.02, "learning_rate": 3.4056338028169014e-05, "loss": 0.687, "step": 1209, "task_loss": 1.0868712663650513 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9849074482917786, "epoch": 1.02, "learning_rate": 3.408450704225352e-05, "loss": 0.9183, "step": 1210, "task_loss": 1.5958986282348633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0909024477005005, "epoch": 1.02, "learning_rate": 3.411267605633803e-05, "loss": 0.9547, "step": 1211, "task_loss": 1.6481575965881348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0945838689804077, "epoch": 1.02, "learning_rate": 3.4140845070422536e-05, "loss": 0.9395, "step": 1212, "task_loss": 1.709041953086853 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6183662414550781, "epoch": 1.03, "learning_rate": 3.416901408450704e-05, "loss": 0.6722, "step": 1213, "task_loss": 0.4922850430011749 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7425100803375244, "epoch": 1.03, "learning_rate": 3.419718309859155e-05, "loss": 0.7634, "step": 1214, "task_loss": 0.36757317185401917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9270923733711243, "epoch": 1.03, "learning_rate": 3.422535211267606e-05, "loss": 0.7364, "step": 1215, "task_loss": 1.4397999048233032 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0387612581253052, "epoch": 1.03, "learning_rate": 3.4253521126760565e-05, "loss": 0.8029, "step": 1216, "task_loss": 0.766712486743927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8965452313423157, "epoch": 1.03, "learning_rate": 3.428169014084507e-05, "loss": 0.8633, "step": 1217, "task_loss": 0.4307364821434021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7357679605484009, "epoch": 1.03, "learning_rate": 3.430985915492958e-05, "loss": 0.6925, "step": 1218, "task_loss": 2.608794927597046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8549785614013672, "epoch": 1.03, "learning_rate": 3.433802816901409e-05, "loss": 0.7881, "step": 1219, "task_loss": 0.8181943297386169 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7123785018920898, "epoch": 1.03, "learning_rate": 3.4366197183098594e-05, "loss": 0.8426, "step": 1220, "task_loss": 0.692875325679779 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7644048929214478, "epoch": 1.03, "learning_rate": 3.4394366197183094e-05, "loss": 0.6814, "step": 1221, "task_loss": 1.2170567512512207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7671277523040771, "epoch": 1.03, "learning_rate": 3.442253521126761e-05, "loss": 0.7011, "step": 1222, "task_loss": 0.9596631526947021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5179318189620972, "epoch": 1.03, "learning_rate": 3.4450704225352116e-05, "loss": 0.5611, "step": 1223, "task_loss": 0.12929952144622803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9294768571853638, "epoch": 1.03, "learning_rate": 3.447887323943662e-05, "loss": 0.9369, "step": 1224, "task_loss": 1.5789263248443604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4444693326950073, "epoch": 1.04, "learning_rate": 3.450704225352113e-05, "loss": 0.5975, "step": 1225, "task_loss": 0.42791566252708435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4421626329421997, "epoch": 1.04, "learning_rate": 3.453521126760563e-05, "loss": 0.8113, "step": 1226, "task_loss": 0.32182690501213074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7147836685180664, "epoch": 1.04, "learning_rate": 3.4563380281690145e-05, "loss": 0.8179, "step": 1227, "task_loss": 0.8937325477600098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2031548023223877, "epoch": 1.04, "learning_rate": 3.459154929577465e-05, "loss": 0.8704, "step": 1228, "task_loss": 1.2828608751296997 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6520639657974243, "epoch": 1.04, "learning_rate": 3.461971830985916e-05, "loss": 0.6956, "step": 1229, "task_loss": 0.4400368332862854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8684521317481995, "epoch": 1.04, "learning_rate": 3.464788732394366e-05, "loss": 0.9689, "step": 1230, "task_loss": 0.6189638376235962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7574407458305359, "epoch": 1.04, "learning_rate": 3.4676056338028174e-05, "loss": 0.7939, "step": 1231, "task_loss": 0.4986882507801056 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0259532928466797, "epoch": 1.04, "learning_rate": 3.470422535211268e-05, "loss": 0.6997, "step": 1232, "task_loss": 0.8594432473182678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7895364165306091, "epoch": 1.04, "learning_rate": 3.473239436619719e-05, "loss": 0.7194, "step": 1233, "task_loss": 1.7608203887939453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8813434839248657, "epoch": 1.04, "learning_rate": 3.476056338028169e-05, "loss": 0.869, "step": 1234, "task_loss": 0.8294102549552917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5496876239776611, "epoch": 1.04, "learning_rate": 3.4788732394366196e-05, "loss": 0.7477, "step": 1235, "task_loss": 1.3527982234954834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6474922299385071, "epoch": 1.04, "learning_rate": 3.481690140845071e-05, "loss": 0.69, "step": 1236, "task_loss": 1.1965404748916626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6866661906242371, "epoch": 1.05, "learning_rate": 3.484507042253522e-05, "loss": 0.7543, "step": 1237, "task_loss": 1.1704741716384888 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.7254645824432373, "epoch": 1.05, "learning_rate": 3.487323943661972e-05, "loss": 1.0765, "step": 1238, "task_loss": 1.3405183553695679 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6111165285110474, "epoch": 1.05, "learning_rate": 3.4901408450704225e-05, "loss": 1.095, "step": 1239, "task_loss": 0.17204031348228455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7005902528762817, "epoch": 1.05, "learning_rate": 3.492957746478873e-05, "loss": 0.7734, "step": 1240, "task_loss": 0.4967375099658966 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5905947089195251, "epoch": 1.05, "learning_rate": 3.4957746478873246e-05, "loss": 0.6945, "step": 1241, "task_loss": 0.1094353199005127 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7181411981582642, "epoch": 1.05, "learning_rate": 3.498591549295775e-05, "loss": 0.7184, "step": 1242, "task_loss": 0.6629283428192139 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8046637177467346, "epoch": 1.05, "learning_rate": 3.5014084507042254e-05, "loss": 0.7162, "step": 1243, "task_loss": 1.0194084644317627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6250340938568115, "epoch": 1.05, "learning_rate": 3.504225352112676e-05, "loss": 0.6875, "step": 1244, "task_loss": 0.541187584400177 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8407703042030334, "epoch": 1.05, "learning_rate": 3.507042253521127e-05, "loss": 0.9001, "step": 1245, "task_loss": 1.2834328413009644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5556535124778748, "epoch": 1.05, "learning_rate": 3.5098591549295776e-05, "loss": 0.8061, "step": 1246, "task_loss": 0.8866622447967529 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7095454931259155, "epoch": 1.05, "learning_rate": 3.512676056338028e-05, "loss": 0.8652, "step": 1247, "task_loss": 0.23205532133579254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.831397533416748, "epoch": 1.05, "learning_rate": 3.515492957746479e-05, "loss": 0.8294, "step": 1248, "task_loss": 0.6734315156936646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.040311336517334, "epoch": 1.06, "learning_rate": 3.51830985915493e-05, "loss": 0.7715, "step": 1249, "task_loss": 1.196539282798767 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9563837647438049, "epoch": 1.06, "learning_rate": 3.5211267605633805e-05, "loss": 0.8995, "step": 1250, "task_loss": 1.4706432819366455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6572422385215759, "epoch": 1.06, "learning_rate": 3.523943661971831e-05, "loss": 0.7538, "step": 1251, "task_loss": 0.818042516708374 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7606565952301025, "epoch": 1.06, "learning_rate": 3.526760563380282e-05, "loss": 0.7855, "step": 1252, "task_loss": 1.0046520233154297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6900460124015808, "epoch": 1.06, "learning_rate": 3.5295774647887326e-05, "loss": 0.915, "step": 1253, "task_loss": 1.532382845878601 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6132869720458984, "epoch": 1.06, "learning_rate": 3.5323943661971834e-05, "loss": 0.7636, "step": 1254, "task_loss": 1.2558724880218506 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6906678676605225, "epoch": 1.06, "learning_rate": 3.5352112676056334e-05, "loss": 0.766, "step": 1255, "task_loss": 0.11208156496286392 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7141894102096558, "epoch": 1.06, "learning_rate": 3.538028169014085e-05, "loss": 0.6758, "step": 1256, "task_loss": 0.6671800017356873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6225650310516357, "epoch": 1.06, "learning_rate": 3.5408450704225355e-05, "loss": 0.7616, "step": 1257, "task_loss": 0.31204259395599365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7383618950843811, "epoch": 1.06, "learning_rate": 3.543661971830986e-05, "loss": 0.7773, "step": 1258, "task_loss": 1.4213207960128784 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3776629567146301, "epoch": 1.06, "learning_rate": 3.546478873239436e-05, "loss": 0.5108, "step": 1259, "task_loss": 0.06904461234807968 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7822449207305908, "epoch": 1.07, "learning_rate": 3.549295774647888e-05, "loss": 0.9544, "step": 1260, "task_loss": 0.4792971611022949 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5510671138763428, "epoch": 1.07, "learning_rate": 3.5521126760563384e-05, "loss": 0.6793, "step": 1261, "task_loss": 1.098591923713684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.554914653301239, "epoch": 1.07, "learning_rate": 3.554929577464789e-05, "loss": 0.7546, "step": 1262, "task_loss": 0.8954185843467712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6111783385276794, "epoch": 1.07, "learning_rate": 3.557746478873239e-05, "loss": 0.828, "step": 1263, "task_loss": 0.7794609069824219 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0496890544891357, "epoch": 1.07, "learning_rate": 3.56056338028169e-05, "loss": 1.0405, "step": 1264, "task_loss": 1.9746778011322021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6859597563743591, "epoch": 1.07, "learning_rate": 3.5633802816901413e-05, "loss": 0.7542, "step": 1265, "task_loss": 0.766257107257843 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9770196676254272, "epoch": 1.07, "learning_rate": 3.566197183098592e-05, "loss": 0.842, "step": 1266, "task_loss": 0.9241597652435303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7313793301582336, "epoch": 1.07, "learning_rate": 3.569014084507042e-05, "loss": 0.8746, "step": 1267, "task_loss": 1.913285493850708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9250802993774414, "epoch": 1.07, "learning_rate": 3.571830985915493e-05, "loss": 0.8673, "step": 1268, "task_loss": 0.7990307211875916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5878286361694336, "epoch": 1.07, "learning_rate": 3.5746478873239436e-05, "loss": 0.8124, "step": 1269, "task_loss": 0.42495110630989075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6998294591903687, "epoch": 1.07, "learning_rate": 3.577464788732395e-05, "loss": 0.8133, "step": 1270, "task_loss": 1.1675562858581543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4835773706436157, "epoch": 1.07, "learning_rate": 3.580281690140846e-05, "loss": 1.046, "step": 1271, "task_loss": 0.48136740922927856 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7225725650787354, "epoch": 1.08, "learning_rate": 3.583098591549296e-05, "loss": 0.6938, "step": 1272, "task_loss": 0.9115102291107178 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5379626750946045, "epoch": 1.08, "learning_rate": 3.5859154929577465e-05, "loss": 0.8298, "step": 1273, "task_loss": 0.8082537055015564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6137171387672424, "epoch": 1.08, "learning_rate": 3.588732394366197e-05, "loss": 0.7764, "step": 1274, "task_loss": 1.040844202041626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9455809593200684, "epoch": 1.08, "learning_rate": 3.5915492957746486e-05, "loss": 0.8659, "step": 1275, "task_loss": 1.1844301223754883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6414105892181396, "epoch": 1.08, "learning_rate": 3.5943661971830986e-05, "loss": 0.7991, "step": 1276, "task_loss": 0.5063562989234924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.549480676651001, "epoch": 1.08, "learning_rate": 3.5971830985915494e-05, "loss": 0.8311, "step": 1277, "task_loss": 0.43166977167129517 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8726398348808289, "epoch": 1.08, "learning_rate": 3.6e-05, "loss": 0.8569, "step": 1278, "task_loss": 1.1902964115142822 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.772813081741333, "epoch": 1.08, "learning_rate": 3.602816901408451e-05, "loss": 0.7259, "step": 1279, "task_loss": 0.37119442224502563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0929890871047974, "epoch": 1.08, "learning_rate": 3.6056338028169015e-05, "loss": 0.8003, "step": 1280, "task_loss": 0.9647442102432251 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7997719049453735, "epoch": 1.08, "learning_rate": 3.608450704225352e-05, "loss": 0.8479, "step": 1281, "task_loss": 1.055337905883789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1603193283081055, "epoch": 1.08, "learning_rate": 3.611267605633803e-05, "loss": 0.7489, "step": 1282, "task_loss": 1.6504595279693604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6377028822898865, "epoch": 1.08, "learning_rate": 3.614084507042254e-05, "loss": 0.8583, "step": 1283, "task_loss": 0.9708558917045593 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9410070776939392, "epoch": 1.09, "learning_rate": 3.6169014084507044e-05, "loss": 0.7004, "step": 1284, "task_loss": 0.3442003130912781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6020750999450684, "epoch": 1.09, "learning_rate": 3.619718309859155e-05, "loss": 0.5501, "step": 1285, "task_loss": 0.5716943144798279 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6549991369247437, "epoch": 1.09, "learning_rate": 3.622535211267606e-05, "loss": 0.8281, "step": 1286, "task_loss": 0.5831762552261353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4124181568622589, "epoch": 1.09, "learning_rate": 3.6253521126760566e-05, "loss": 0.783, "step": 1287, "task_loss": 0.06972641497850418 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7036275863647461, "epoch": 1.09, "learning_rate": 3.6281690140845073e-05, "loss": 0.6749, "step": 1288, "task_loss": 0.6497281193733215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5104941129684448, "epoch": 1.09, "learning_rate": 3.630985915492958e-05, "loss": 0.8103, "step": 1289, "task_loss": 0.09877430647611618 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8388316631317139, "epoch": 1.09, "learning_rate": 3.633802816901409e-05, "loss": 0.813, "step": 1290, "task_loss": 0.7490025758743286 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6802901029586792, "epoch": 1.09, "learning_rate": 3.6366197183098595e-05, "loss": 0.8596, "step": 1291, "task_loss": 1.0177067518234253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5577201843261719, "epoch": 1.09, "learning_rate": 3.63943661971831e-05, "loss": 0.7907, "step": 1292, "task_loss": 0.705917239189148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9185020327568054, "epoch": 1.09, "learning_rate": 3.64225352112676e-05, "loss": 0.7579, "step": 1293, "task_loss": 0.714583694934845 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4762297570705414, "epoch": 1.09, "learning_rate": 3.645070422535212e-05, "loss": 0.6115, "step": 1294, "task_loss": 0.509657621383667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.719886302947998, "epoch": 1.09, "learning_rate": 3.6478873239436624e-05, "loss": 0.7005, "step": 1295, "task_loss": 0.6613854169845581 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7538480758666992, "epoch": 1.1, "learning_rate": 3.650704225352113e-05, "loss": 0.6476, "step": 1296, "task_loss": 0.19480088353157043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5861597061157227, "epoch": 1.1, "learning_rate": 3.653521126760563e-05, "loss": 0.531, "step": 1297, "task_loss": 0.26570937037467957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5029035210609436, "epoch": 1.1, "learning_rate": 3.656338028169014e-05, "loss": 0.4622, "step": 1298, "task_loss": 0.5075638294219971 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4162977933883667, "epoch": 1.1, "learning_rate": 3.659154929577465e-05, "loss": 0.9142, "step": 1299, "task_loss": 1.6295454502105713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8458637595176697, "epoch": 1.1, "learning_rate": 3.661971830985916e-05, "loss": 0.7595, "step": 1300, "task_loss": 0.33450421690940857 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.858117938041687, "epoch": 1.1, "learning_rate": 3.664788732394366e-05, "loss": 0.7565, "step": 1301, "task_loss": 1.2290905714035034 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7413697242736816, "epoch": 1.1, "learning_rate": 3.667605633802817e-05, "loss": 0.6924, "step": 1302, "task_loss": 1.1581050157546997 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8725029230117798, "epoch": 1.1, "learning_rate": 3.6704225352112675e-05, "loss": 0.7511, "step": 1303, "task_loss": 0.8451574444770813 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6860213875770569, "epoch": 1.1, "learning_rate": 3.673239436619719e-05, "loss": 0.7371, "step": 1304, "task_loss": 1.367508053779602 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5642271041870117, "epoch": 1.1, "learning_rate": 3.676056338028169e-05, "loss": 0.9491, "step": 1305, "task_loss": 0.5634865164756775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6156296730041504, "epoch": 1.1, "learning_rate": 3.67887323943662e-05, "loss": 0.6687, "step": 1306, "task_loss": 0.8175805807113647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7799545526504517, "epoch": 1.1, "learning_rate": 3.6816901408450704e-05, "loss": 0.7471, "step": 1307, "task_loss": 0.35390737652778625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.060184121131897, "epoch": 1.11, "learning_rate": 3.684507042253521e-05, "loss": 0.9033, "step": 1308, "task_loss": 1.10116708278656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0425946712493896, "epoch": 1.11, "learning_rate": 3.687323943661972e-05, "loss": 0.7784, "step": 1309, "task_loss": 0.9533910155296326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.917415976524353, "epoch": 1.11, "learning_rate": 3.6901408450704226e-05, "loss": 0.7737, "step": 1310, "task_loss": 1.8243242502212524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7757285833358765, "epoch": 1.11, "learning_rate": 3.692957746478873e-05, "loss": 0.7652, "step": 1311, "task_loss": 0.5608454346656799 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7432516813278198, "epoch": 1.11, "learning_rate": 3.695774647887324e-05, "loss": 0.6602, "step": 1312, "task_loss": 0.31056028604507446 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8425228595733643, "epoch": 1.11, "learning_rate": 3.698591549295775e-05, "loss": 0.7002, "step": 1313, "task_loss": 1.7012228965759277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9212037324905396, "epoch": 1.11, "learning_rate": 3.7014084507042255e-05, "loss": 0.7826, "step": 1314, "task_loss": 1.2410392761230469 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5926636457443237, "epoch": 1.11, "learning_rate": 3.704225352112676e-05, "loss": 0.6269, "step": 1315, "task_loss": 0.6078276038169861 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7096782326698303, "epoch": 1.11, "learning_rate": 3.707042253521127e-05, "loss": 0.6782, "step": 1316, "task_loss": 0.7626842856407166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7662361860275269, "epoch": 1.11, "learning_rate": 3.709859154929578e-05, "loss": 0.7875, "step": 1317, "task_loss": 0.7043976187705994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49353694915771484, "epoch": 1.11, "learning_rate": 3.7126760563380284e-05, "loss": 0.6974, "step": 1318, "task_loss": 0.5637612342834473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6509162187576294, "epoch": 1.11, "learning_rate": 3.715492957746479e-05, "loss": 0.6556, "step": 1319, "task_loss": 0.5995513200759888 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5150904655456543, "epoch": 1.12, "learning_rate": 3.71830985915493e-05, "loss": 0.831, "step": 1320, "task_loss": 0.43913036584854126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0928218364715576, "epoch": 1.12, "learning_rate": 3.7211267605633806e-05, "loss": 0.7373, "step": 1321, "task_loss": 1.0779298543930054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.740695595741272, "epoch": 1.12, "learning_rate": 3.723943661971831e-05, "loss": 0.9163, "step": 1322, "task_loss": 1.884770393371582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.740666925907135, "epoch": 1.12, "learning_rate": 3.726760563380282e-05, "loss": 0.7479, "step": 1323, "task_loss": 0.8885055780410767 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8488104343414307, "epoch": 1.12, "learning_rate": 3.729577464788733e-05, "loss": 0.8376, "step": 1324, "task_loss": 0.519652783870697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8062671422958374, "epoch": 1.12, "learning_rate": 3.7323943661971835e-05, "loss": 0.8165, "step": 1325, "task_loss": 0.5618849992752075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5601645112037659, "epoch": 1.12, "learning_rate": 3.735211267605634e-05, "loss": 0.7358, "step": 1326, "task_loss": 0.6365836262702942 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5128225684165955, "epoch": 1.12, "learning_rate": 3.738028169014084e-05, "loss": 0.8503, "step": 1327, "task_loss": 0.6467136740684509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6482024192810059, "epoch": 1.12, "learning_rate": 3.740845070422536e-05, "loss": 0.6229, "step": 1328, "task_loss": 0.5975654125213623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.502515435218811, "epoch": 1.12, "learning_rate": 3.7436619718309864e-05, "loss": 0.9566, "step": 1329, "task_loss": 0.25512880086898804 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8315339088439941, "epoch": 1.12, "learning_rate": 3.746478873239437e-05, "loss": 0.8478, "step": 1330, "task_loss": 1.1492507457733154 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.813011646270752, "epoch": 1.13, "learning_rate": 3.749295774647887e-05, "loss": 0.7969, "step": 1331, "task_loss": 1.2511141300201416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5783092379570007, "epoch": 1.13, "learning_rate": 3.752112676056338e-05, "loss": 0.8815, "step": 1332, "task_loss": 0.7081245183944702 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5627174973487854, "epoch": 1.13, "learning_rate": 3.754929577464789e-05, "loss": 0.7537, "step": 1333, "task_loss": 0.4291169047355652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47017714381217957, "epoch": 1.13, "learning_rate": 3.75774647887324e-05, "loss": 0.7331, "step": 1334, "task_loss": 1.004648208618164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9714229106903076, "epoch": 1.13, "learning_rate": 3.76056338028169e-05, "loss": 0.9644, "step": 1335, "task_loss": 1.3506898880004883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7662819623947144, "epoch": 1.13, "learning_rate": 3.763380281690141e-05, "loss": 0.9737, "step": 1336, "task_loss": 1.3562350273132324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4037238359451294, "epoch": 1.13, "learning_rate": 3.7661971830985915e-05, "loss": 0.9654, "step": 1337, "task_loss": 0.6441592574119568 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.01736581325531, "epoch": 1.13, "learning_rate": 3.769014084507043e-05, "loss": 0.8509, "step": 1338, "task_loss": 1.8972861766815186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7947677373886108, "epoch": 1.13, "learning_rate": 3.771830985915493e-05, "loss": 0.6794, "step": 1339, "task_loss": 1.1238175630569458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8995745182037354, "epoch": 1.13, "learning_rate": 3.774647887323944e-05, "loss": 0.7463, "step": 1340, "task_loss": 1.7476989030838013 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5612698793411255, "epoch": 1.13, "learning_rate": 3.7774647887323944e-05, "loss": 0.8292, "step": 1341, "task_loss": 0.24017272889614105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37615880370140076, "epoch": 1.13, "learning_rate": 3.780281690140845e-05, "loss": 0.8868, "step": 1342, "task_loss": 0.3772379457950592 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5153158903121948, "epoch": 1.14, "learning_rate": 3.783098591549296e-05, "loss": 0.7332, "step": 1343, "task_loss": 0.19181449711322784 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.564118504524231, "epoch": 1.14, "learning_rate": 3.7859154929577466e-05, "loss": 0.9418, "step": 1344, "task_loss": 1.5227885246276855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6710333824157715, "epoch": 1.14, "learning_rate": 3.788732394366197e-05, "loss": 0.7432, "step": 1345, "task_loss": 0.8282901048660278 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9012320041656494, "epoch": 1.14, "learning_rate": 3.791549295774648e-05, "loss": 0.8825, "step": 1346, "task_loss": 1.1163285970687866 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5712893009185791, "epoch": 1.14, "learning_rate": 3.794366197183099e-05, "loss": 0.7851, "step": 1347, "task_loss": 0.49462440609931946 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.570472240447998, "epoch": 1.14, "learning_rate": 3.7971830985915495e-05, "loss": 0.8712, "step": 1348, "task_loss": 1.0322186946868896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47600072622299194, "epoch": 1.14, "learning_rate": 3.8e-05, "loss": 0.6013, "step": 1349, "task_loss": 0.7563858032226562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5716865062713623, "epoch": 1.14, "learning_rate": 3.802816901408451e-05, "loss": 0.6489, "step": 1350, "task_loss": 1.6566252708435059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49220454692840576, "epoch": 1.14, "learning_rate": 3.8056338028169017e-05, "loss": 0.7974, "step": 1351, "task_loss": 0.9204698204994202 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8018453121185303, "epoch": 1.14, "learning_rate": 3.8084507042253524e-05, "loss": 0.9372, "step": 1352, "task_loss": 1.1570929288864136 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2013646364212036, "epoch": 1.14, "learning_rate": 3.811267605633803e-05, "loss": 1.057, "step": 1353, "task_loss": 0.6653337478637695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4932442307472229, "epoch": 1.14, "learning_rate": 3.814084507042254e-05, "loss": 0.6407, "step": 1354, "task_loss": 0.5450311303138733 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5508944988250732, "epoch": 1.15, "learning_rate": 3.8169014084507046e-05, "loss": 0.6934, "step": 1355, "task_loss": 0.4687301814556122 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1773220300674438, "epoch": 1.15, "learning_rate": 3.8197183098591546e-05, "loss": 1.0188, "step": 1356, "task_loss": 0.4553040564060211 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49309512972831726, "epoch": 1.15, "learning_rate": 3.822535211267606e-05, "loss": 0.8781, "step": 1357, "task_loss": 0.28266268968582153 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8344337344169617, "epoch": 1.15, "learning_rate": 3.825352112676057e-05, "loss": 0.9748, "step": 1358, "task_loss": 1.720163106918335 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.072000503540039, "epoch": 1.15, "learning_rate": 3.8281690140845075e-05, "loss": 0.7161, "step": 1359, "task_loss": 0.2949625551700592 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4227490425109863, "epoch": 1.15, "learning_rate": 3.8309859154929575e-05, "loss": 0.9029, "step": 1360, "task_loss": 0.7863182425498962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.161776065826416, "epoch": 1.15, "learning_rate": 3.833802816901408e-05, "loss": 0.9003, "step": 1361, "task_loss": 1.0454025268554688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5051529407501221, "epoch": 1.15, "learning_rate": 3.8366197183098596e-05, "loss": 0.7621, "step": 1362, "task_loss": 0.49472156167030334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45620864629745483, "epoch": 1.15, "learning_rate": 3.8394366197183104e-05, "loss": 0.829, "step": 1363, "task_loss": 0.7446763515472412 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4824894666671753, "epoch": 1.15, "learning_rate": 3.8422535211267604e-05, "loss": 0.7465, "step": 1364, "task_loss": 0.7177490592002869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43231409788131714, "epoch": 1.15, "learning_rate": 3.845070422535211e-05, "loss": 0.6063, "step": 1365, "task_loss": 1.3803067207336426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9414101839065552, "epoch": 1.15, "learning_rate": 3.847887323943662e-05, "loss": 0.8662, "step": 1366, "task_loss": 1.1798189878463745 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48943471908569336, "epoch": 1.16, "learning_rate": 3.850704225352113e-05, "loss": 0.6163, "step": 1367, "task_loss": 0.45160919427871704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.017007827758789, "epoch": 1.16, "learning_rate": 3.853521126760564e-05, "loss": 0.8259, "step": 1368, "task_loss": 0.8634839653968811 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5405937433242798, "epoch": 1.16, "learning_rate": 3.856338028169014e-05, "loss": 0.9796, "step": 1369, "task_loss": 0.6813457012176514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5586112141609192, "epoch": 1.16, "learning_rate": 3.859154929577465e-05, "loss": 0.6258, "step": 1370, "task_loss": 0.8454182147979736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7199917435646057, "epoch": 1.16, "learning_rate": 3.8619718309859155e-05, "loss": 0.6414, "step": 1371, "task_loss": 1.5096521377563477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4697726368904114, "epoch": 1.16, "learning_rate": 3.864788732394367e-05, "loss": 0.729, "step": 1372, "task_loss": 0.4536629021167755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8570728302001953, "epoch": 1.16, "learning_rate": 3.867605633802817e-05, "loss": 0.8744, "step": 1373, "task_loss": 0.8850362300872803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6022423505783081, "epoch": 1.16, "learning_rate": 3.8704225352112677e-05, "loss": 0.8715, "step": 1374, "task_loss": 1.3966511487960815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7351111769676208, "epoch": 1.16, "learning_rate": 3.8732394366197184e-05, "loss": 0.7921, "step": 1375, "task_loss": 1.3739149570465088 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8857632875442505, "epoch": 1.16, "learning_rate": 3.87605633802817e-05, "loss": 0.8269, "step": 1376, "task_loss": 0.8521987795829773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9163758754730225, "epoch": 1.16, "learning_rate": 3.87887323943662e-05, "loss": 0.7638, "step": 1377, "task_loss": 0.9611953496932983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9297792911529541, "epoch": 1.16, "learning_rate": 3.8816901408450706e-05, "loss": 0.8163, "step": 1378, "task_loss": 0.22395800054073334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5782598853111267, "epoch": 1.17, "learning_rate": 3.884507042253521e-05, "loss": 0.6709, "step": 1379, "task_loss": 0.7647679448127747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7844232320785522, "epoch": 1.17, "learning_rate": 3.887323943661972e-05, "loss": 0.7591, "step": 1380, "task_loss": 0.6664432287216187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0925613641738892, "epoch": 1.17, "learning_rate": 3.890140845070423e-05, "loss": 0.8473, "step": 1381, "task_loss": 1.4349753856658936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0698866844177246, "epoch": 1.17, "learning_rate": 3.8929577464788735e-05, "loss": 0.7608, "step": 1382, "task_loss": 0.9003656506538391 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8592077493667603, "epoch": 1.17, "learning_rate": 3.895774647887324e-05, "loss": 0.8426, "step": 1383, "task_loss": 2.6794750690460205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9933497905731201, "epoch": 1.17, "learning_rate": 3.898591549295775e-05, "loss": 0.687, "step": 1384, "task_loss": 0.7604245543479919 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8535611629486084, "epoch": 1.17, "learning_rate": 3.9014084507042256e-05, "loss": 0.7222, "step": 1385, "task_loss": 0.6163215041160583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.72929847240448, "epoch": 1.17, "learning_rate": 3.9042253521126764e-05, "loss": 0.7214, "step": 1386, "task_loss": 0.7554467916488647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.335517555475235, "epoch": 1.17, "learning_rate": 3.907042253521127e-05, "loss": 0.8252, "step": 1387, "task_loss": 0.33588606119155884 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5570566654205322, "epoch": 1.17, "learning_rate": 3.909859154929578e-05, "loss": 0.6968, "step": 1388, "task_loss": 0.43593066930770874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5047706365585327, "epoch": 1.17, "learning_rate": 3.9126760563380285e-05, "loss": 0.5897, "step": 1389, "task_loss": 1.1096971035003662 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0606225728988647, "epoch": 1.17, "learning_rate": 3.9154929577464786e-05, "loss": 0.9759, "step": 1390, "task_loss": 1.0042649507522583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5720828771591187, "epoch": 1.18, "learning_rate": 3.91830985915493e-05, "loss": 0.7701, "step": 1391, "task_loss": 0.7499518990516663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.918779730796814, "epoch": 1.18, "learning_rate": 3.921126760563381e-05, "loss": 0.8901, "step": 1392, "task_loss": 0.44356319308280945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.639498233795166, "epoch": 1.18, "learning_rate": 3.9239436619718314e-05, "loss": 0.8945, "step": 1393, "task_loss": 1.0957310199737549 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7458616495132446, "epoch": 1.18, "learning_rate": 3.9267605633802815e-05, "loss": 0.8344, "step": 1394, "task_loss": 1.20658278465271 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5476272106170654, "epoch": 1.18, "learning_rate": 3.929577464788732e-05, "loss": 0.8759, "step": 1395, "task_loss": 0.842934250831604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.949516773223877, "epoch": 1.18, "learning_rate": 3.9323943661971836e-05, "loss": 0.8425, "step": 1396, "task_loss": 1.1704927682876587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6456067562103271, "epoch": 1.18, "learning_rate": 3.935211267605634e-05, "loss": 0.7357, "step": 1397, "task_loss": 1.445674180984497 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6861137747764587, "epoch": 1.18, "learning_rate": 3.9380281690140844e-05, "loss": 0.7815, "step": 1398, "task_loss": 0.5069717764854431 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6344183683395386, "epoch": 1.18, "learning_rate": 3.940845070422535e-05, "loss": 0.6535, "step": 1399, "task_loss": 1.7080676555633545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0612473487854004, "epoch": 1.18, "learning_rate": 3.943661971830986e-05, "loss": 0.8291, "step": 1400, "task_loss": 1.1031296253204346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7825841307640076, "epoch": 1.18, "learning_rate": 3.946478873239437e-05, "loss": 0.9265, "step": 1401, "task_loss": 1.4064406156539917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7861604690551758, "epoch": 1.19, "learning_rate": 3.949295774647887e-05, "loss": 0.6241, "step": 1402, "task_loss": 0.8486780524253845 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7487976551055908, "epoch": 1.19, "learning_rate": 3.952112676056338e-05, "loss": 0.5328, "step": 1403, "task_loss": 1.8587334156036377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7075173854827881, "epoch": 1.19, "learning_rate": 3.954929577464789e-05, "loss": 0.6368, "step": 1404, "task_loss": 0.8138731718063354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6649550199508667, "epoch": 1.19, "learning_rate": 3.9577464788732395e-05, "loss": 0.8321, "step": 1405, "task_loss": 0.9761863350868225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9430279731750488, "epoch": 1.19, "learning_rate": 3.96056338028169e-05, "loss": 0.6707, "step": 1406, "task_loss": 1.458341360092163 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8182626962661743, "epoch": 1.19, "learning_rate": 3.963380281690141e-05, "loss": 0.9693, "step": 1407, "task_loss": 1.335984706878662 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7342658638954163, "epoch": 1.19, "learning_rate": 3.9661971830985916e-05, "loss": 0.697, "step": 1408, "task_loss": 1.3257267475128174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8703567385673523, "epoch": 1.19, "learning_rate": 3.9690140845070424e-05, "loss": 0.9402, "step": 1409, "task_loss": 1.193984866142273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9058358669281006, "epoch": 1.19, "learning_rate": 3.971830985915493e-05, "loss": 0.7122, "step": 1410, "task_loss": 1.4998536109924316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5770800113677979, "epoch": 1.19, "learning_rate": 3.974647887323944e-05, "loss": 1.0827, "step": 1411, "task_loss": 1.6231870651245117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4893024265766144, "epoch": 1.19, "learning_rate": 3.9774647887323945e-05, "loss": 0.7539, "step": 1412, "task_loss": 0.8779321908950806 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9525576233863831, "epoch": 1.19, "learning_rate": 3.980281690140845e-05, "loss": 0.553, "step": 1413, "task_loss": 0.8644983172416687 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6452755928039551, "epoch": 1.2, "learning_rate": 3.983098591549296e-05, "loss": 0.6284, "step": 1414, "task_loss": 1.0893397331237793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7170165777206421, "epoch": 1.2, "learning_rate": 3.985915492957747e-05, "loss": 0.848, "step": 1415, "task_loss": 1.5706171989440918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5917027592658997, "epoch": 1.2, "learning_rate": 3.9887323943661974e-05, "loss": 0.7523, "step": 1416, "task_loss": 0.9570057392120361 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.982469916343689, "epoch": 1.2, "learning_rate": 3.991549295774648e-05, "loss": 0.94, "step": 1417, "task_loss": 1.5264694690704346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7338570356369019, "epoch": 1.2, "learning_rate": 3.994366197183099e-05, "loss": 0.8222, "step": 1418, "task_loss": 1.105000376701355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49749359488487244, "epoch": 1.2, "learning_rate": 3.9971830985915496e-05, "loss": 0.8415, "step": 1419, "task_loss": 0.8792486190795898 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5241503715515137, "epoch": 1.2, "learning_rate": 4e-05, "loss": 0.7951, "step": 1420, "task_loss": 0.11043868958950043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.30800101161003113, "epoch": 1.2, "learning_rate": 4.002816901408451e-05, "loss": 0.5366, "step": 1421, "task_loss": 0.0944603979587555 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2810603380203247, "epoch": 1.2, "learning_rate": 4.005633802816902e-05, "loss": 0.884, "step": 1422, "task_loss": 2.224720001220703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.34068262577056885, "epoch": 1.2, "learning_rate": 4.0084507042253525e-05, "loss": 0.6329, "step": 1423, "task_loss": 0.6567180156707764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7345470190048218, "epoch": 1.2, "learning_rate": 4.0112676056338025e-05, "loss": 0.6627, "step": 1424, "task_loss": 1.1017460823059082 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7441085577011108, "epoch": 1.2, "learning_rate": 4.014084507042254e-05, "loss": 0.8451, "step": 1425, "task_loss": 1.1204769611358643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41622674465179443, "epoch": 1.21, "learning_rate": 4.016901408450705e-05, "loss": 0.5114, "step": 1426, "task_loss": 0.9898133873939514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9604610204696655, "epoch": 1.21, "learning_rate": 4.0197183098591554e-05, "loss": 0.8228, "step": 1427, "task_loss": 1.6443889141082764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9369698762893677, "epoch": 1.21, "learning_rate": 4.0225352112676054e-05, "loss": 0.9926, "step": 1428, "task_loss": 1.7056491374969482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9260783195495605, "epoch": 1.21, "learning_rate": 4.025352112676056e-05, "loss": 0.7973, "step": 1429, "task_loss": 1.0944664478302002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8343417048454285, "epoch": 1.21, "learning_rate": 4.0281690140845076e-05, "loss": 0.709, "step": 1430, "task_loss": 0.8264630436897278 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4543207585811615, "epoch": 1.21, "learning_rate": 4.030985915492958e-05, "loss": 0.6515, "step": 1431, "task_loss": 0.9087828397750854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.427203893661499, "epoch": 1.21, "learning_rate": 4.0338028169014083e-05, "loss": 1.0075, "step": 1432, "task_loss": 1.7724955081939697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46120402216911316, "epoch": 1.21, "learning_rate": 4.036619718309859e-05, "loss": 0.6048, "step": 1433, "task_loss": 0.2906125783920288 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5413203239440918, "epoch": 1.21, "learning_rate": 4.03943661971831e-05, "loss": 0.6059, "step": 1434, "task_loss": 0.8125205039978027 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4173925817012787, "epoch": 1.21, "learning_rate": 4.042253521126761e-05, "loss": 0.6457, "step": 1435, "task_loss": 0.5229185819625854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9813892841339111, "epoch": 1.21, "learning_rate": 4.045070422535211e-05, "loss": 0.7014, "step": 1436, "task_loss": 0.7678688168525696 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9574260711669922, "epoch": 1.21, "learning_rate": 4.047887323943662e-05, "loss": 0.6997, "step": 1437, "task_loss": 0.8040425181388855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7205783724784851, "epoch": 1.22, "learning_rate": 4.050704225352113e-05, "loss": 0.8122, "step": 1438, "task_loss": 0.7689061164855957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.5196071863174438, "epoch": 1.22, "learning_rate": 4.053521126760564e-05, "loss": 0.9953, "step": 1439, "task_loss": 1.5849289894104004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.425736665725708, "epoch": 1.22, "learning_rate": 4.056338028169014e-05, "loss": 0.6668, "step": 1440, "task_loss": 0.2395995408296585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8614277243614197, "epoch": 1.22, "learning_rate": 4.059154929577465e-05, "loss": 0.9433, "step": 1441, "task_loss": 0.6502854228019714 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7915602922439575, "epoch": 1.22, "learning_rate": 4.0619718309859156e-05, "loss": 0.6728, "step": 1442, "task_loss": 0.8777129650115967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6988682150840759, "epoch": 1.22, "learning_rate": 4.064788732394366e-05, "loss": 0.8165, "step": 1443, "task_loss": 0.8379138708114624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.52309250831604, "epoch": 1.22, "learning_rate": 4.067605633802817e-05, "loss": 0.691, "step": 1444, "task_loss": 0.3691547214984894 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9614949226379395, "epoch": 1.22, "learning_rate": 4.070422535211268e-05, "loss": 0.9305, "step": 1445, "task_loss": 0.48714780807495117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0589027404785156, "epoch": 1.22, "learning_rate": 4.0732394366197185e-05, "loss": 0.8103, "step": 1446, "task_loss": 1.1771272420883179 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.541944146156311, "epoch": 1.22, "learning_rate": 4.076056338028169e-05, "loss": 0.9964, "step": 1447, "task_loss": 1.6577328443527222 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9546430706977844, "epoch": 1.22, "learning_rate": 4.07887323943662e-05, "loss": 0.669, "step": 1448, "task_loss": 0.7795661091804504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6378626823425293, "epoch": 1.22, "learning_rate": 4.081690140845071e-05, "loss": 0.6847, "step": 1449, "task_loss": 0.5675938129425049 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1401309967041016, "epoch": 1.23, "learning_rate": 4.0845070422535214e-05, "loss": 0.951, "step": 1450, "task_loss": 1.6893963813781738 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7616292834281921, "epoch": 1.23, "learning_rate": 4.087323943661972e-05, "loss": 0.8076, "step": 1451, "task_loss": 0.960985541343689 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0787506103515625, "epoch": 1.23, "learning_rate": 4.090140845070423e-05, "loss": 0.8095, "step": 1452, "task_loss": 2.19832706451416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6812751889228821, "epoch": 1.23, "learning_rate": 4.092957746478873e-05, "loss": 0.744, "step": 1453, "task_loss": 0.5638813972473145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5005697011947632, "epoch": 1.23, "learning_rate": 4.095774647887324e-05, "loss": 0.7415, "step": 1454, "task_loss": 0.8492183089256287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6663601994514465, "epoch": 1.23, "learning_rate": 4.098591549295775e-05, "loss": 0.6286, "step": 1455, "task_loss": 0.7005186676979065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5757356882095337, "epoch": 1.23, "learning_rate": 4.101408450704226e-05, "loss": 0.9352, "step": 1456, "task_loss": 0.3871944546699524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5835959911346436, "epoch": 1.23, "learning_rate": 4.104225352112676e-05, "loss": 0.7587, "step": 1457, "task_loss": 0.4098961651325226 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5202710032463074, "epoch": 1.23, "learning_rate": 4.1070422535211265e-05, "loss": 0.919, "step": 1458, "task_loss": 1.1881924867630005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6732559204101562, "epoch": 1.23, "learning_rate": 4.109859154929578e-05, "loss": 0.8497, "step": 1459, "task_loss": 0.9882545471191406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8253720998764038, "epoch": 1.23, "learning_rate": 4.1126760563380286e-05, "loss": 0.8272, "step": 1460, "task_loss": 1.9540684223175049 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8993015289306641, "epoch": 1.23, "learning_rate": 4.115492957746479e-05, "loss": 0.7926, "step": 1461, "task_loss": 1.1706955432891846 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5250594615936279, "epoch": 1.24, "learning_rate": 4.1183098591549294e-05, "loss": 0.7582, "step": 1462, "task_loss": 0.3079270124435425 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.619388222694397, "epoch": 1.24, "learning_rate": 4.12112676056338e-05, "loss": 0.7224, "step": 1463, "task_loss": 0.5681352019309998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5910869836807251, "epoch": 1.24, "learning_rate": 4.1239436619718315e-05, "loss": 0.734, "step": 1464, "task_loss": 1.6270041465759277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5703544020652771, "epoch": 1.24, "learning_rate": 4.126760563380282e-05, "loss": 0.7523, "step": 1465, "task_loss": 0.8180013298988342 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8054250478744507, "epoch": 1.24, "learning_rate": 4.129577464788732e-05, "loss": 0.705, "step": 1466, "task_loss": 0.8697779774665833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5813971161842346, "epoch": 1.24, "learning_rate": 4.132394366197183e-05, "loss": 0.5815, "step": 1467, "task_loss": 0.5868406295776367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7796499133110046, "epoch": 1.24, "learning_rate": 4.1352112676056344e-05, "loss": 0.7437, "step": 1468, "task_loss": 1.1720917224884033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.8090413808822632, "epoch": 1.24, "learning_rate": 4.138028169014085e-05, "loss": 0.823, "step": 1469, "task_loss": 1.8783974647521973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.581538736820221, "epoch": 1.24, "learning_rate": 4.140845070422535e-05, "loss": 0.7519, "step": 1470, "task_loss": 0.5573990345001221 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8268152475357056, "epoch": 1.24, "learning_rate": 4.143661971830986e-05, "loss": 0.6233, "step": 1471, "task_loss": 1.1309012174606323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7492552399635315, "epoch": 1.24, "learning_rate": 4.146478873239437e-05, "loss": 0.8294, "step": 1472, "task_loss": 0.46743038296699524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5401409864425659, "epoch": 1.24, "learning_rate": 4.149295774647888e-05, "loss": 0.5577, "step": 1473, "task_loss": 0.4421212375164032 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5641607046127319, "epoch": 1.25, "learning_rate": 4.152112676056338e-05, "loss": 0.6165, "step": 1474, "task_loss": 0.44506072998046875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4754176735877991, "epoch": 1.25, "learning_rate": 4.154929577464789e-05, "loss": 0.664, "step": 1475, "task_loss": 1.1528629064559937 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7620433568954468, "epoch": 1.25, "learning_rate": 4.1577464788732396e-05, "loss": 0.6327, "step": 1476, "task_loss": 1.5153850317001343 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6689141392707825, "epoch": 1.25, "learning_rate": 4.16056338028169e-05, "loss": 0.6849, "step": 1477, "task_loss": 0.44533202052116394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5448932647705078, "epoch": 1.25, "learning_rate": 4.163380281690141e-05, "loss": 0.5701, "step": 1478, "task_loss": 0.9568987488746643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5239968299865723, "epoch": 1.25, "learning_rate": 4.166197183098592e-05, "loss": 0.6328, "step": 1479, "task_loss": 0.7752301692962646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44894275069236755, "epoch": 1.25, "learning_rate": 4.1690140845070425e-05, "loss": 0.5689, "step": 1480, "task_loss": 0.7350797057151794 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.054861307144165, "epoch": 1.25, "learning_rate": 4.171830985915493e-05, "loss": 0.7421, "step": 1481, "task_loss": 1.2436907291412354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6947156190872192, "epoch": 1.25, "learning_rate": 4.174647887323944e-05, "loss": 0.6968, "step": 1482, "task_loss": 1.9594483375549316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6655158996582031, "epoch": 1.25, "learning_rate": 4.1774647887323946e-05, "loss": 0.8951, "step": 1483, "task_loss": 1.9203691482543945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7013247013092041, "epoch": 1.25, "learning_rate": 4.1802816901408454e-05, "loss": 0.7885, "step": 1484, "task_loss": 0.3448171615600586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5212112665176392, "epoch": 1.26, "learning_rate": 4.183098591549296e-05, "loss": 0.9575, "step": 1485, "task_loss": 0.7477326393127441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5705157518386841, "epoch": 1.26, "learning_rate": 4.185915492957747e-05, "loss": 0.6083, "step": 1486, "task_loss": 0.8950374126434326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48674139380455017, "epoch": 1.26, "learning_rate": 4.188732394366197e-05, "loss": 0.6288, "step": 1487, "task_loss": 0.45856380462646484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0393720865249634, "epoch": 1.26, "learning_rate": 4.191549295774648e-05, "loss": 0.8585, "step": 1488, "task_loss": 0.7326593399047852 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5466912984848022, "epoch": 1.26, "learning_rate": 4.194366197183099e-05, "loss": 0.8062, "step": 1489, "task_loss": 0.7875824570655823 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7664721012115479, "epoch": 1.26, "learning_rate": 4.19718309859155e-05, "loss": 0.7845, "step": 1490, "task_loss": 0.6366121768951416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6253582239151001, "epoch": 1.26, "learning_rate": 4.2e-05, "loss": 0.7607, "step": 1491, "task_loss": 0.7679975032806396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8254096508026123, "epoch": 1.26, "learning_rate": 4.2028169014084505e-05, "loss": 0.5868, "step": 1492, "task_loss": 1.3259700536727905 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.906609058380127, "epoch": 1.26, "learning_rate": 4.205633802816902e-05, "loss": 0.6933, "step": 1493, "task_loss": 1.4294419288635254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8529675006866455, "epoch": 1.26, "learning_rate": 4.2084507042253526e-05, "loss": 0.9228, "step": 1494, "task_loss": 0.5000584125518799 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7976648807525635, "epoch": 1.26, "learning_rate": 4.211267605633803e-05, "loss": 0.7462, "step": 1495, "task_loss": 1.4522374868392944 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5766510963439941, "epoch": 1.26, "learning_rate": 4.2140845070422534e-05, "loss": 0.5288, "step": 1496, "task_loss": 0.5663628578186035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9638341665267944, "epoch": 1.27, "learning_rate": 4.216901408450705e-05, "loss": 0.7274, "step": 1497, "task_loss": 0.6561776399612427 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6617642045021057, "epoch": 1.27, "learning_rate": 4.2197183098591555e-05, "loss": 0.6631, "step": 1498, "task_loss": 0.9563338160514832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5959641933441162, "epoch": 1.27, "learning_rate": 4.2225352112676056e-05, "loss": 0.6495, "step": 1499, "task_loss": 1.1985416412353516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7782770395278931, "epoch": 1.27, "learning_rate": 4.225352112676056e-05, "loss": 0.8561, "step": 1500, "task_loss": 2.391960620880127 }, { "epoch": 1.27, "eval_accuracy": 0.8945346534653466, "eval_loss": 0.41603487730026245, "eval_runtime": 206.7936, "eval_samples_per_second": 122.102, "eval_steps_per_second": 0.957, "step": 1500 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3701817989349365, "epoch": 1.27, "learning_rate": 4.228169014084507e-05, "loss": 0.7922, "step": 1501, "task_loss": 1.4408962726593018 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.82663893699646, "epoch": 1.27, "learning_rate": 4.2309859154929584e-05, "loss": 0.6608, "step": 1502, "task_loss": 0.8414306640625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37749165296554565, "epoch": 1.27, "learning_rate": 4.2338028169014085e-05, "loss": 0.4369, "step": 1503, "task_loss": 1.0078134536743164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9166648387908936, "epoch": 1.27, "learning_rate": 4.236619718309859e-05, "loss": 0.7907, "step": 1504, "task_loss": 0.649882435798645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1793951988220215, "epoch": 1.27, "learning_rate": 4.23943661971831e-05, "loss": 0.8407, "step": 1505, "task_loss": 0.7870943546295166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.623225748538971, "epoch": 1.27, "learning_rate": 4.2422535211267606e-05, "loss": 0.732, "step": 1506, "task_loss": 0.9778086543083191 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4311503469944, "epoch": 1.27, "learning_rate": 4.2450704225352114e-05, "loss": 0.7615, "step": 1507, "task_loss": 0.7427660226821899 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0191118717193604, "epoch": 1.27, "learning_rate": 4.247887323943662e-05, "loss": 0.7351, "step": 1508, "task_loss": 0.9643348455429077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.887663722038269, "epoch": 1.28, "learning_rate": 4.250704225352113e-05, "loss": 0.7607, "step": 1509, "task_loss": 0.48451733589172363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4863021969795227, "epoch": 1.28, "learning_rate": 4.2535211267605635e-05, "loss": 0.6676, "step": 1510, "task_loss": 1.4860094785690308 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.723047137260437, "epoch": 1.28, "learning_rate": 4.256338028169014e-05, "loss": 0.7184, "step": 1511, "task_loss": 0.6996059417724609 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8457887172698975, "epoch": 1.28, "learning_rate": 4.259154929577465e-05, "loss": 0.6875, "step": 1512, "task_loss": 1.0112577676773071 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9305227994918823, "epoch": 1.28, "learning_rate": 4.261971830985916e-05, "loss": 0.6323, "step": 1513, "task_loss": 0.5598640441894531 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8993477821350098, "epoch": 1.28, "learning_rate": 4.2647887323943664e-05, "loss": 0.8259, "step": 1514, "task_loss": 1.6478748321533203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4169367849826813, "epoch": 1.28, "learning_rate": 4.267605633802817e-05, "loss": 0.7097, "step": 1515, "task_loss": 0.025843456387519836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7501055002212524, "epoch": 1.28, "learning_rate": 4.270422535211268e-05, "loss": 0.7811, "step": 1516, "task_loss": 0.9411833882331848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5275061726570129, "epoch": 1.28, "learning_rate": 4.2732394366197186e-05, "loss": 0.7822, "step": 1517, "task_loss": 1.361840009689331 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7466787695884705, "epoch": 1.28, "learning_rate": 4.2760563380281693e-05, "loss": 0.8339, "step": 1518, "task_loss": 0.9366869926452637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6969200968742371, "epoch": 1.28, "learning_rate": 4.27887323943662e-05, "loss": 0.7871, "step": 1519, "task_loss": 0.1534590870141983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9175844192504883, "epoch": 1.28, "learning_rate": 4.281690140845071e-05, "loss": 0.7708, "step": 1520, "task_loss": 1.9030206203460693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6826875805854797, "epoch": 1.29, "learning_rate": 4.284507042253521e-05, "loss": 0.7049, "step": 1521, "task_loss": 1.443913221359253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8458816409111023, "epoch": 1.29, "learning_rate": 4.287323943661972e-05, "loss": 0.7016, "step": 1522, "task_loss": 1.2492772340774536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47657832503318787, "epoch": 1.29, "learning_rate": 4.290140845070423e-05, "loss": 0.6246, "step": 1523, "task_loss": 0.0971175953745842 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8459151983261108, "epoch": 1.29, "learning_rate": 4.292957746478874e-05, "loss": 0.7696, "step": 1524, "task_loss": 0.9176093339920044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9244289398193359, "epoch": 1.29, "learning_rate": 4.295774647887324e-05, "loss": 0.8049, "step": 1525, "task_loss": 1.436991810798645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44274523854255676, "epoch": 1.29, "learning_rate": 4.298591549295775e-05, "loss": 0.6156, "step": 1526, "task_loss": 0.09825313091278076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6414744257926941, "epoch": 1.29, "learning_rate": 4.301408450704226e-05, "loss": 0.7228, "step": 1527, "task_loss": 0.7976199984550476 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7300910353660583, "epoch": 1.29, "learning_rate": 4.3042253521126766e-05, "loss": 0.6451, "step": 1528, "task_loss": 0.3670691251754761 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6434690356254578, "epoch": 1.29, "learning_rate": 4.3070422535211266e-05, "loss": 0.9089, "step": 1529, "task_loss": 1.0599827766418457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6121619939804077, "epoch": 1.29, "learning_rate": 4.3098591549295774e-05, "loss": 0.6698, "step": 1530, "task_loss": 0.6525076627731323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5947908163070679, "epoch": 1.29, "learning_rate": 4.312676056338029e-05, "loss": 0.7732, "step": 1531, "task_loss": 0.510790228843689 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7159149646759033, "epoch": 1.29, "learning_rate": 4.3154929577464795e-05, "loss": 0.5616, "step": 1532, "task_loss": 1.438185691833496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.799270749092102, "epoch": 1.3, "learning_rate": 4.3183098591549295e-05, "loss": 0.7192, "step": 1533, "task_loss": 1.0623685121536255 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7644717693328857, "epoch": 1.3, "learning_rate": 4.32112676056338e-05, "loss": 0.7922, "step": 1534, "task_loss": 1.1115410327911377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49740421772003174, "epoch": 1.3, "learning_rate": 4.323943661971831e-05, "loss": 0.7516, "step": 1535, "task_loss": 0.611073911190033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1715893745422363, "epoch": 1.3, "learning_rate": 4.3267605633802824e-05, "loss": 0.8563, "step": 1536, "task_loss": 0.797667384147644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8696463108062744, "epoch": 1.3, "learning_rate": 4.3295774647887324e-05, "loss": 0.7221, "step": 1537, "task_loss": 1.080940842628479 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0973002910614014, "epoch": 1.3, "learning_rate": 4.332394366197183e-05, "loss": 0.7016, "step": 1538, "task_loss": 0.6256407499313354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7354416847229004, "epoch": 1.3, "learning_rate": 4.335211267605634e-05, "loss": 0.7778, "step": 1539, "task_loss": 0.6478419303894043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1563811302185059, "epoch": 1.3, "learning_rate": 4.3380281690140846e-05, "loss": 0.7436, "step": 1540, "task_loss": 1.4439857006072998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4815584421157837, "epoch": 1.3, "learning_rate": 4.340845070422535e-05, "loss": 0.5114, "step": 1541, "task_loss": 0.34868770837783813 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8675858974456787, "epoch": 1.3, "learning_rate": 4.343661971830986e-05, "loss": 0.7771, "step": 1542, "task_loss": 1.2117412090301514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5264592170715332, "epoch": 1.3, "learning_rate": 4.346478873239437e-05, "loss": 0.758, "step": 1543, "task_loss": 0.5890882015228271 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.185166835784912, "epoch": 1.3, "learning_rate": 4.3492957746478875e-05, "loss": 0.7742, "step": 1544, "task_loss": 2.3525924682617188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6669719815254211, "epoch": 1.31, "learning_rate": 4.352112676056338e-05, "loss": 0.5994, "step": 1545, "task_loss": 0.840576171875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49212849140167236, "epoch": 1.31, "learning_rate": 4.354929577464789e-05, "loss": 0.5237, "step": 1546, "task_loss": 0.45892518758773804 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.567849338054657, "epoch": 1.31, "learning_rate": 4.35774647887324e-05, "loss": 0.7323, "step": 1547, "task_loss": 0.41283532977104187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8312389254570007, "epoch": 1.31, "learning_rate": 4.3605633802816904e-05, "loss": 0.6698, "step": 1548, "task_loss": 1.7425544261932373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8055087924003601, "epoch": 1.31, "learning_rate": 4.363380281690141e-05, "loss": 0.7367, "step": 1549, "task_loss": 0.6188901662826538 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9969013929367065, "epoch": 1.31, "learning_rate": 4.366197183098591e-05, "loss": 0.7566, "step": 1550, "task_loss": 0.6705337762832642 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48357468843460083, "epoch": 1.31, "learning_rate": 4.3690140845070426e-05, "loss": 0.6729, "step": 1551, "task_loss": 0.41909509897232056 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7552318572998047, "epoch": 1.31, "learning_rate": 4.371830985915493e-05, "loss": 0.8841, "step": 1552, "task_loss": 0.8598719239234924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7271772027015686, "epoch": 1.31, "learning_rate": 4.374647887323944e-05, "loss": 0.6663, "step": 1553, "task_loss": 1.1557353734970093 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6499906778335571, "epoch": 1.31, "learning_rate": 4.377464788732394e-05, "loss": 0.6898, "step": 1554, "task_loss": 0.9630948305130005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.542984127998352, "epoch": 1.31, "learning_rate": 4.3802816901408455e-05, "loss": 0.6676, "step": 1555, "task_loss": 0.7775415182113647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7684942483901978, "epoch": 1.32, "learning_rate": 4.383098591549296e-05, "loss": 0.7417, "step": 1556, "task_loss": 0.8520320057868958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5003653764724731, "epoch": 1.32, "learning_rate": 4.385915492957747e-05, "loss": 0.5853, "step": 1557, "task_loss": 0.9681815505027771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6999002695083618, "epoch": 1.32, "learning_rate": 4.388732394366197e-05, "loss": 0.7303, "step": 1558, "task_loss": 0.46428707242012024 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5340083241462708, "epoch": 1.32, "learning_rate": 4.391549295774648e-05, "loss": 0.6845, "step": 1559, "task_loss": 1.261164903640747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8254595398902893, "epoch": 1.32, "learning_rate": 4.394366197183099e-05, "loss": 0.7834, "step": 1560, "task_loss": 1.1078704595565796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0518798828125, "epoch": 1.32, "learning_rate": 4.39718309859155e-05, "loss": 0.8875, "step": 1561, "task_loss": 0.927859902381897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3842101991176605, "epoch": 1.32, "learning_rate": 4.4000000000000006e-05, "loss": 0.4522, "step": 1562, "task_loss": 0.11398382484912872 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6807934045791626, "epoch": 1.32, "learning_rate": 4.4028169014084506e-05, "loss": 0.8432, "step": 1563, "task_loss": 0.5401532053947449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6663694977760315, "epoch": 1.32, "learning_rate": 4.405633802816901e-05, "loss": 0.7077, "step": 1564, "task_loss": 1.4222922325134277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8710164427757263, "epoch": 1.32, "learning_rate": 4.408450704225353e-05, "loss": 0.5614, "step": 1565, "task_loss": 0.39943355321884155 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0948066711425781, "epoch": 1.32, "learning_rate": 4.4112676056338035e-05, "loss": 0.9351, "step": 1566, "task_loss": 0.779899001121521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6991064548492432, "epoch": 1.32, "learning_rate": 4.4140845070422535e-05, "loss": 0.6366, "step": 1567, "task_loss": 0.6332132816314697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7855426073074341, "epoch": 1.33, "learning_rate": 4.416901408450704e-05, "loss": 0.7496, "step": 1568, "task_loss": 0.7602503299713135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5742967128753662, "epoch": 1.33, "learning_rate": 4.419718309859155e-05, "loss": 0.8434, "step": 1569, "task_loss": 0.7736462950706482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3363397717475891, "epoch": 1.33, "learning_rate": 4.4225352112676064e-05, "loss": 0.5154, "step": 1570, "task_loss": 0.5790085792541504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6399105787277222, "epoch": 1.33, "learning_rate": 4.4253521126760564e-05, "loss": 0.5074, "step": 1571, "task_loss": 0.5580586194992065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5467866659164429, "epoch": 1.33, "learning_rate": 4.428169014084507e-05, "loss": 0.4928, "step": 1572, "task_loss": 0.2578856348991394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41350653767585754, "epoch": 1.33, "learning_rate": 4.430985915492958e-05, "loss": 0.5029, "step": 1573, "task_loss": 0.08998537808656693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6312925815582275, "epoch": 1.33, "learning_rate": 4.4338028169014086e-05, "loss": 0.547, "step": 1574, "task_loss": 0.710391640663147 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5399519205093384, "epoch": 1.33, "learning_rate": 4.436619718309859e-05, "loss": 0.5491, "step": 1575, "task_loss": 1.1737101078033447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46089524030685425, "epoch": 1.33, "learning_rate": 4.43943661971831e-05, "loss": 0.4725, "step": 1576, "task_loss": 0.2197568714618683 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4201352596282959, "epoch": 1.33, "learning_rate": 4.442253521126761e-05, "loss": 0.5006, "step": 1577, "task_loss": 0.24145811796188354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7045038938522339, "epoch": 1.33, "learning_rate": 4.4450704225352115e-05, "loss": 0.7575, "step": 1578, "task_loss": 0.14815276861190796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5899484753608704, "epoch": 1.33, "learning_rate": 4.447887323943662e-05, "loss": 0.7878, "step": 1579, "task_loss": 1.2533992528915405 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7741119861602783, "epoch": 1.34, "learning_rate": 4.450704225352113e-05, "loss": 0.6783, "step": 1580, "task_loss": 0.8779782056808472 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.703798234462738, "epoch": 1.34, "learning_rate": 4.4535211267605637e-05, "loss": 0.7993, "step": 1581, "task_loss": 1.0945786237716675 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6185953617095947, "epoch": 1.34, "learning_rate": 4.4563380281690144e-05, "loss": 0.6986, "step": 1582, "task_loss": 1.1421067714691162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38674333691596985, "epoch": 1.34, "learning_rate": 4.459154929577465e-05, "loss": 0.6026, "step": 1583, "task_loss": 0.5906066298484802 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9399058818817139, "epoch": 1.34, "learning_rate": 4.461971830985916e-05, "loss": 0.7688, "step": 1584, "task_loss": 1.2046525478363037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.920735239982605, "epoch": 1.34, "learning_rate": 4.4647887323943666e-05, "loss": 0.7085, "step": 1585, "task_loss": 0.38639959692955017 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5597467422485352, "epoch": 1.34, "learning_rate": 4.467605633802817e-05, "loss": 0.6571, "step": 1586, "task_loss": 0.3095024526119232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6345034837722778, "epoch": 1.34, "learning_rate": 4.470422535211268e-05, "loss": 0.5954, "step": 1587, "task_loss": 0.30299389362335205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8126978278160095, "epoch": 1.34, "learning_rate": 4.473239436619718e-05, "loss": 0.7011, "step": 1588, "task_loss": 0.4249064028263092 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3777484893798828, "epoch": 1.34, "learning_rate": 4.4760563380281695e-05, "loss": 0.6387, "step": 1589, "task_loss": 0.5378279685974121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.746435284614563, "epoch": 1.34, "learning_rate": 4.47887323943662e-05, "loss": 0.7699, "step": 1590, "task_loss": 0.8520925641059875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8152430057525635, "epoch": 1.34, "learning_rate": 4.481690140845071e-05, "loss": 0.6377, "step": 1591, "task_loss": 0.7060929536819458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5515363812446594, "epoch": 1.35, "learning_rate": 4.484507042253521e-05, "loss": 0.5346, "step": 1592, "task_loss": 1.0614104270935059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8067082762718201, "epoch": 1.35, "learning_rate": 4.487323943661972e-05, "loss": 0.5925, "step": 1593, "task_loss": 0.973268985748291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3601953387260437, "epoch": 1.35, "learning_rate": 4.490140845070423e-05, "loss": 0.564, "step": 1594, "task_loss": 0.6473004817962646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5404622554779053, "epoch": 1.35, "learning_rate": 4.492957746478874e-05, "loss": 0.6333, "step": 1595, "task_loss": 0.16768519580364227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3098883628845215, "epoch": 1.35, "learning_rate": 4.495774647887324e-05, "loss": 0.7722, "step": 1596, "task_loss": 1.1064550876617432 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5654951333999634, "epoch": 1.35, "learning_rate": 4.4985915492957746e-05, "loss": 0.7023, "step": 1597, "task_loss": 0.9072781801223755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5111460089683533, "epoch": 1.35, "learning_rate": 4.501408450704225e-05, "loss": 0.6201, "step": 1598, "task_loss": 0.5623824000358582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.372096061706543, "epoch": 1.35, "learning_rate": 4.504225352112677e-05, "loss": 0.9342, "step": 1599, "task_loss": 1.5337944030761719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5938840508460999, "epoch": 1.35, "learning_rate": 4.507042253521127e-05, "loss": 0.7117, "step": 1600, "task_loss": 0.6396026611328125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5004254579544067, "epoch": 1.35, "learning_rate": 4.5098591549295775e-05, "loss": 0.7833, "step": 1601, "task_loss": 1.2400044202804565 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6157674193382263, "epoch": 1.35, "learning_rate": 4.512676056338028e-05, "loss": 0.7289, "step": 1602, "task_loss": 0.40302708745002747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8182985782623291, "epoch": 1.35, "learning_rate": 4.515492957746479e-05, "loss": 0.6799, "step": 1603, "task_loss": 0.48437926173210144 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3704138398170471, "epoch": 1.36, "learning_rate": 4.5183098591549297e-05, "loss": 0.7657, "step": 1604, "task_loss": 0.26412326097488403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7256048917770386, "epoch": 1.36, "learning_rate": 4.5211267605633804e-05, "loss": 0.6874, "step": 1605, "task_loss": 0.19763152301311493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8576298356056213, "epoch": 1.36, "learning_rate": 4.523943661971831e-05, "loss": 0.7063, "step": 1606, "task_loss": 0.8036271333694458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6846177577972412, "epoch": 1.36, "learning_rate": 4.526760563380282e-05, "loss": 0.7224, "step": 1607, "task_loss": 0.9296252131462097 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8795887231826782, "epoch": 1.36, "learning_rate": 4.5295774647887326e-05, "loss": 0.701, "step": 1608, "task_loss": 0.9749605059623718 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4876706600189209, "epoch": 1.36, "learning_rate": 4.532394366197183e-05, "loss": 0.5273, "step": 1609, "task_loss": 0.9442476630210876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9191760420799255, "epoch": 1.36, "learning_rate": 4.535211267605634e-05, "loss": 0.903, "step": 1610, "task_loss": 1.2675002813339233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0397008657455444, "epoch": 1.36, "learning_rate": 4.538028169014085e-05, "loss": 0.9053, "step": 1611, "task_loss": 0.9247954487800598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8439387679100037, "epoch": 1.36, "learning_rate": 4.5408450704225355e-05, "loss": 0.701, "step": 1612, "task_loss": 0.7955954670906067 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4201531708240509, "epoch": 1.36, "learning_rate": 4.543661971830986e-05, "loss": 0.5991, "step": 1613, "task_loss": 0.6970077157020569 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6968681812286377, "epoch": 1.36, "learning_rate": 4.546478873239437e-05, "loss": 0.603, "step": 1614, "task_loss": 1.2739346027374268 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5191566944122314, "epoch": 1.36, "learning_rate": 4.5492957746478876e-05, "loss": 0.551, "step": 1615, "task_loss": 0.5945554971694946 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.659951388835907, "epoch": 1.37, "learning_rate": 4.5521126760563384e-05, "loss": 0.6596, "step": 1616, "task_loss": 0.6435373425483704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7264826893806458, "epoch": 1.37, "learning_rate": 4.554929577464789e-05, "loss": 0.5811, "step": 1617, "task_loss": 0.7037597894668579 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3778638243675232, "epoch": 1.37, "learning_rate": 4.55774647887324e-05, "loss": 0.5639, "step": 1618, "task_loss": 0.9364375472068787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4664233326911926, "epoch": 1.37, "learning_rate": 4.5605633802816905e-05, "loss": 0.6283, "step": 1619, "task_loss": 0.3525354266166687 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48258882761001587, "epoch": 1.37, "learning_rate": 4.563380281690141e-05, "loss": 0.723, "step": 1620, "task_loss": 0.048179417848587036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9769374132156372, "epoch": 1.37, "learning_rate": 4.566197183098592e-05, "loss": 0.7774, "step": 1621, "task_loss": 0.9386952519416809 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44537508487701416, "epoch": 1.37, "learning_rate": 4.569014084507042e-05, "loss": 0.6961, "step": 1622, "task_loss": 0.688592791557312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4803049564361572, "epoch": 1.37, "learning_rate": 4.5718309859154934e-05, "loss": 0.5264, "step": 1623, "task_loss": 1.2706339359283447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7576771378517151, "epoch": 1.37, "learning_rate": 4.574647887323944e-05, "loss": 0.7053, "step": 1624, "task_loss": 1.0819844007492065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5964647531509399, "epoch": 1.37, "learning_rate": 4.577464788732395e-05, "loss": 0.776, "step": 1625, "task_loss": 0.9251953363418579 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6509608626365662, "epoch": 1.37, "learning_rate": 4.580281690140845e-05, "loss": 0.6557, "step": 1626, "task_loss": 0.6880785822868347 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.593774676322937, "epoch": 1.38, "learning_rate": 4.5830985915492957e-05, "loss": 0.5495, "step": 1627, "task_loss": 1.0458431243896484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6694990992546082, "epoch": 1.38, "learning_rate": 4.585915492957747e-05, "loss": 0.6394, "step": 1628, "task_loss": 0.6873940825462341 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8138754367828369, "epoch": 1.38, "learning_rate": 4.588732394366198e-05, "loss": 0.6942, "step": 1629, "task_loss": 0.6464247703552246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7374907732009888, "epoch": 1.38, "learning_rate": 4.591549295774648e-05, "loss": 0.7856, "step": 1630, "task_loss": 0.9992972612380981 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7398185729980469, "epoch": 1.38, "learning_rate": 4.5943661971830986e-05, "loss": 0.5943, "step": 1631, "task_loss": 1.232700228691101 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4812476336956024, "epoch": 1.38, "learning_rate": 4.597183098591549e-05, "loss": 0.6492, "step": 1632, "task_loss": 0.9382060170173645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46417078375816345, "epoch": 1.38, "learning_rate": 4.600000000000001e-05, "loss": 0.5702, "step": 1633, "task_loss": 0.3191852867603302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8275821208953857, "epoch": 1.38, "learning_rate": 4.602816901408451e-05, "loss": 0.9403, "step": 1634, "task_loss": 1.2224198579788208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9413549900054932, "epoch": 1.38, "learning_rate": 4.6056338028169015e-05, "loss": 0.7617, "step": 1635, "task_loss": 0.7549138069152832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5871737599372864, "epoch": 1.38, "learning_rate": 4.608450704225352e-05, "loss": 0.6801, "step": 1636, "task_loss": 1.3034164905548096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6202578544616699, "epoch": 1.38, "learning_rate": 4.611267605633803e-05, "loss": 0.8044, "step": 1637, "task_loss": 0.5311557054519653 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.005652904510498, "epoch": 1.38, "learning_rate": 4.6140845070422536e-05, "loss": 0.8017, "step": 1638, "task_loss": 1.7919282913208008 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6073235273361206, "epoch": 1.39, "learning_rate": 4.6169014084507044e-05, "loss": 0.7203, "step": 1639, "task_loss": 0.5981647372245789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5321988463401794, "epoch": 1.39, "learning_rate": 4.619718309859155e-05, "loss": 0.7767, "step": 1640, "task_loss": 1.2166517972946167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3905990421772003, "epoch": 1.39, "learning_rate": 4.622535211267606e-05, "loss": 0.6165, "step": 1641, "task_loss": 0.5900824069976807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.31377795338630676, "epoch": 1.39, "learning_rate": 4.6253521126760565e-05, "loss": 0.6328, "step": 1642, "task_loss": 0.02757611684501171 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9510294795036316, "epoch": 1.39, "learning_rate": 4.628169014084507e-05, "loss": 0.8472, "step": 1643, "task_loss": 0.6467332243919373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9068036675453186, "epoch": 1.39, "learning_rate": 4.630985915492958e-05, "loss": 0.6127, "step": 1644, "task_loss": 0.7072877883911133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7350319623947144, "epoch": 1.39, "learning_rate": 4.633802816901409e-05, "loss": 0.6444, "step": 1645, "task_loss": 0.8510034084320068 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5910776257514954, "epoch": 1.39, "learning_rate": 4.6366197183098594e-05, "loss": 0.6956, "step": 1646, "task_loss": 0.3775237798690796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6216821074485779, "epoch": 1.39, "learning_rate": 4.63943661971831e-05, "loss": 0.5952, "step": 1647, "task_loss": 1.11821448802948 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6410535573959351, "epoch": 1.39, "learning_rate": 4.642253521126761e-05, "loss": 0.671, "step": 1648, "task_loss": 0.7640874981880188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5683889985084534, "epoch": 1.39, "learning_rate": 4.6450704225352116e-05, "loss": 0.8037, "step": 1649, "task_loss": 1.2884925603866577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4782760739326477, "epoch": 1.39, "learning_rate": 4.647887323943662e-05, "loss": 0.706, "step": 1650, "task_loss": 0.8123883605003357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6763464212417603, "epoch": 1.4, "learning_rate": 4.6507042253521124e-05, "loss": 0.8179, "step": 1651, "task_loss": 0.773410975933075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3630438446998596, "epoch": 1.4, "learning_rate": 4.653521126760564e-05, "loss": 0.7071, "step": 1652, "task_loss": 0.2774551212787628 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48728662729263306, "epoch": 1.4, "learning_rate": 4.6563380281690145e-05, "loss": 0.5677, "step": 1653, "task_loss": 0.3113407492637634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6041675806045532, "epoch": 1.4, "learning_rate": 4.659154929577465e-05, "loss": 0.6528, "step": 1654, "task_loss": 1.2924110889434814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3602876663208008, "epoch": 1.4, "learning_rate": 4.661971830985915e-05, "loss": 0.696, "step": 1655, "task_loss": 0.09083258360624313 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7984553575515747, "epoch": 1.4, "learning_rate": 4.664788732394366e-05, "loss": 0.7322, "step": 1656, "task_loss": 0.4074188768863678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5338616371154785, "epoch": 1.4, "learning_rate": 4.6676056338028174e-05, "loss": 0.6984, "step": 1657, "task_loss": 1.3142259120941162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40345728397369385, "epoch": 1.4, "learning_rate": 4.670422535211268e-05, "loss": 0.6494, "step": 1658, "task_loss": 0.17130941152572632 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7828407287597656, "epoch": 1.4, "learning_rate": 4.673239436619719e-05, "loss": 0.5009, "step": 1659, "task_loss": 1.0953917503356934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5598666071891785, "epoch": 1.4, "learning_rate": 4.676056338028169e-05, "loss": 0.7247, "step": 1660, "task_loss": 0.22864079475402832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7103433609008789, "epoch": 1.4, "learning_rate": 4.6788732394366196e-05, "loss": 0.7511, "step": 1661, "task_loss": 0.6112599968910217 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6918916702270508, "epoch": 1.4, "learning_rate": 4.681690140845071e-05, "loss": 0.5303, "step": 1662, "task_loss": 0.2006126493215561 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7058935761451721, "epoch": 1.41, "learning_rate": 4.684507042253522e-05, "loss": 0.694, "step": 1663, "task_loss": 0.8665253520011902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5211086273193359, "epoch": 1.41, "learning_rate": 4.687323943661972e-05, "loss": 0.76, "step": 1664, "task_loss": 0.2786552309989929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5105043053627014, "epoch": 1.41, "learning_rate": 4.6901408450704225e-05, "loss": 0.6091, "step": 1665, "task_loss": 0.6466337442398071 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.187175989151001, "epoch": 1.41, "learning_rate": 4.692957746478873e-05, "loss": 0.8059, "step": 1666, "task_loss": 1.0322376489639282 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8007494211196899, "epoch": 1.41, "learning_rate": 4.6957746478873247e-05, "loss": 0.8035, "step": 1667, "task_loss": 1.5951308012008667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5322604775428772, "epoch": 1.41, "learning_rate": 4.698591549295775e-05, "loss": 0.4752, "step": 1668, "task_loss": 0.5678632259368896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.29829883575439453, "epoch": 1.41, "learning_rate": 4.7014084507042254e-05, "loss": 0.5926, "step": 1669, "task_loss": 0.9094152450561523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6462032794952393, "epoch": 1.41, "learning_rate": 4.704225352112676e-05, "loss": 0.6934, "step": 1670, "task_loss": 0.590508222579956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2933039665222168, "epoch": 1.41, "learning_rate": 4.707042253521127e-05, "loss": 0.4964, "step": 1671, "task_loss": 0.5093124508857727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.604745090007782, "epoch": 1.41, "learning_rate": 4.7098591549295776e-05, "loss": 0.7375, "step": 1672, "task_loss": 0.4759080111980438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46596410870552063, "epoch": 1.41, "learning_rate": 4.712676056338028e-05, "loss": 0.8209, "step": 1673, "task_loss": 0.7101446390151978 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4847882390022278, "epoch": 1.41, "learning_rate": 4.715492957746479e-05, "loss": 0.6404, "step": 1674, "task_loss": 0.2981914281845093 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6020488739013672, "epoch": 1.42, "learning_rate": 4.71830985915493e-05, "loss": 0.5821, "step": 1675, "task_loss": 0.47541120648384094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5112794041633606, "epoch": 1.42, "learning_rate": 4.7211267605633805e-05, "loss": 0.4681, "step": 1676, "task_loss": 0.9648087620735168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0668083429336548, "epoch": 1.42, "learning_rate": 4.723943661971831e-05, "loss": 0.7234, "step": 1677, "task_loss": 0.5064386129379272 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0904996395111084, "epoch": 1.42, "learning_rate": 4.726760563380282e-05, "loss": 0.9536, "step": 1678, "task_loss": 0.8612411618232727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.471343457698822, "epoch": 1.42, "learning_rate": 4.729577464788733e-05, "loss": 0.633, "step": 1679, "task_loss": 0.7127698063850403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4019850790500641, "epoch": 1.42, "learning_rate": 4.7323943661971834e-05, "loss": 0.6514, "step": 1680, "task_loss": 0.8699890375137329 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6972496509552002, "epoch": 1.42, "learning_rate": 4.735211267605634e-05, "loss": 0.5258, "step": 1681, "task_loss": 0.9608989953994751 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5665212869644165, "epoch": 1.42, "learning_rate": 4.738028169014085e-05, "loss": 0.4864, "step": 1682, "task_loss": 0.3217461109161377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7031760215759277, "epoch": 1.42, "learning_rate": 4.7408450704225356e-05, "loss": 0.4713, "step": 1683, "task_loss": 0.23230141401290894 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48633596301078796, "epoch": 1.42, "learning_rate": 4.743661971830986e-05, "loss": 0.5437, "step": 1684, "task_loss": 1.0164079666137695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7596257925033569, "epoch": 1.42, "learning_rate": 4.7464788732394363e-05, "loss": 0.744, "step": 1685, "task_loss": 0.4292535185813904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.64918053150177, "epoch": 1.42, "learning_rate": 4.749295774647888e-05, "loss": 0.613, "step": 1686, "task_loss": 0.5599294304847717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43048256635665894, "epoch": 1.43, "learning_rate": 4.7521126760563385e-05, "loss": 0.6673, "step": 1687, "task_loss": 0.7167811393737793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6201004385948181, "epoch": 1.43, "learning_rate": 4.754929577464789e-05, "loss": 0.7049, "step": 1688, "task_loss": 0.6512680053710938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49391043186187744, "epoch": 1.43, "learning_rate": 4.757746478873239e-05, "loss": 0.6781, "step": 1689, "task_loss": 0.5704379677772522 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5168977975845337, "epoch": 1.43, "learning_rate": 4.76056338028169e-05, "loss": 0.8071, "step": 1690, "task_loss": 0.3104069232940674 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7273399829864502, "epoch": 1.43, "learning_rate": 4.7633802816901414e-05, "loss": 0.6757, "step": 1691, "task_loss": 1.2106983661651611 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8390233516693115, "epoch": 1.43, "learning_rate": 4.766197183098592e-05, "loss": 0.517, "step": 1692, "task_loss": 1.0718122720718384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.551091730594635, "epoch": 1.43, "learning_rate": 4.769014084507042e-05, "loss": 0.6235, "step": 1693, "task_loss": 0.07249514013528824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3700655698776245, "epoch": 1.43, "learning_rate": 4.771830985915493e-05, "loss": 0.7134, "step": 1694, "task_loss": 0.07154733687639236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6866461038589478, "epoch": 1.43, "learning_rate": 4.7746478873239436e-05, "loss": 0.6708, "step": 1695, "task_loss": 1.29938542842865 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6798800826072693, "epoch": 1.43, "learning_rate": 4.777464788732395e-05, "loss": 0.7324, "step": 1696, "task_loss": 1.183453917503357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9477272033691406, "epoch": 1.43, "learning_rate": 4.780281690140845e-05, "loss": 0.7289, "step": 1697, "task_loss": 1.8966968059539795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4756430983543396, "epoch": 1.44, "learning_rate": 4.783098591549296e-05, "loss": 0.7531, "step": 1698, "task_loss": 1.0449330806732178 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.630846381187439, "epoch": 1.44, "learning_rate": 4.7859154929577465e-05, "loss": 0.7821, "step": 1699, "task_loss": 1.8963607549667358 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6619269847869873, "epoch": 1.44, "learning_rate": 4.788732394366197e-05, "loss": 0.9894, "step": 1700, "task_loss": 0.8408291339874268 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.640231728553772, "epoch": 1.44, "learning_rate": 4.791549295774648e-05, "loss": 0.7756, "step": 1701, "task_loss": 1.1331188678741455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43245363235473633, "epoch": 1.44, "learning_rate": 4.794366197183099e-05, "loss": 0.5697, "step": 1702, "task_loss": 0.9229650497436523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46145695447921753, "epoch": 1.44, "learning_rate": 4.7971830985915494e-05, "loss": 0.646, "step": 1703, "task_loss": 0.915656328201294 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40333157777786255, "epoch": 1.44, "learning_rate": 4.8e-05, "loss": 0.6202, "step": 1704, "task_loss": 0.3361087143421173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.753397524356842, "epoch": 1.44, "learning_rate": 4.8028169014084515e-05, "loss": 0.7937, "step": 1705, "task_loss": 1.1469022035598755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5252906084060669, "epoch": 1.44, "learning_rate": 4.8056338028169016e-05, "loss": 0.7107, "step": 1706, "task_loss": 1.1360138654708862 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7311115264892578, "epoch": 1.44, "learning_rate": 4.808450704225352e-05, "loss": 0.6374, "step": 1707, "task_loss": 0.7260037064552307 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4606877565383911, "epoch": 1.44, "learning_rate": 4.811267605633803e-05, "loss": 0.5541, "step": 1708, "task_loss": 0.1935642808675766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1537079811096191, "epoch": 1.44, "learning_rate": 4.814084507042254e-05, "loss": 0.7161, "step": 1709, "task_loss": 0.8487422466278076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9706428647041321, "epoch": 1.45, "learning_rate": 4.8169014084507045e-05, "loss": 0.7159, "step": 1710, "task_loss": 0.8153154850006104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6801272630691528, "epoch": 1.45, "learning_rate": 4.819718309859155e-05, "loss": 0.8424, "step": 1711, "task_loss": 1.5993804931640625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5495634078979492, "epoch": 1.45, "learning_rate": 4.822535211267606e-05, "loss": 0.6038, "step": 1712, "task_loss": 0.7332903742790222 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40992289781570435, "epoch": 1.45, "learning_rate": 4.8253521126760566e-05, "loss": 0.6271, "step": 1713, "task_loss": 0.6718761920928955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5004311800003052, "epoch": 1.45, "learning_rate": 4.8281690140845074e-05, "loss": 0.7362, "step": 1714, "task_loss": 1.753472924232483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44008588790893555, "epoch": 1.45, "learning_rate": 4.830985915492958e-05, "loss": 0.5271, "step": 1715, "task_loss": 0.652900218963623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5127201080322266, "epoch": 1.45, "learning_rate": 4.833802816901409e-05, "loss": 0.5959, "step": 1716, "task_loss": 0.2781747579574585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6766136288642883, "epoch": 1.45, "learning_rate": 4.8366197183098595e-05, "loss": 0.651, "step": 1717, "task_loss": 1.0642186403274536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6986705660820007, "epoch": 1.45, "learning_rate": 4.83943661971831e-05, "loss": 0.6736, "step": 1718, "task_loss": 1.452517032623291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8774093389511108, "epoch": 1.45, "learning_rate": 4.84225352112676e-05, "loss": 0.7971, "step": 1719, "task_loss": 0.5574795007705688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6652801036834717, "epoch": 1.45, "learning_rate": 4.845070422535212e-05, "loss": 0.795, "step": 1720, "task_loss": 1.1690058708190918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8377767205238342, "epoch": 1.45, "learning_rate": 4.8478873239436624e-05, "loss": 0.5275, "step": 1721, "task_loss": 1.272325038909912 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6717647314071655, "epoch": 1.46, "learning_rate": 4.850704225352113e-05, "loss": 0.7104, "step": 1722, "task_loss": 0.990061342716217 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40907326340675354, "epoch": 1.46, "learning_rate": 4.853521126760563e-05, "loss": 0.6054, "step": 1723, "task_loss": 0.6067416071891785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6697214841842651, "epoch": 1.46, "learning_rate": 4.856338028169014e-05, "loss": 0.6997, "step": 1724, "task_loss": 0.9226334095001221 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8197699785232544, "epoch": 1.46, "learning_rate": 4.8591549295774653e-05, "loss": 0.8772, "step": 1725, "task_loss": 1.5548620223999023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5488120913505554, "epoch": 1.46, "learning_rate": 4.861971830985916e-05, "loss": 0.6337, "step": 1726, "task_loss": 0.7327488660812378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6403170228004456, "epoch": 1.46, "learning_rate": 4.864788732394366e-05, "loss": 0.7662, "step": 1727, "task_loss": 0.5010391473770142 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37367990612983704, "epoch": 1.46, "learning_rate": 4.867605633802817e-05, "loss": 0.4755, "step": 1728, "task_loss": 0.24142694473266602 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5752547979354858, "epoch": 1.46, "learning_rate": 4.8704225352112676e-05, "loss": 0.7167, "step": 1729, "task_loss": 0.19948241114616394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.178580641746521, "epoch": 1.46, "learning_rate": 4.873239436619719e-05, "loss": 0.6603, "step": 1730, "task_loss": 0.8581843972206116 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6714542508125305, "epoch": 1.46, "learning_rate": 4.876056338028169e-05, "loss": 0.6248, "step": 1731, "task_loss": 0.6137378215789795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7051255702972412, "epoch": 1.46, "learning_rate": 4.87887323943662e-05, "loss": 0.7824, "step": 1732, "task_loss": 0.6099342107772827 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5463262796401978, "epoch": 1.46, "learning_rate": 4.8816901408450705e-05, "loss": 0.794, "step": 1733, "task_loss": 0.8290018439292908 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6452159881591797, "epoch": 1.47, "learning_rate": 4.884507042253522e-05, "loss": 0.6002, "step": 1734, "task_loss": 1.5022836923599243 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5066654682159424, "epoch": 1.47, "learning_rate": 4.887323943661972e-05, "loss": 0.4403, "step": 1735, "task_loss": 0.22324420511722565 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7978217601776123, "epoch": 1.47, "learning_rate": 4.8901408450704226e-05, "loss": 0.6448, "step": 1736, "task_loss": 1.0378270149230957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0954978466033936, "epoch": 1.47, "learning_rate": 4.8929577464788734e-05, "loss": 0.6768, "step": 1737, "task_loss": 0.5352761745452881 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7706320285797119, "epoch": 1.47, "learning_rate": 4.895774647887324e-05, "loss": 0.6284, "step": 1738, "task_loss": 1.1697399616241455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45554298162460327, "epoch": 1.47, "learning_rate": 4.898591549295775e-05, "loss": 0.5134, "step": 1739, "task_loss": 0.4110294580459595 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.453569233417511, "epoch": 1.47, "learning_rate": 4.9014084507042255e-05, "loss": 0.6389, "step": 1740, "task_loss": 0.342339426279068 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43892428278923035, "epoch": 1.47, "learning_rate": 4.904225352112676e-05, "loss": 0.4931, "step": 1741, "task_loss": 1.4942964315414429 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6022356152534485, "epoch": 1.47, "learning_rate": 4.907042253521127e-05, "loss": 0.6562, "step": 1742, "task_loss": 1.7500537633895874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5086394548416138, "epoch": 1.47, "learning_rate": 4.909859154929578e-05, "loss": 0.6551, "step": 1743, "task_loss": 0.767167329788208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7103742361068726, "epoch": 1.47, "learning_rate": 4.9126760563380284e-05, "loss": 0.6312, "step": 1744, "task_loss": 0.8919826149940491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5152264833450317, "epoch": 1.47, "learning_rate": 4.915492957746479e-05, "loss": 0.6893, "step": 1745, "task_loss": 0.27684393525123596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7589502334594727, "epoch": 1.48, "learning_rate": 4.91830985915493e-05, "loss": 0.6977, "step": 1746, "task_loss": 1.6448665857315063 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.876821756362915, "epoch": 1.48, "learning_rate": 4.9211267605633806e-05, "loss": 0.7055, "step": 1747, "task_loss": 0.4326273798942566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.63507080078125, "epoch": 1.48, "learning_rate": 4.923943661971831e-05, "loss": 0.7002, "step": 1748, "task_loss": 1.0050069093704224 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7446730136871338, "epoch": 1.48, "learning_rate": 4.926760563380282e-05, "loss": 0.6175, "step": 1749, "task_loss": 1.0677016973495483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4185749888420105, "epoch": 1.48, "learning_rate": 4.929577464788733e-05, "loss": 0.5881, "step": 1750, "task_loss": 0.7596157789230347 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6640006303787231, "epoch": 1.48, "learning_rate": 4.9323943661971835e-05, "loss": 0.6238, "step": 1751, "task_loss": 0.8177092671394348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3937869369983673, "epoch": 1.48, "learning_rate": 4.9352112676056336e-05, "loss": 0.6965, "step": 1752, "task_loss": 0.27502498030662537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.33980023860931396, "epoch": 1.48, "learning_rate": 4.938028169014084e-05, "loss": 0.5322, "step": 1753, "task_loss": 0.3384622037410736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7791434526443481, "epoch": 1.48, "learning_rate": 4.940845070422536e-05, "loss": 0.699, "step": 1754, "task_loss": 0.6799315810203552 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.958435595035553, "epoch": 1.48, "learning_rate": 4.9436619718309864e-05, "loss": 0.7993, "step": 1755, "task_loss": 0.9463890790939331 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4931311309337616, "epoch": 1.48, "learning_rate": 4.946478873239437e-05, "loss": 0.5059, "step": 1756, "task_loss": 0.8569930791854858 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9269417524337769, "epoch": 1.48, "learning_rate": 4.949295774647887e-05, "loss": 0.7773, "step": 1757, "task_loss": 1.5328443050384521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6841145157814026, "epoch": 1.49, "learning_rate": 4.952112676056338e-05, "loss": 0.8293, "step": 1758, "task_loss": 1.2151539325714111 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4345683753490448, "epoch": 1.49, "learning_rate": 4.954929577464789e-05, "loss": 0.6384, "step": 1759, "task_loss": 0.5382832884788513 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5993403792381287, "epoch": 1.49, "learning_rate": 4.95774647887324e-05, "loss": 0.7904, "step": 1760, "task_loss": 2.0251951217651367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7951532006263733, "epoch": 1.49, "learning_rate": 4.96056338028169e-05, "loss": 0.6964, "step": 1761, "task_loss": 1.33260977268219 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6838171482086182, "epoch": 1.49, "learning_rate": 4.963380281690141e-05, "loss": 0.8582, "step": 1762, "task_loss": 0.6893925070762634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6183147430419922, "epoch": 1.49, "learning_rate": 4.966197183098592e-05, "loss": 0.6214, "step": 1763, "task_loss": 0.3137242794036865 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3793777823448181, "epoch": 1.49, "learning_rate": 4.969014084507043e-05, "loss": 0.5471, "step": 1764, "task_loss": 0.06447605043649673 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.24995794892311096, "epoch": 1.49, "learning_rate": 4.971830985915493e-05, "loss": 0.5572, "step": 1765, "task_loss": 0.17739669978618622 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4889288544654846, "epoch": 1.49, "learning_rate": 4.974647887323944e-05, "loss": 0.7477, "step": 1766, "task_loss": 0.4853813052177429 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.023239016532898, "epoch": 1.49, "learning_rate": 4.9774647887323944e-05, "loss": 0.7254, "step": 1767, "task_loss": 1.1800516843795776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.827078640460968, "epoch": 1.49, "learning_rate": 4.980281690140846e-05, "loss": 0.6807, "step": 1768, "task_loss": 0.9607373476028442 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0390146970748901, "epoch": 1.5, "learning_rate": 4.983098591549296e-05, "loss": 0.7093, "step": 1769, "task_loss": 0.941625714302063 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.746138870716095, "epoch": 1.5, "learning_rate": 4.9859154929577466e-05, "loss": 0.7918, "step": 1770, "task_loss": 1.6039825677871704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.629546046257019, "epoch": 1.5, "learning_rate": 4.9887323943661973e-05, "loss": 0.6581, "step": 1771, "task_loss": 0.6128548979759216 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.79425048828125, "epoch": 1.5, "learning_rate": 4.991549295774648e-05, "loss": 0.7481, "step": 1772, "task_loss": 1.4998985528945923 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5611071586608887, "epoch": 1.5, "learning_rate": 4.994366197183099e-05, "loss": 0.6589, "step": 1773, "task_loss": 0.9800018072128296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.704261064529419, "epoch": 1.5, "learning_rate": 4.9971830985915495e-05, "loss": 0.5492, "step": 1774, "task_loss": 1.3363583087921143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4659772217273712, "epoch": 1.5, "learning_rate": 5e-05, "loss": 0.3957, "step": 1775, "task_loss": 0.21420785784721375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7589311599731445, "epoch": 1.5, "learning_rate": 4.9996869129618034e-05, "loss": 0.7359, "step": 1776, "task_loss": 0.9890809655189514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5468038320541382, "epoch": 1.5, "learning_rate": 4.999373825923607e-05, "loss": 0.5934, "step": 1777, "task_loss": 1.0633552074432373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7291760444641113, "epoch": 1.5, "learning_rate": 4.9990607388854105e-05, "loss": 0.5676, "step": 1778, "task_loss": 0.40489307045936584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5862628221511841, "epoch": 1.5, "learning_rate": 4.9987476518472136e-05, "loss": 0.5201, "step": 1779, "task_loss": 0.2626260817050934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3920353949069977, "epoch": 1.5, "learning_rate": 4.998434564809017e-05, "loss": 0.6475, "step": 1780, "task_loss": 1.132348656654358 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.820918083190918, "epoch": 1.51, "learning_rate": 4.998121477770821e-05, "loss": 0.7581, "step": 1781, "task_loss": 0.9222025871276855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6620943546295166, "epoch": 1.51, "learning_rate": 4.997808390732624e-05, "loss": 0.6791, "step": 1782, "task_loss": 0.4763490557670593 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5768100023269653, "epoch": 1.51, "learning_rate": 4.997495303694427e-05, "loss": 0.6261, "step": 1783, "task_loss": 1.0265660285949707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4834175109863281, "epoch": 1.51, "learning_rate": 4.99718221665623e-05, "loss": 0.8234, "step": 1784, "task_loss": 1.556879997253418 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8264639377593994, "epoch": 1.51, "learning_rate": 4.996869129618034e-05, "loss": 0.8484, "step": 1785, "task_loss": 0.5273158550262451 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49966955184936523, "epoch": 1.51, "learning_rate": 4.996556042579837e-05, "loss": 0.564, "step": 1786, "task_loss": 0.9590334892272949 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5889289379119873, "epoch": 1.51, "learning_rate": 4.9962429555416404e-05, "loss": 0.6611, "step": 1787, "task_loss": 0.4693489968776703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7299202680587769, "epoch": 1.51, "learning_rate": 4.995929868503444e-05, "loss": 0.5524, "step": 1788, "task_loss": 1.431998610496521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6850101947784424, "epoch": 1.51, "learning_rate": 4.9956167814652474e-05, "loss": 0.6793, "step": 1789, "task_loss": 1.528543472290039 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4579393267631531, "epoch": 1.51, "learning_rate": 4.9953036944270506e-05, "loss": 0.5074, "step": 1790, "task_loss": 0.3179674744606018 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6380869150161743, "epoch": 1.51, "learning_rate": 4.9949906073888545e-05, "loss": 0.65, "step": 1791, "task_loss": 0.1815675050020218 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9062588214874268, "epoch": 1.51, "learning_rate": 4.9946775203506577e-05, "loss": 0.7858, "step": 1792, "task_loss": 1.2457330226898193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7081063985824585, "epoch": 1.52, "learning_rate": 4.9943644333124615e-05, "loss": 0.685, "step": 1793, "task_loss": 0.087989442050457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.218354269862175, "epoch": 1.52, "learning_rate": 4.994051346274265e-05, "loss": 0.4174, "step": 1794, "task_loss": 0.35041168332099915 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7589148283004761, "epoch": 1.52, "learning_rate": 4.993738259236068e-05, "loss": 0.8537, "step": 1795, "task_loss": 0.7867473363876343 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4449131488800049, "epoch": 1.52, "learning_rate": 4.993425172197872e-05, "loss": 0.5687, "step": 1796, "task_loss": 0.31527191400527954 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.488155722618103, "epoch": 1.52, "learning_rate": 4.993112085159675e-05, "loss": 0.7606, "step": 1797, "task_loss": 0.7264242768287659 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46588680148124695, "epoch": 1.52, "learning_rate": 4.992798998121478e-05, "loss": 0.681, "step": 1798, "task_loss": 1.511401891708374 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6547958850860596, "epoch": 1.52, "learning_rate": 4.992485911083281e-05, "loss": 0.7054, "step": 1799, "task_loss": 1.8270134925842285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4431312084197998, "epoch": 1.52, "learning_rate": 4.992172824045085e-05, "loss": 0.4505, "step": 1800, "task_loss": 0.2967238426208496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46827447414398193, "epoch": 1.52, "learning_rate": 4.991859737006888e-05, "loss": 0.5899, "step": 1801, "task_loss": 0.3745695650577545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8080461621284485, "epoch": 1.52, "learning_rate": 4.9915466499686915e-05, "loss": 0.6657, "step": 1802, "task_loss": 0.38682907819747925 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46412205696105957, "epoch": 1.52, "learning_rate": 4.991233562930495e-05, "loss": 0.668, "step": 1803, "task_loss": 1.6101449728012085 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4977133870124817, "epoch": 1.52, "learning_rate": 4.9909204758922985e-05, "loss": 0.6419, "step": 1804, "task_loss": 0.42898091673851013 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8670452833175659, "epoch": 1.53, "learning_rate": 4.990607388854102e-05, "loss": 0.6058, "step": 1805, "task_loss": 2.060584545135498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7038534283638, "epoch": 1.53, "learning_rate": 4.990294301815905e-05, "loss": 0.6232, "step": 1806, "task_loss": 1.1460820436477661 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43152838945388794, "epoch": 1.53, "learning_rate": 4.989981214777709e-05, "loss": 0.495, "step": 1807, "task_loss": 0.6902475953102112 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8095934391021729, "epoch": 1.53, "learning_rate": 4.989668127739512e-05, "loss": 0.6588, "step": 1808, "task_loss": 0.713253378868103 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5866916179656982, "epoch": 1.53, "learning_rate": 4.989355040701315e-05, "loss": 0.6411, "step": 1809, "task_loss": 1.5942386388778687 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4574407935142517, "epoch": 1.53, "learning_rate": 4.989041953663118e-05, "loss": 0.5922, "step": 1810, "task_loss": 0.25790825486183167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6200160980224609, "epoch": 1.53, "learning_rate": 4.988728866624922e-05, "loss": 0.7298, "step": 1811, "task_loss": 0.6222794055938721 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7348558902740479, "epoch": 1.53, "learning_rate": 4.988415779586725e-05, "loss": 0.8533, "step": 1812, "task_loss": 1.1796276569366455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7472292184829712, "epoch": 1.53, "learning_rate": 4.9881026925485285e-05, "loss": 0.7828, "step": 1813, "task_loss": 0.7248402237892151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6385415196418762, "epoch": 1.53, "learning_rate": 4.987789605510332e-05, "loss": 0.7694, "step": 1814, "task_loss": 0.7425999045372009 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.716332197189331, "epoch": 1.53, "learning_rate": 4.9874765184721355e-05, "loss": 0.7536, "step": 1815, "task_loss": 1.3053377866744995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5453732013702393, "epoch": 1.53, "learning_rate": 4.987163431433939e-05, "loss": 0.6461, "step": 1816, "task_loss": 1.1363106966018677 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5644437074661255, "epoch": 1.54, "learning_rate": 4.986850344395742e-05, "loss": 0.777, "step": 1817, "task_loss": 0.7945980429649353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7382049560546875, "epoch": 1.54, "learning_rate": 4.986537257357546e-05, "loss": 0.631, "step": 1818, "task_loss": 0.5975694060325623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44151127338409424, "epoch": 1.54, "learning_rate": 4.986224170319349e-05, "loss": 0.6433, "step": 1819, "task_loss": 0.8343964219093323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.722470760345459, "epoch": 1.54, "learning_rate": 4.985911083281152e-05, "loss": 0.6414, "step": 1820, "task_loss": 0.7751318216323853 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8326852917671204, "epoch": 1.54, "learning_rate": 4.985597996242955e-05, "loss": 0.6585, "step": 1821, "task_loss": 1.2828439474105835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7740627527236938, "epoch": 1.54, "learning_rate": 4.985284909204759e-05, "loss": 0.6751, "step": 1822, "task_loss": 0.6138177514076233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.537439227104187, "epoch": 1.54, "learning_rate": 4.984971822166562e-05, "loss": 0.7133, "step": 1823, "task_loss": 0.7864573001861572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4803377091884613, "epoch": 1.54, "learning_rate": 4.9846587351283655e-05, "loss": 0.6603, "step": 1824, "task_loss": 0.6513344049453735 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.460914671421051, "epoch": 1.54, "learning_rate": 4.984345648090169e-05, "loss": 0.5309, "step": 1825, "task_loss": 0.5043463706970215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4908214211463928, "epoch": 1.54, "learning_rate": 4.9840325610519725e-05, "loss": 0.5785, "step": 1826, "task_loss": 1.3462622165679932 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3570498824119568, "epoch": 1.54, "learning_rate": 4.9837194740137763e-05, "loss": 0.4514, "step": 1827, "task_loss": 0.26234447956085205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6105209589004517, "epoch": 1.54, "learning_rate": 4.9834063869755795e-05, "loss": 0.5727, "step": 1828, "task_loss": 1.4233118295669556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38166531920433044, "epoch": 1.55, "learning_rate": 4.9830932999373834e-05, "loss": 0.476, "step": 1829, "task_loss": 0.7092889547348022 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.608270525932312, "epoch": 1.55, "learning_rate": 4.9827802128991866e-05, "loss": 0.5901, "step": 1830, "task_loss": 1.1557390689849854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5318359136581421, "epoch": 1.55, "learning_rate": 4.98246712586099e-05, "loss": 0.7141, "step": 1831, "task_loss": 0.6606122851371765 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.571493923664093, "epoch": 1.55, "learning_rate": 4.982154038822793e-05, "loss": 0.7032, "step": 1832, "task_loss": 1.0071773529052734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4722786247730255, "epoch": 1.55, "learning_rate": 4.981840951784597e-05, "loss": 0.4888, "step": 1833, "task_loss": 0.9370937943458557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46872323751449585, "epoch": 1.55, "learning_rate": 4.9815278647464e-05, "loss": 0.5162, "step": 1834, "task_loss": 1.3404901027679443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7788828611373901, "epoch": 1.55, "learning_rate": 4.981214777708203e-05, "loss": 0.6133, "step": 1835, "task_loss": 0.6192160248756409 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9818916320800781, "epoch": 1.55, "learning_rate": 4.980901690670006e-05, "loss": 0.7321, "step": 1836, "task_loss": 1.2386287450790405 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5778181552886963, "epoch": 1.55, "learning_rate": 4.98058860363181e-05, "loss": 0.7182, "step": 1837, "task_loss": 0.32417237758636475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5331680774688721, "epoch": 1.55, "learning_rate": 4.980275516593613e-05, "loss": 0.6216, "step": 1838, "task_loss": 0.7085524797439575 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9092373251914978, "epoch": 1.55, "learning_rate": 4.9799624295554165e-05, "loss": 0.8151, "step": 1839, "task_loss": 0.9767289757728577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3825393617153168, "epoch": 1.56, "learning_rate": 4.9796493425172204e-05, "loss": 0.6358, "step": 1840, "task_loss": 1.0692232847213745 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7195698618888855, "epoch": 1.56, "learning_rate": 4.9793362554790235e-05, "loss": 0.7661, "step": 1841, "task_loss": 1.07028067111969 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5334628820419312, "epoch": 1.56, "learning_rate": 4.979023168440827e-05, "loss": 0.6391, "step": 1842, "task_loss": 1.2108945846557617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5728611946105957, "epoch": 1.56, "learning_rate": 4.97871008140263e-05, "loss": 0.6743, "step": 1843, "task_loss": 0.5248939990997314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8412822484970093, "epoch": 1.56, "learning_rate": 4.978396994364434e-05, "loss": 0.5886, "step": 1844, "task_loss": 0.5057944655418396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.784916877746582, "epoch": 1.56, "learning_rate": 4.978083907326237e-05, "loss": 0.7943, "step": 1845, "task_loss": 0.7054234147071838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6050792336463928, "epoch": 1.56, "learning_rate": 4.97777082028804e-05, "loss": 0.6653, "step": 1846, "task_loss": 0.743058979511261 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6797387599945068, "epoch": 1.56, "learning_rate": 4.977457733249843e-05, "loss": 0.5427, "step": 1847, "task_loss": 1.0078208446502686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.013283371925354, "epoch": 1.56, "learning_rate": 4.977144646211647e-05, "loss": 0.6383, "step": 1848, "task_loss": 1.0155601501464844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5210666656494141, "epoch": 1.56, "learning_rate": 4.97683155917345e-05, "loss": 0.526, "step": 1849, "task_loss": 1.4587905406951904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7302477359771729, "epoch": 1.56, "learning_rate": 4.9765184721352535e-05, "loss": 0.6605, "step": 1850, "task_loss": 0.9275100231170654 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8575241565704346, "epoch": 1.56, "learning_rate": 4.9762053850970574e-05, "loss": 0.6863, "step": 1851, "task_loss": 0.7441763281822205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6409611701965332, "epoch": 1.57, "learning_rate": 4.9758922980588605e-05, "loss": 0.568, "step": 1852, "task_loss": 0.8352595567703247 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6701617240905762, "epoch": 1.57, "learning_rate": 4.975579211020664e-05, "loss": 0.6826, "step": 1853, "task_loss": 1.1050704717636108 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5685806274414062, "epoch": 1.57, "learning_rate": 4.975266123982467e-05, "loss": 0.7567, "step": 1854, "task_loss": 0.7252973318099976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5129014253616333, "epoch": 1.57, "learning_rate": 4.974953036944271e-05, "loss": 0.6102, "step": 1855, "task_loss": 0.6139322519302368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.34985101222991943, "epoch": 1.57, "learning_rate": 4.974639949906074e-05, "loss": 0.6686, "step": 1856, "task_loss": 0.06662418693304062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46754616498947144, "epoch": 1.57, "learning_rate": 4.974326862867877e-05, "loss": 0.6294, "step": 1857, "task_loss": 0.3787352740764618 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7425840497016907, "epoch": 1.57, "learning_rate": 4.974013775829681e-05, "loss": 0.6034, "step": 1858, "task_loss": 0.6082272529602051 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8619138598442078, "epoch": 1.57, "learning_rate": 4.973700688791484e-05, "loss": 0.8147, "step": 1859, "task_loss": 1.5062973499298096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9642199873924255, "epoch": 1.57, "learning_rate": 4.973387601753288e-05, "loss": 0.8482, "step": 1860, "task_loss": 2.0624330043792725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37015020847320557, "epoch": 1.57, "learning_rate": 4.973074514715091e-05, "loss": 0.5978, "step": 1861, "task_loss": 0.23315726220607758 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8714903593063354, "epoch": 1.57, "learning_rate": 4.9727614276768944e-05, "loss": 0.5779, "step": 1862, "task_loss": 1.106894612312317 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46105077862739563, "epoch": 1.57, "learning_rate": 4.972448340638698e-05, "loss": 0.5495, "step": 1863, "task_loss": 0.33445456624031067 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6958836317062378, "epoch": 1.58, "learning_rate": 4.9721352536005014e-05, "loss": 0.651, "step": 1864, "task_loss": 1.3073015213012695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9411340951919556, "epoch": 1.58, "learning_rate": 4.9718221665623046e-05, "loss": 0.8908, "step": 1865, "task_loss": 1.8183186054229736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5974377393722534, "epoch": 1.58, "learning_rate": 4.9715090795241084e-05, "loss": 0.6495, "step": 1866, "task_loss": 0.40484994649887085 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.649840772151947, "epoch": 1.58, "learning_rate": 4.9711959924859116e-05, "loss": 0.8113, "step": 1867, "task_loss": 1.0594204664230347 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5160443782806396, "epoch": 1.58, "learning_rate": 4.970882905447715e-05, "loss": 0.6364, "step": 1868, "task_loss": 0.3267112374305725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49156510829925537, "epoch": 1.58, "learning_rate": 4.970569818409518e-05, "loss": 0.7124, "step": 1869, "task_loss": 0.967833399772644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5583096742630005, "epoch": 1.58, "learning_rate": 4.970256731371322e-05, "loss": 0.564, "step": 1870, "task_loss": 0.270713746547699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9654182195663452, "epoch": 1.58, "learning_rate": 4.969943644333125e-05, "loss": 0.8151, "step": 1871, "task_loss": 1.4135959148406982 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6948229074478149, "epoch": 1.58, "learning_rate": 4.969630557294928e-05, "loss": 0.6895, "step": 1872, "task_loss": 0.5128327012062073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5274141430854797, "epoch": 1.58, "learning_rate": 4.9693174702567313e-05, "loss": 0.5025, "step": 1873, "task_loss": 0.6782694458961487 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47294849157333374, "epoch": 1.58, "learning_rate": 4.969004383218535e-05, "loss": 0.6328, "step": 1874, "task_loss": 0.7425934076309204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9146326780319214, "epoch": 1.58, "learning_rate": 4.9686912961803384e-05, "loss": 0.746, "step": 1875, "task_loss": 0.7183727025985718 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4129660725593567, "epoch": 1.59, "learning_rate": 4.9683782091421416e-05, "loss": 0.6906, "step": 1876, "task_loss": 0.33734527230262756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.31612256169319153, "epoch": 1.59, "learning_rate": 4.9680651221039454e-05, "loss": 0.5292, "step": 1877, "task_loss": 0.473185658454895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3645424246788025, "epoch": 1.59, "learning_rate": 4.9677520350657486e-05, "loss": 0.7389, "step": 1878, "task_loss": 0.42236536741256714 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6450278162956238, "epoch": 1.59, "learning_rate": 4.967438948027552e-05, "loss": 0.531, "step": 1879, "task_loss": 0.502147376537323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8462682366371155, "epoch": 1.59, "learning_rate": 4.967125860989355e-05, "loss": 0.6969, "step": 1880, "task_loss": 1.1848641633987427 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7974748015403748, "epoch": 1.59, "learning_rate": 4.966812773951159e-05, "loss": 0.6025, "step": 1881, "task_loss": 0.9261400103569031 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5537360906600952, "epoch": 1.59, "learning_rate": 4.966499686912962e-05, "loss": 0.6897, "step": 1882, "task_loss": 0.1728622168302536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.611894428730011, "epoch": 1.59, "learning_rate": 4.966186599874765e-05, "loss": 0.6724, "step": 1883, "task_loss": 0.7432882785797119 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6963502764701843, "epoch": 1.59, "learning_rate": 4.9658735128365683e-05, "loss": 0.7866, "step": 1884, "task_loss": 0.7535934448242188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38894689083099365, "epoch": 1.59, "learning_rate": 4.965560425798372e-05, "loss": 0.6377, "step": 1885, "task_loss": 0.717616081237793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38068193197250366, "epoch": 1.59, "learning_rate": 4.9652473387601754e-05, "loss": 0.7113, "step": 1886, "task_loss": 0.6385071277618408 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6118330955505371, "epoch": 1.59, "learning_rate": 4.9649342517219786e-05, "loss": 0.6031, "step": 1887, "task_loss": 0.7858465313911438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7376946210861206, "epoch": 1.6, "learning_rate": 4.9646211646837824e-05, "loss": 0.7976, "step": 1888, "task_loss": 1.4276291131973267 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7895321249961853, "epoch": 1.6, "learning_rate": 4.9643080776455856e-05, "loss": 0.6158, "step": 1889, "task_loss": 0.6673716902732849 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6097396016120911, "epoch": 1.6, "learning_rate": 4.963994990607389e-05, "loss": 0.832, "step": 1890, "task_loss": 1.9537951946258545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6436938047409058, "epoch": 1.6, "learning_rate": 4.963681903569192e-05, "loss": 0.6908, "step": 1891, "task_loss": 0.9204204082489014 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5961222052574158, "epoch": 1.6, "learning_rate": 4.963368816530996e-05, "loss": 0.483, "step": 1892, "task_loss": 0.24744339287281036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7518224716186523, "epoch": 1.6, "learning_rate": 4.963055729492799e-05, "loss": 0.7209, "step": 1893, "task_loss": 1.5790899991989136 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5728371143341064, "epoch": 1.6, "learning_rate": 4.962742642454603e-05, "loss": 0.6284, "step": 1894, "task_loss": 0.5015727281570435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6766525506973267, "epoch": 1.6, "learning_rate": 4.962429555416406e-05, "loss": 0.5945, "step": 1895, "task_loss": 1.5083414316177368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7807190418243408, "epoch": 1.6, "learning_rate": 4.96211646837821e-05, "loss": 0.6959, "step": 1896, "task_loss": 1.3658125400543213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7218337059020996, "epoch": 1.6, "learning_rate": 4.961803381340013e-05, "loss": 0.6599, "step": 1897, "task_loss": 1.0971128940582275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.33474114537239075, "epoch": 1.6, "learning_rate": 4.961490294301816e-05, "loss": 0.4758, "step": 1898, "task_loss": 0.27137675881385803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.444968581199646, "epoch": 1.6, "learning_rate": 4.96117720726362e-05, "loss": 0.556, "step": 1899, "task_loss": 0.6097999215126038 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6328514218330383, "epoch": 1.61, "learning_rate": 4.960864120225423e-05, "loss": 0.5153, "step": 1900, "task_loss": 0.6395464539527893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6463197469711304, "epoch": 1.61, "learning_rate": 4.9605510331872264e-05, "loss": 0.6608, "step": 1901, "task_loss": 1.0098474025726318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3754156827926636, "epoch": 1.61, "learning_rate": 4.9602379461490296e-05, "loss": 0.6518, "step": 1902, "task_loss": 0.1963372528553009 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.431393027305603, "epoch": 1.61, "learning_rate": 4.9599248591108335e-05, "loss": 0.7657, "step": 1903, "task_loss": 0.7314988374710083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49419891834259033, "epoch": 1.61, "learning_rate": 4.9596117720726366e-05, "loss": 0.5716, "step": 1904, "task_loss": 0.8949008584022522 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4104131758213043, "epoch": 1.61, "learning_rate": 4.95929868503444e-05, "loss": 0.7443, "step": 1905, "task_loss": 0.7403879165649414 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.69712233543396, "epoch": 1.61, "learning_rate": 4.958985597996243e-05, "loss": 0.5786, "step": 1906, "task_loss": 1.1130852699279785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1562131643295288, "epoch": 1.61, "learning_rate": 4.958672510958047e-05, "loss": 0.8875, "step": 1907, "task_loss": 0.9969403743743896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8369267582893372, "epoch": 1.61, "learning_rate": 4.95835942391985e-05, "loss": 0.7038, "step": 1908, "task_loss": 0.8778485655784607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6184611320495605, "epoch": 1.61, "learning_rate": 4.958046336881653e-05, "loss": 0.8991, "step": 1909, "task_loss": 0.49932044744491577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.42341554164886475, "epoch": 1.61, "learning_rate": 4.957733249843457e-05, "loss": 0.6186, "step": 1910, "task_loss": 1.1440263986587524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5245881080627441, "epoch": 1.61, "learning_rate": 4.95742016280526e-05, "loss": 0.4534, "step": 1911, "task_loss": 1.2103773355484009 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7973243594169617, "epoch": 1.62, "learning_rate": 4.9571070757670634e-05, "loss": 0.7673, "step": 1912, "task_loss": 1.1234196424484253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6281956434249878, "epoch": 1.62, "learning_rate": 4.9567939887288666e-05, "loss": 0.76, "step": 1913, "task_loss": 0.7964915037155151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6769474744796753, "epoch": 1.62, "learning_rate": 4.9564809016906705e-05, "loss": 0.6881, "step": 1914, "task_loss": 0.5975040197372437 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5171736478805542, "epoch": 1.62, "learning_rate": 4.9561678146524736e-05, "loss": 0.5682, "step": 1915, "task_loss": 0.8132036328315735 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3539310097694397, "epoch": 1.62, "learning_rate": 4.955854727614277e-05, "loss": 0.6061, "step": 1916, "task_loss": 0.938351571559906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43796542286872864, "epoch": 1.62, "learning_rate": 4.95554164057608e-05, "loss": 0.4974, "step": 1917, "task_loss": 0.45349541306495667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6885635256767273, "epoch": 1.62, "learning_rate": 4.955228553537884e-05, "loss": 0.6672, "step": 1918, "task_loss": 0.7289242744445801 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5212783217430115, "epoch": 1.62, "learning_rate": 4.954915466499687e-05, "loss": 0.6399, "step": 1919, "task_loss": 0.5375036597251892 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.954554557800293, "epoch": 1.62, "learning_rate": 4.95460237946149e-05, "loss": 0.7478, "step": 1920, "task_loss": 0.8465198874473572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8649784326553345, "epoch": 1.62, "learning_rate": 4.9542892924232934e-05, "loss": 0.7012, "step": 1921, "task_loss": 2.0679683685302734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5553555488586426, "epoch": 1.62, "learning_rate": 4.953976205385097e-05, "loss": 0.6321, "step": 1922, "task_loss": 0.8277266025543213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38108664751052856, "epoch": 1.63, "learning_rate": 4.9536631183469004e-05, "loss": 0.6543, "step": 1923, "task_loss": 0.2204054743051529 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6473585963249207, "epoch": 1.63, "learning_rate": 4.9533500313087036e-05, "loss": 0.4589, "step": 1924, "task_loss": 0.8858429193496704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5503647923469543, "epoch": 1.63, "learning_rate": 4.9530369442705075e-05, "loss": 0.6157, "step": 1925, "task_loss": 1.1575253009796143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6086478233337402, "epoch": 1.63, "learning_rate": 4.9527238572323106e-05, "loss": 0.5778, "step": 1926, "task_loss": 0.7648385167121887 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.35494664311408997, "epoch": 1.63, "learning_rate": 4.9524107701941145e-05, "loss": 0.525, "step": 1927, "task_loss": 0.44571173191070557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6289347410202026, "epoch": 1.63, "learning_rate": 4.952097683155918e-05, "loss": 0.6942, "step": 1928, "task_loss": 2.0364298820495605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7279799580574036, "epoch": 1.63, "learning_rate": 4.951784596117721e-05, "loss": 0.7336, "step": 1929, "task_loss": 0.38676583766937256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7602465152740479, "epoch": 1.63, "learning_rate": 4.951471509079525e-05, "loss": 0.7469, "step": 1930, "task_loss": 0.5870030522346497 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6314072608947754, "epoch": 1.63, "learning_rate": 4.951158422041328e-05, "loss": 0.7256, "step": 1931, "task_loss": 2.087371349334717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7365673780441284, "epoch": 1.63, "learning_rate": 4.950845335003131e-05, "loss": 0.6645, "step": 1932, "task_loss": 0.5904118418693542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6646330952644348, "epoch": 1.63, "learning_rate": 4.950532247964935e-05, "loss": 0.615, "step": 1933, "task_loss": 1.0293835401535034 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41148483753204346, "epoch": 1.63, "learning_rate": 4.950219160926738e-05, "loss": 0.6611, "step": 1934, "task_loss": 0.23426881432533264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47328197956085205, "epoch": 1.64, "learning_rate": 4.949906073888541e-05, "loss": 0.6199, "step": 1935, "task_loss": 1.0920230150222778 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6245715618133545, "epoch": 1.64, "learning_rate": 4.949592986850345e-05, "loss": 0.5839, "step": 1936, "task_loss": 0.3793618977069855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5313767194747925, "epoch": 1.64, "learning_rate": 4.949279899812148e-05, "loss": 0.5664, "step": 1937, "task_loss": 0.4905346930027008 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40442922711372375, "epoch": 1.64, "learning_rate": 4.9489668127739515e-05, "loss": 0.6156, "step": 1938, "task_loss": 0.7184382081031799 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0248744487762451, "epoch": 1.64, "learning_rate": 4.9486537257357547e-05, "loss": 0.7475, "step": 1939, "task_loss": 1.445002555847168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49010708928108215, "epoch": 1.64, "learning_rate": 4.9483406386975585e-05, "loss": 0.5469, "step": 1940, "task_loss": 0.7649798393249512 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40869268774986267, "epoch": 1.64, "learning_rate": 4.948027551659362e-05, "loss": 0.5429, "step": 1941, "task_loss": 0.6507436633110046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5798522233963013, "epoch": 1.64, "learning_rate": 4.947714464621165e-05, "loss": 0.6123, "step": 1942, "task_loss": 0.7554382681846619 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5164743661880493, "epoch": 1.64, "learning_rate": 4.947401377582968e-05, "loss": 0.7167, "step": 1943, "task_loss": 0.7159159779548645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6421444416046143, "epoch": 1.64, "learning_rate": 4.947088290544772e-05, "loss": 0.5765, "step": 1944, "task_loss": 1.0378459692001343 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49714013934135437, "epoch": 1.64, "learning_rate": 4.946775203506575e-05, "loss": 0.7207, "step": 1945, "task_loss": 0.6052864193916321 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6423880457878113, "epoch": 1.64, "learning_rate": 4.946462116468378e-05, "loss": 0.6891, "step": 1946, "task_loss": 1.6096755266189575 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4944945275783539, "epoch": 1.65, "learning_rate": 4.946149029430182e-05, "loss": 0.5347, "step": 1947, "task_loss": 0.4242657423019409 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8928776383399963, "epoch": 1.65, "learning_rate": 4.945835942391985e-05, "loss": 0.9216, "step": 1948, "task_loss": 1.4361083507537842 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9358099699020386, "epoch": 1.65, "learning_rate": 4.9455228553537885e-05, "loss": 0.6402, "step": 1949, "task_loss": 1.248407006263733 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4649122357368469, "epoch": 1.65, "learning_rate": 4.9452097683155916e-05, "loss": 0.675, "step": 1950, "task_loss": 0.577364981174469 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3630506992340088, "epoch": 1.65, "learning_rate": 4.9448966812773955e-05, "loss": 0.6125, "step": 1951, "task_loss": 1.114147663116455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5133869647979736, "epoch": 1.65, "learning_rate": 4.944583594239199e-05, "loss": 0.5935, "step": 1952, "task_loss": 0.3679982125759125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7500715255737305, "epoch": 1.65, "learning_rate": 4.944270507201002e-05, "loss": 0.6887, "step": 1953, "task_loss": 0.543660044670105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8992449641227722, "epoch": 1.65, "learning_rate": 4.943957420162805e-05, "loss": 0.7484, "step": 1954, "task_loss": 0.832017183303833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5735136866569519, "epoch": 1.65, "learning_rate": 4.943644333124609e-05, "loss": 0.7995, "step": 1955, "task_loss": 1.1599526405334473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6302834749221802, "epoch": 1.65, "learning_rate": 4.943331246086412e-05, "loss": 0.6661, "step": 1956, "task_loss": 0.3010315001010895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6885948181152344, "epoch": 1.65, "learning_rate": 4.943018159048215e-05, "loss": 0.6881, "step": 1957, "task_loss": 0.7518966794013977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.755750298500061, "epoch": 1.65, "learning_rate": 4.9427050720100184e-05, "loss": 0.7154, "step": 1958, "task_loss": 0.6935897469520569 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.36743319034576416, "epoch": 1.66, "learning_rate": 4.942391984971822e-05, "loss": 0.5368, "step": 1959, "task_loss": 0.5644461512565613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7725592255592346, "epoch": 1.66, "learning_rate": 4.9420788979336255e-05, "loss": 0.6846, "step": 1960, "task_loss": 0.41620486974716187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.39005398750305176, "epoch": 1.66, "learning_rate": 4.941765810895429e-05, "loss": 0.4699, "step": 1961, "task_loss": 0.5392753481864929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48912060260772705, "epoch": 1.66, "learning_rate": 4.9414527238572325e-05, "loss": 0.5679, "step": 1962, "task_loss": 0.20118527114391327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7551258206367493, "epoch": 1.66, "learning_rate": 4.9411396368190364e-05, "loss": 0.5452, "step": 1963, "task_loss": 0.14876854419708252 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7719982266426086, "epoch": 1.66, "learning_rate": 4.9408265497808395e-05, "loss": 0.7333, "step": 1964, "task_loss": 0.9910011291503906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6281237602233887, "epoch": 1.66, "learning_rate": 4.940513462742643e-05, "loss": 0.4845, "step": 1965, "task_loss": 0.7331114411354065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.723662257194519, "epoch": 1.66, "learning_rate": 4.9402003757044466e-05, "loss": 0.7275, "step": 1966, "task_loss": 1.0949897766113281 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7545567750930786, "epoch": 1.66, "learning_rate": 4.93988728866625e-05, "loss": 0.5968, "step": 1967, "task_loss": 1.1954057216644287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4971545338630676, "epoch": 1.66, "learning_rate": 4.939574201628053e-05, "loss": 0.5143, "step": 1968, "task_loss": 0.9052865505218506 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6326043009757996, "epoch": 1.66, "learning_rate": 4.939261114589856e-05, "loss": 0.4374, "step": 1969, "task_loss": 0.5534011721611023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4748438894748688, "epoch": 1.66, "learning_rate": 4.93894802755166e-05, "loss": 0.6876, "step": 1970, "task_loss": 0.4051666855812073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3269922733306885, "epoch": 1.67, "learning_rate": 4.938634940513463e-05, "loss": 0.5996, "step": 1971, "task_loss": 1.2984964847564697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4338199198246002, "epoch": 1.67, "learning_rate": 4.938321853475266e-05, "loss": 0.4849, "step": 1972, "task_loss": 0.7718173861503601 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7763398885726929, "epoch": 1.67, "learning_rate": 4.93800876643707e-05, "loss": 0.6696, "step": 1973, "task_loss": 0.9196620583534241 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8411092758178711, "epoch": 1.67, "learning_rate": 4.9376956793988733e-05, "loss": 0.7312, "step": 1974, "task_loss": 1.2120000123977661 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9371830224990845, "epoch": 1.67, "learning_rate": 4.9373825923606765e-05, "loss": 0.5641, "step": 1975, "task_loss": 1.0051378011703491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5774310231208801, "epoch": 1.67, "learning_rate": 4.93706950532248e-05, "loss": 0.68, "step": 1976, "task_loss": 1.1707119941711426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6136159896850586, "epoch": 1.67, "learning_rate": 4.9367564182842836e-05, "loss": 0.6724, "step": 1977, "task_loss": 1.3529040813446045 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4807935953140259, "epoch": 1.67, "learning_rate": 4.936443331246087e-05, "loss": 0.7414, "step": 1978, "task_loss": 0.992839515209198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6059003472328186, "epoch": 1.67, "learning_rate": 4.93613024420789e-05, "loss": 0.5897, "step": 1979, "task_loss": 1.304392695426941 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41444510221481323, "epoch": 1.67, "learning_rate": 4.935817157169693e-05, "loss": 0.5583, "step": 1980, "task_loss": 0.3821897804737091 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5459423065185547, "epoch": 1.67, "learning_rate": 4.935504070131497e-05, "loss": 0.5173, "step": 1981, "task_loss": 0.7471437454223633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47253596782684326, "epoch": 1.67, "learning_rate": 4.9351909830933e-05, "loss": 0.4923, "step": 1982, "task_loss": 1.1227154731750488 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3687297999858856, "epoch": 1.68, "learning_rate": 4.934877896055103e-05, "loss": 0.5459, "step": 1983, "task_loss": 0.5791575908660889 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7077852487564087, "epoch": 1.68, "learning_rate": 4.934564809016907e-05, "loss": 0.6304, "step": 1984, "task_loss": 0.5971851348876953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5832208395004272, "epoch": 1.68, "learning_rate": 4.93425172197871e-05, "loss": 0.6258, "step": 1985, "task_loss": 0.4751685857772827 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3852139413356781, "epoch": 1.68, "learning_rate": 4.9339386349405135e-05, "loss": 0.5782, "step": 1986, "task_loss": 0.7024917602539062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7285100221633911, "epoch": 1.68, "learning_rate": 4.933625547902317e-05, "loss": 0.7267, "step": 1987, "task_loss": 1.963628888130188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6865188479423523, "epoch": 1.68, "learning_rate": 4.9333124608641205e-05, "loss": 0.7319, "step": 1988, "task_loss": 1.2825442552566528 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1501373052597046, "epoch": 1.68, "learning_rate": 4.932999373825924e-05, "loss": 0.6868, "step": 1989, "task_loss": 1.0284918546676636 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8097833395004272, "epoch": 1.68, "learning_rate": 4.932686286787727e-05, "loss": 0.6955, "step": 1990, "task_loss": 0.964862048625946 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.30223289132118225, "epoch": 1.68, "learning_rate": 4.93237319974953e-05, "loss": 0.4465, "step": 1991, "task_loss": 0.667451024055481 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.550812840461731, "epoch": 1.68, "learning_rate": 4.932060112711334e-05, "loss": 0.505, "step": 1992, "task_loss": 0.5724264979362488 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3076070547103882, "epoch": 1.68, "learning_rate": 4.931747025673137e-05, "loss": 0.5195, "step": 1993, "task_loss": 0.7102048993110657 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4364529848098755, "epoch": 1.69, "learning_rate": 4.93143393863494e-05, "loss": 0.4866, "step": 1994, "task_loss": 0.9032068252563477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7616747617721558, "epoch": 1.69, "learning_rate": 4.931120851596744e-05, "loss": 0.5937, "step": 1995, "task_loss": 0.4781607985496521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48160672187805176, "epoch": 1.69, "learning_rate": 4.930807764558547e-05, "loss": 0.5314, "step": 1996, "task_loss": 0.5243744850158691 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.506239116191864, "epoch": 1.69, "learning_rate": 4.930494677520351e-05, "loss": 0.6673, "step": 1997, "task_loss": 0.7698559761047363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1524841785430908, "epoch": 1.69, "learning_rate": 4.9301815904821544e-05, "loss": 0.71, "step": 1998, "task_loss": 1.2391326427459717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.36105772852897644, "epoch": 1.69, "learning_rate": 4.929868503443958e-05, "loss": 0.4396, "step": 1999, "task_loss": 0.4166907072067261 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4137156009674072, "epoch": 1.69, "learning_rate": 4.9295554164057614e-05, "loss": 0.5777, "step": 2000, "task_loss": 0.1113438829779625 }, { "epoch": 1.69, "eval_accuracy": 0.9006336633663367, "eval_loss": 0.36640891432762146, "eval_runtime": 208.1301, "eval_samples_per_second": 121.318, "eval_steps_per_second": 0.951, "step": 2000 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7018259763717651, "epoch": 1.69, "learning_rate": 4.9292423293675646e-05, "loss": 0.5719, "step": 2001, "task_loss": 0.2629460096359253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5679777264595032, "epoch": 1.69, "learning_rate": 4.928929242329368e-05, "loss": 0.579, "step": 2002, "task_loss": 0.3460458815097809 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43855440616607666, "epoch": 1.69, "learning_rate": 4.9286161552911716e-05, "loss": 0.6825, "step": 2003, "task_loss": 0.930694580078125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7002033591270447, "epoch": 1.69, "learning_rate": 4.928303068252975e-05, "loss": 0.5247, "step": 2004, "task_loss": 0.5329486727714539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.646618664264679, "epoch": 1.69, "learning_rate": 4.927989981214778e-05, "loss": 0.6485, "step": 2005, "task_loss": 0.4069245159626007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0145593881607056, "epoch": 1.7, "learning_rate": 4.927676894176581e-05, "loss": 0.882, "step": 2006, "task_loss": 2.492241859436035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.42476505041122437, "epoch": 1.7, "learning_rate": 4.927363807138385e-05, "loss": 0.5209, "step": 2007, "task_loss": 0.13505129516124725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1062614917755127, "epoch": 1.7, "learning_rate": 4.927050720100188e-05, "loss": 0.6577, "step": 2008, "task_loss": 1.6635442972183228 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7027348279953003, "epoch": 1.7, "learning_rate": 4.9267376330619914e-05, "loss": 0.6095, "step": 2009, "task_loss": 0.3977372348308563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3768814206123352, "epoch": 1.7, "learning_rate": 4.926424546023795e-05, "loss": 0.713, "step": 2010, "task_loss": 1.0592989921569824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5463868975639343, "epoch": 1.7, "learning_rate": 4.9261114589855984e-05, "loss": 0.6438, "step": 2011, "task_loss": 0.650419294834137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.42588698863983154, "epoch": 1.7, "learning_rate": 4.9257983719474016e-05, "loss": 0.4957, "step": 2012, "task_loss": 0.4630742073059082 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.685448169708252, "epoch": 1.7, "learning_rate": 4.925485284909205e-05, "loss": 0.5833, "step": 2013, "task_loss": 1.310046911239624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.39456626772880554, "epoch": 1.7, "learning_rate": 4.9251721978710086e-05, "loss": 0.4542, "step": 2014, "task_loss": 0.47526127099990845 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7558485269546509, "epoch": 1.7, "learning_rate": 4.924859110832812e-05, "loss": 0.6641, "step": 2015, "task_loss": 0.5466424226760864 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5756458640098572, "epoch": 1.7, "learning_rate": 4.924546023794615e-05, "loss": 0.6041, "step": 2016, "task_loss": 0.9603175520896912 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.722676694393158, "epoch": 1.7, "learning_rate": 4.924232936756418e-05, "loss": 0.742, "step": 2017, "task_loss": 2.0924301147460938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.691618800163269, "epoch": 1.71, "learning_rate": 4.923919849718222e-05, "loss": 0.5845, "step": 2018, "task_loss": 1.0205845832824707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4835856258869171, "epoch": 1.71, "learning_rate": 4.923606762680025e-05, "loss": 0.5819, "step": 2019, "task_loss": 0.638346791267395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5968512296676636, "epoch": 1.71, "learning_rate": 4.9232936756418283e-05, "loss": 0.6626, "step": 2020, "task_loss": 0.9660597443580627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.50050950050354, "epoch": 1.71, "learning_rate": 4.922980588603632e-05, "loss": 0.688, "step": 2021, "task_loss": 0.673224687576294 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.89590984582901, "epoch": 1.71, "learning_rate": 4.9226675015654354e-05, "loss": 0.6938, "step": 2022, "task_loss": 0.36511263251304626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4525848627090454, "epoch": 1.71, "learning_rate": 4.9223544145272386e-05, "loss": 0.4711, "step": 2023, "task_loss": 2.65280818939209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5653413534164429, "epoch": 1.71, "learning_rate": 4.922041327489042e-05, "loss": 0.6813, "step": 2024, "task_loss": 0.7345885634422302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.580802321434021, "epoch": 1.71, "learning_rate": 4.9217282404508456e-05, "loss": 0.778, "step": 2025, "task_loss": 1.44764244556427 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4792664051055908, "epoch": 1.71, "learning_rate": 4.921415153412649e-05, "loss": 0.4886, "step": 2026, "task_loss": 0.6550425887107849 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6716933250427246, "epoch": 1.71, "learning_rate": 4.921102066374452e-05, "loss": 0.6739, "step": 2027, "task_loss": 1.2890682220458984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45043429732322693, "epoch": 1.71, "learning_rate": 4.920788979336256e-05, "loss": 0.5488, "step": 2028, "task_loss": 0.41126132011413574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.577710747718811, "epoch": 1.71, "learning_rate": 4.920475892298059e-05, "loss": 0.6201, "step": 2029, "task_loss": 0.4359762966632843 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48747503757476807, "epoch": 1.72, "learning_rate": 4.920162805259863e-05, "loss": 0.6496, "step": 2030, "task_loss": 0.7561209201812744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8348122835159302, "epoch": 1.72, "learning_rate": 4.919849718221666e-05, "loss": 0.7238, "step": 2031, "task_loss": 0.9416124224662781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4957389831542969, "epoch": 1.72, "learning_rate": 4.919536631183469e-05, "loss": 0.5096, "step": 2032, "task_loss": 0.297063946723938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4176046848297119, "epoch": 1.72, "learning_rate": 4.919223544145273e-05, "loss": 0.4662, "step": 2033, "task_loss": 1.4843019247055054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48003995418548584, "epoch": 1.72, "learning_rate": 4.918910457107076e-05, "loss": 0.623, "step": 2034, "task_loss": 0.7919873595237732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.557610034942627, "epoch": 1.72, "learning_rate": 4.9185973700688794e-05, "loss": 0.5895, "step": 2035, "task_loss": 0.7549609541893005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7372152805328369, "epoch": 1.72, "learning_rate": 4.918284283030683e-05, "loss": 0.7089, "step": 2036, "task_loss": 1.0130817890167236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6063181161880493, "epoch": 1.72, "learning_rate": 4.9179711959924864e-05, "loss": 0.6167, "step": 2037, "task_loss": 1.4348359107971191 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8438919186592102, "epoch": 1.72, "learning_rate": 4.9176581089542896e-05, "loss": 0.5514, "step": 2038, "task_loss": 0.9761325120925903 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5961563587188721, "epoch": 1.72, "learning_rate": 4.917345021916093e-05, "loss": 0.6892, "step": 2039, "task_loss": 0.5040318369865417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6199002265930176, "epoch": 1.72, "learning_rate": 4.9170319348778967e-05, "loss": 0.6097, "step": 2040, "task_loss": 1.1280938386917114 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5494731068611145, "epoch": 1.72, "learning_rate": 4.9167188478397e-05, "loss": 0.5492, "step": 2041, "task_loss": 0.7358270883560181 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7027154564857483, "epoch": 1.73, "learning_rate": 4.916405760801503e-05, "loss": 0.7005, "step": 2042, "task_loss": 0.47275885939598083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4182558059692383, "epoch": 1.73, "learning_rate": 4.916092673763306e-05, "loss": 0.4456, "step": 2043, "task_loss": 0.9683384895324707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.34182876348495483, "epoch": 1.73, "learning_rate": 4.91577958672511e-05, "loss": 0.6713, "step": 2044, "task_loss": 0.5590824484825134 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.34680500626564026, "epoch": 1.73, "learning_rate": 4.915466499686913e-05, "loss": 0.5232, "step": 2045, "task_loss": 1.5110278129577637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2685874104499817, "epoch": 1.73, "learning_rate": 4.9151534126487164e-05, "loss": 0.5065, "step": 2046, "task_loss": 0.3219241797924042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4738891124725342, "epoch": 1.73, "learning_rate": 4.91484032561052e-05, "loss": 0.6959, "step": 2047, "task_loss": 1.094277024269104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.34121495485305786, "epoch": 1.73, "learning_rate": 4.9145272385723234e-05, "loss": 0.5441, "step": 2048, "task_loss": 0.4475323259830475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5802171230316162, "epoch": 1.73, "learning_rate": 4.9142141515341266e-05, "loss": 0.5542, "step": 2049, "task_loss": 1.0223356485366821 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6329468488693237, "epoch": 1.73, "learning_rate": 4.91390106449593e-05, "loss": 0.7662, "step": 2050, "task_loss": 0.5947673916816711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4295281767845154, "epoch": 1.73, "learning_rate": 4.9135879774577336e-05, "loss": 0.6475, "step": 2051, "task_loss": 0.310517281293869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38155925273895264, "epoch": 1.73, "learning_rate": 4.913274890419537e-05, "loss": 0.615, "step": 2052, "task_loss": 0.8374943733215332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5222368836402893, "epoch": 1.73, "learning_rate": 4.91296180338134e-05, "loss": 0.4413, "step": 2053, "task_loss": 0.7955223321914673 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3367731273174286, "epoch": 1.74, "learning_rate": 4.912648716343143e-05, "loss": 0.5884, "step": 2054, "task_loss": 0.20172403752803802 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7912696599960327, "epoch": 1.74, "learning_rate": 4.912335629304947e-05, "loss": 0.5577, "step": 2055, "task_loss": 1.4537066221237183 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6863619685173035, "epoch": 1.74, "learning_rate": 4.91202254226675e-05, "loss": 0.6425, "step": 2056, "task_loss": 0.688111424446106 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5390403866767883, "epoch": 1.74, "learning_rate": 4.9117094552285534e-05, "loss": 0.6213, "step": 2057, "task_loss": 1.159805178642273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5281963348388672, "epoch": 1.74, "learning_rate": 4.911396368190357e-05, "loss": 0.4335, "step": 2058, "task_loss": 0.9060367345809937 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5215946435928345, "epoch": 1.74, "learning_rate": 4.9110832811521604e-05, "loss": 0.4732, "step": 2059, "task_loss": 1.5111451148986816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.42220818996429443, "epoch": 1.74, "learning_rate": 4.9107701941139636e-05, "loss": 0.5739, "step": 2060, "task_loss": 0.1371811330318451 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3609325885772705, "epoch": 1.74, "learning_rate": 4.910457107075767e-05, "loss": 0.5775, "step": 2061, "task_loss": 0.060314346104860306 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6961934566497803, "epoch": 1.74, "learning_rate": 4.9101440200375706e-05, "loss": 0.5225, "step": 2062, "task_loss": 0.9735041856765747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.27555352449417114, "epoch": 1.74, "learning_rate": 4.909830932999374e-05, "loss": 0.3439, "step": 2063, "task_loss": 0.03225020319223404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5072891712188721, "epoch": 1.74, "learning_rate": 4.909517845961178e-05, "loss": 0.4552, "step": 2064, "task_loss": 1.0870237350463867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.707290768623352, "epoch": 1.75, "learning_rate": 4.909204758922981e-05, "loss": 0.5853, "step": 2065, "task_loss": 0.6081264019012451 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.4060214757919312, "epoch": 1.75, "learning_rate": 4.908891671884785e-05, "loss": 1.1267, "step": 2066, "task_loss": 1.127398133277893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8365243673324585, "epoch": 1.75, "learning_rate": 4.908578584846588e-05, "loss": 0.6841, "step": 2067, "task_loss": 1.1914780139923096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.1417372226715088, "epoch": 1.75, "learning_rate": 4.908265497808391e-05, "loss": 0.8157, "step": 2068, "task_loss": 0.6716783046722412 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4661165177822113, "epoch": 1.75, "learning_rate": 4.907952410770194e-05, "loss": 0.5928, "step": 2069, "task_loss": 0.8545433878898621 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48185107111930847, "epoch": 1.75, "learning_rate": 4.907639323731998e-05, "loss": 0.6948, "step": 2070, "task_loss": 0.05954224616289139 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4463898837566376, "epoch": 1.75, "learning_rate": 4.907326236693801e-05, "loss": 0.5184, "step": 2071, "task_loss": 0.23890015482902527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7773299813270569, "epoch": 1.75, "learning_rate": 4.9070131496556045e-05, "loss": 0.7219, "step": 2072, "task_loss": 0.23574751615524292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5636759996414185, "epoch": 1.75, "learning_rate": 4.906700062617408e-05, "loss": 0.7084, "step": 2073, "task_loss": 0.4014042615890503 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4411720633506775, "epoch": 1.75, "learning_rate": 4.9063869755792115e-05, "loss": 0.521, "step": 2074, "task_loss": 0.5234248638153076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4884949028491974, "epoch": 1.75, "learning_rate": 4.906073888541015e-05, "loss": 0.5953, "step": 2075, "task_loss": 0.9362673759460449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4500921368598938, "epoch": 1.75, "learning_rate": 4.905760801502818e-05, "loss": 0.4188, "step": 2076, "task_loss": 0.3185476064682007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5606488585472107, "epoch": 1.76, "learning_rate": 4.905447714464622e-05, "loss": 0.5489, "step": 2077, "task_loss": 0.14564985036849976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4630191922187805, "epoch": 1.76, "learning_rate": 4.905134627426425e-05, "loss": 0.5245, "step": 2078, "task_loss": 0.577075183391571 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3914957046508789, "epoch": 1.76, "learning_rate": 4.904821540388228e-05, "loss": 0.586, "step": 2079, "task_loss": 0.21617145836353302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4646104574203491, "epoch": 1.76, "learning_rate": 4.904508453350031e-05, "loss": 0.6525, "step": 2080, "task_loss": 0.9521306753158569 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5909013748168945, "epoch": 1.76, "learning_rate": 4.904195366311835e-05, "loss": 0.7381, "step": 2081, "task_loss": 0.7526659369468689 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8249051570892334, "epoch": 1.76, "learning_rate": 4.903882279273638e-05, "loss": 0.7484, "step": 2082, "task_loss": 0.7452606558799744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.596991777420044, "epoch": 1.76, "learning_rate": 4.9035691922354414e-05, "loss": 0.517, "step": 2083, "task_loss": 2.459826707839966 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3626992106437683, "epoch": 1.76, "learning_rate": 4.903256105197245e-05, "loss": 0.4604, "step": 2084, "task_loss": 0.11691419780254364 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41758930683135986, "epoch": 1.76, "learning_rate": 4.9029430181590485e-05, "loss": 0.5549, "step": 2085, "task_loss": 2.065662145614624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3848503828048706, "epoch": 1.76, "learning_rate": 4.9026299311208517e-05, "loss": 0.4655, "step": 2086, "task_loss": 0.5120450854301453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2197413444519043, "epoch": 1.76, "learning_rate": 4.902316844082655e-05, "loss": 0.4522, "step": 2087, "task_loss": 0.30466586351394653 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7243236899375916, "epoch": 1.76, "learning_rate": 4.902003757044459e-05, "loss": 0.5933, "step": 2088, "task_loss": 0.3135186731815338 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41405266523361206, "epoch": 1.77, "learning_rate": 4.901690670006262e-05, "loss": 0.5895, "step": 2089, "task_loss": 1.3180618286132812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5005056262016296, "epoch": 1.77, "learning_rate": 4.901377582968065e-05, "loss": 0.6134, "step": 2090, "task_loss": 0.7872045636177063 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4454622268676758, "epoch": 1.77, "learning_rate": 4.901064495929868e-05, "loss": 0.6777, "step": 2091, "task_loss": 0.7301774621009827 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.33388280868530273, "epoch": 1.77, "learning_rate": 4.900751408891672e-05, "loss": 0.5627, "step": 2092, "task_loss": 0.5217288136482239 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.494853675365448, "epoch": 1.77, "learning_rate": 4.900438321853475e-05, "loss": 0.6132, "step": 2093, "task_loss": 1.1620210409164429 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6449175477027893, "epoch": 1.77, "learning_rate": 4.9001252348152784e-05, "loss": 0.6142, "step": 2094, "task_loss": 0.8698065280914307 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5967389941215515, "epoch": 1.77, "learning_rate": 4.899812147777082e-05, "loss": 0.4615, "step": 2095, "task_loss": 0.9925934076309204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4799313247203827, "epoch": 1.77, "learning_rate": 4.8994990607388855e-05, "loss": 0.528, "step": 2096, "task_loss": 0.9510855674743652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7690541744232178, "epoch": 1.77, "learning_rate": 4.899185973700689e-05, "loss": 0.7087, "step": 2097, "task_loss": 1.3609060049057007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3533988296985626, "epoch": 1.77, "learning_rate": 4.8988728866624925e-05, "loss": 0.5156, "step": 2098, "task_loss": 0.8872549533843994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.480049192905426, "epoch": 1.77, "learning_rate": 4.898559799624296e-05, "loss": 0.6161, "step": 2099, "task_loss": 0.7815600633621216 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4116857051849365, "epoch": 1.77, "learning_rate": 4.8982467125860995e-05, "loss": 0.486, "step": 2100, "task_loss": 0.3901171088218689 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4409099817276001, "epoch": 1.78, "learning_rate": 4.897933625547903e-05, "loss": 0.5123, "step": 2101, "task_loss": 0.48921525478363037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8763812184333801, "epoch": 1.78, "learning_rate": 4.897620538509706e-05, "loss": 0.7333, "step": 2102, "task_loss": 1.0154145956039429 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37436163425445557, "epoch": 1.78, "learning_rate": 4.89730745147151e-05, "loss": 0.5683, "step": 2103, "task_loss": 0.41537272930145264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5571658611297607, "epoch": 1.78, "learning_rate": 4.896994364433313e-05, "loss": 0.5434, "step": 2104, "task_loss": 1.0850716829299927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.2591012716293335, "epoch": 1.78, "learning_rate": 4.896681277395116e-05, "loss": 0.8047, "step": 2105, "task_loss": 1.1168330907821655 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.689893364906311, "epoch": 1.78, "learning_rate": 4.896368190356919e-05, "loss": 0.6425, "step": 2106, "task_loss": 1.3808951377868652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7274289131164551, "epoch": 1.78, "learning_rate": 4.896055103318723e-05, "loss": 0.8153, "step": 2107, "task_loss": 1.0645067691802979 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5709022283554077, "epoch": 1.78, "learning_rate": 4.895742016280526e-05, "loss": 0.5475, "step": 2108, "task_loss": 0.9172461032867432 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8034793138504028, "epoch": 1.78, "learning_rate": 4.8954289292423295e-05, "loss": 0.69, "step": 2109, "task_loss": 1.639829158782959 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3763461709022522, "epoch": 1.78, "learning_rate": 4.8951158422041334e-05, "loss": 0.5466, "step": 2110, "task_loss": 0.5938164591789246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5774704217910767, "epoch": 1.78, "learning_rate": 4.8948027551659365e-05, "loss": 0.6243, "step": 2111, "task_loss": 0.3134887218475342 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6293296217918396, "epoch": 1.78, "learning_rate": 4.89448966812774e-05, "loss": 0.7504, "step": 2112, "task_loss": 0.22134238481521606 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.39036330580711365, "epoch": 1.79, "learning_rate": 4.894176581089543e-05, "loss": 0.5992, "step": 2113, "task_loss": 0.09032367169857025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5955778360366821, "epoch": 1.79, "learning_rate": 4.893863494051347e-05, "loss": 0.5707, "step": 2114, "task_loss": 0.7912930250167847 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8447633981704712, "epoch": 1.79, "learning_rate": 4.89355040701315e-05, "loss": 0.6797, "step": 2115, "task_loss": 1.1877171993255615 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4283994734287262, "epoch": 1.79, "learning_rate": 4.893237319974953e-05, "loss": 0.5651, "step": 2116, "task_loss": 0.9739860892295837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5256165862083435, "epoch": 1.79, "learning_rate": 4.892924232936756e-05, "loss": 0.6712, "step": 2117, "task_loss": 0.33050039410591125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.726043701171875, "epoch": 1.79, "learning_rate": 4.89261114589856e-05, "loss": 0.5461, "step": 2118, "task_loss": 0.6563425660133362 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7229875326156616, "epoch": 1.79, "learning_rate": 4.892298058860363e-05, "loss": 0.5403, "step": 2119, "task_loss": 1.492161512374878 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45323097705841064, "epoch": 1.79, "learning_rate": 4.8919849718221665e-05, "loss": 0.923, "step": 2120, "task_loss": 0.8250668048858643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4360085129737854, "epoch": 1.79, "learning_rate": 4.8916718847839703e-05, "loss": 0.6097, "step": 2121, "task_loss": 1.199126124382019 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5901917815208435, "epoch": 1.79, "learning_rate": 4.8913587977457735e-05, "loss": 0.5727, "step": 2122, "task_loss": 0.6161655187606812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7889527678489685, "epoch": 1.79, "learning_rate": 4.891045710707577e-05, "loss": 0.8452, "step": 2123, "task_loss": 1.391523003578186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.729505717754364, "epoch": 1.79, "learning_rate": 4.89073262366938e-05, "loss": 0.7236, "step": 2124, "task_loss": 1.4552538394927979 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 1.0266610383987427, "epoch": 1.8, "learning_rate": 4.890419536631184e-05, "loss": 0.7939, "step": 2125, "task_loss": 1.0114405155181885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5434058904647827, "epoch": 1.8, "learning_rate": 4.890106449592987e-05, "loss": 0.5191, "step": 2126, "task_loss": 0.9051684737205505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6551738381385803, "epoch": 1.8, "learning_rate": 4.88979336255479e-05, "loss": 0.6577, "step": 2127, "task_loss": 0.6065924167633057 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8395287990570068, "epoch": 1.8, "learning_rate": 4.889480275516593e-05, "loss": 0.6926, "step": 2128, "task_loss": 0.9712740778923035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.827460527420044, "epoch": 1.8, "learning_rate": 4.889167188478397e-05, "loss": 0.6317, "step": 2129, "task_loss": 0.1672440618276596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3177676796913147, "epoch": 1.8, "learning_rate": 4.8888541014402e-05, "loss": 0.5893, "step": 2130, "task_loss": 0.2996341586112976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.622376024723053, "epoch": 1.8, "learning_rate": 4.888541014402004e-05, "loss": 0.511, "step": 2131, "task_loss": 0.6662011742591858 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5530962944030762, "epoch": 1.8, "learning_rate": 4.888227927363807e-05, "loss": 0.4772, "step": 2132, "task_loss": 1.3799068927764893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.39280903339385986, "epoch": 1.8, "learning_rate": 4.887914840325611e-05, "loss": 0.4791, "step": 2133, "task_loss": 0.7474582195281982 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9605111479759216, "epoch": 1.8, "learning_rate": 4.8876017532874144e-05, "loss": 0.7368, "step": 2134, "task_loss": 0.7224611043930054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5554821491241455, "epoch": 1.8, "learning_rate": 4.8872886662492175e-05, "loss": 0.565, "step": 2135, "task_loss": 1.2930554151535034 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4907560646533966, "epoch": 1.81, "learning_rate": 4.8869755792110214e-05, "loss": 0.5153, "step": 2136, "task_loss": 0.38207367062568665 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4421542286872864, "epoch": 1.81, "learning_rate": 4.8866624921728246e-05, "loss": 0.4932, "step": 2137, "task_loss": 1.4856674671173096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6923367977142334, "epoch": 1.81, "learning_rate": 4.886349405134628e-05, "loss": 0.683, "step": 2138, "task_loss": 1.5043758153915405 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3068653345108032, "epoch": 1.81, "learning_rate": 4.886036318096431e-05, "loss": 0.403, "step": 2139, "task_loss": 0.5421009659767151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7366296052932739, "epoch": 1.81, "learning_rate": 4.885723231058235e-05, "loss": 0.6499, "step": 2140, "task_loss": 0.8569260239601135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3624809980392456, "epoch": 1.81, "learning_rate": 4.885410144020038e-05, "loss": 0.5115, "step": 2141, "task_loss": 0.07824302464723587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8588196039199829, "epoch": 1.81, "learning_rate": 4.885097056981841e-05, "loss": 0.7043, "step": 2142, "task_loss": 0.9554499983787537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6366950273513794, "epoch": 1.81, "learning_rate": 4.884783969943645e-05, "loss": 0.6665, "step": 2143, "task_loss": 1.2620222568511963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7368850708007812, "epoch": 1.81, "learning_rate": 4.884470882905448e-05, "loss": 0.6095, "step": 2144, "task_loss": 1.1792137622833252 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5826089382171631, "epoch": 1.81, "learning_rate": 4.8841577958672514e-05, "loss": 0.5848, "step": 2145, "task_loss": 0.6086183786392212 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7025729417800903, "epoch": 1.81, "learning_rate": 4.8838447088290545e-05, "loss": 0.5169, "step": 2146, "task_loss": 1.2442482709884644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46185675263404846, "epoch": 1.81, "learning_rate": 4.8835316217908584e-05, "loss": 0.6267, "step": 2147, "task_loss": 0.6671622395515442 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7256494760513306, "epoch": 1.82, "learning_rate": 4.8832185347526616e-05, "loss": 0.5811, "step": 2148, "task_loss": 0.3093624413013458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7267260551452637, "epoch": 1.82, "learning_rate": 4.882905447714465e-05, "loss": 0.5474, "step": 2149, "task_loss": 0.6404131650924683 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.39623990654945374, "epoch": 1.82, "learning_rate": 4.882592360676268e-05, "loss": 0.5643, "step": 2150, "task_loss": 0.14476701617240906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5093360543251038, "epoch": 1.82, "learning_rate": 4.882279273638072e-05, "loss": 0.472, "step": 2151, "task_loss": 0.5220075249671936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.465090811252594, "epoch": 1.82, "learning_rate": 4.881966186599875e-05, "loss": 0.5347, "step": 2152, "task_loss": 0.8201823830604553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8854434490203857, "epoch": 1.82, "learning_rate": 4.881653099561678e-05, "loss": 0.6039, "step": 2153, "task_loss": 0.8482139706611633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8003247976303101, "epoch": 1.82, "learning_rate": 4.881340012523481e-05, "loss": 0.7288, "step": 2154, "task_loss": 1.02394700050354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6002784967422485, "epoch": 1.82, "learning_rate": 4.881026925485285e-05, "loss": 0.6398, "step": 2155, "task_loss": 0.7402263879776001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5710400938987732, "epoch": 1.82, "learning_rate": 4.8807138384470884e-05, "loss": 0.5196, "step": 2156, "task_loss": 1.2978270053863525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5296972990036011, "epoch": 1.82, "learning_rate": 4.8804007514088915e-05, "loss": 0.607, "step": 2157, "task_loss": 0.7336243987083435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37521642446517944, "epoch": 1.82, "learning_rate": 4.8800876643706954e-05, "loss": 0.5288, "step": 2158, "task_loss": 0.4478328227996826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5608843564987183, "epoch": 1.82, "learning_rate": 4.8797745773324986e-05, "loss": 0.543, "step": 2159, "task_loss": 0.9204266667366028 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7064813375473022, "epoch": 1.83, "learning_rate": 4.879461490294302e-05, "loss": 0.6402, "step": 2160, "task_loss": 0.9193848371505737 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3050161898136139, "epoch": 1.83, "learning_rate": 4.879148403256105e-05, "loss": 0.6641, "step": 2161, "task_loss": 0.20055975019931793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7948587536811829, "epoch": 1.83, "learning_rate": 4.878835316217909e-05, "loss": 0.7333, "step": 2162, "task_loss": 0.7384874224662781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.33466649055480957, "epoch": 1.83, "learning_rate": 4.878522229179712e-05, "loss": 0.6593, "step": 2163, "task_loss": 0.09069152921438217 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5731754302978516, "epoch": 1.83, "learning_rate": 4.878209142141516e-05, "loss": 0.5703, "step": 2164, "task_loss": 0.45842090249061584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6033138036727905, "epoch": 1.83, "learning_rate": 4.877896055103319e-05, "loss": 0.5052, "step": 2165, "task_loss": 0.3740144968032837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5677679181098938, "epoch": 1.83, "learning_rate": 4.877582968065122e-05, "loss": 0.4807, "step": 2166, "task_loss": 1.1866438388824463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5365862250328064, "epoch": 1.83, "learning_rate": 4.877269881026926e-05, "loss": 0.7371, "step": 2167, "task_loss": 0.43480488657951355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7548823356628418, "epoch": 1.83, "learning_rate": 4.876956793988729e-05, "loss": 0.6137, "step": 2168, "task_loss": 0.5724943280220032 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7481116652488708, "epoch": 1.83, "learning_rate": 4.876643706950533e-05, "loss": 0.723, "step": 2169, "task_loss": 0.3660881221294403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47073423862457275, "epoch": 1.83, "learning_rate": 4.876330619912336e-05, "loss": 0.4317, "step": 2170, "task_loss": 0.8030180931091309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6687943935394287, "epoch": 1.83, "learning_rate": 4.8760175328741394e-05, "loss": 0.6084, "step": 2171, "task_loss": 1.3962076902389526 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8526927828788757, "epoch": 1.84, "learning_rate": 4.8757044458359426e-05, "loss": 0.5972, "step": 2172, "task_loss": 1.5182141065597534 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9695984125137329, "epoch": 1.84, "learning_rate": 4.8753913587977464e-05, "loss": 0.5844, "step": 2173, "task_loss": 0.487541526556015 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7158949375152588, "epoch": 1.84, "learning_rate": 4.8750782717595496e-05, "loss": 0.6288, "step": 2174, "task_loss": 0.9329524040222168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.28207823634147644, "epoch": 1.84, "learning_rate": 4.874765184721353e-05, "loss": 0.4656, "step": 2175, "task_loss": 0.19819530844688416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.33092135190963745, "epoch": 1.84, "learning_rate": 4.874452097683156e-05, "loss": 0.5663, "step": 2176, "task_loss": 1.1531652212142944 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3137865364551544, "epoch": 1.84, "learning_rate": 4.87413901064496e-05, "loss": 0.4099, "step": 2177, "task_loss": 0.2796710729598999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3901706039905548, "epoch": 1.84, "learning_rate": 4.873825923606763e-05, "loss": 0.6147, "step": 2178, "task_loss": 0.6988053917884827 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5045095682144165, "epoch": 1.84, "learning_rate": 4.873512836568566e-05, "loss": 0.544, "step": 2179, "task_loss": 1.0051348209381104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46893662214279175, "epoch": 1.84, "learning_rate": 4.87319974953037e-05, "loss": 0.4505, "step": 2180, "task_loss": 0.8602231740951538 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.463205486536026, "epoch": 1.84, "learning_rate": 4.872886662492173e-05, "loss": 0.5216, "step": 2181, "task_loss": 0.9686093926429749 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49171945452690125, "epoch": 1.84, "learning_rate": 4.8725735754539764e-05, "loss": 0.4508, "step": 2182, "task_loss": 0.39173591136932373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3818315863609314, "epoch": 1.84, "learning_rate": 4.8722604884157796e-05, "loss": 0.5222, "step": 2183, "task_loss": 1.472856044769287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5648458003997803, "epoch": 1.85, "learning_rate": 4.8719474013775834e-05, "loss": 0.7529, "step": 2184, "task_loss": 1.0871113538742065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8122997283935547, "epoch": 1.85, "learning_rate": 4.8716343143393866e-05, "loss": 0.6786, "step": 2185, "task_loss": 0.655812680721283 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6582978963851929, "epoch": 1.85, "learning_rate": 4.87132122730119e-05, "loss": 0.7032, "step": 2186, "task_loss": 1.7352465391159058 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5490027666091919, "epoch": 1.85, "learning_rate": 4.871008140262993e-05, "loss": 0.4023, "step": 2187, "task_loss": 0.48472151160240173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7001916766166687, "epoch": 1.85, "learning_rate": 4.870695053224797e-05, "loss": 0.6216, "step": 2188, "task_loss": 0.5699008107185364 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4870021939277649, "epoch": 1.85, "learning_rate": 4.8703819661866e-05, "loss": 0.4118, "step": 2189, "task_loss": 0.47853875160217285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46022090315818787, "epoch": 1.85, "learning_rate": 4.870068879148403e-05, "loss": 0.4963, "step": 2190, "task_loss": 0.6397734880447388 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5513700246810913, "epoch": 1.85, "learning_rate": 4.8697557921102064e-05, "loss": 0.5925, "step": 2191, "task_loss": 0.7450762391090393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.36688822507858276, "epoch": 1.85, "learning_rate": 4.86944270507201e-05, "loss": 0.4863, "step": 2192, "task_loss": 0.20471200346946716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46115782856941223, "epoch": 1.85, "learning_rate": 4.8691296180338134e-05, "loss": 0.6307, "step": 2193, "task_loss": 0.6599039435386658 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7066313028335571, "epoch": 1.85, "learning_rate": 4.8688165309956166e-05, "loss": 0.4846, "step": 2194, "task_loss": 0.7911555171012878 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6729583740234375, "epoch": 1.85, "learning_rate": 4.8685034439574204e-05, "loss": 0.5681, "step": 2195, "task_loss": 0.43649423122406006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44678613543510437, "epoch": 1.86, "learning_rate": 4.8681903569192236e-05, "loss": 0.6135, "step": 2196, "task_loss": 0.30738404393196106 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4595412015914917, "epoch": 1.86, "learning_rate": 4.867877269881027e-05, "loss": 0.5655, "step": 2197, "task_loss": 0.5316629409790039 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.24386191368103027, "epoch": 1.86, "learning_rate": 4.8675641828428306e-05, "loss": 0.5996, "step": 2198, "task_loss": 0.3736094534397125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6645935773849487, "epoch": 1.86, "learning_rate": 4.867251095804634e-05, "loss": 0.6208, "step": 2199, "task_loss": 0.9722223281860352 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.34003517031669617, "epoch": 1.86, "learning_rate": 4.866938008766438e-05, "loss": 0.5625, "step": 2200, "task_loss": 0.549965500831604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6636495590209961, "epoch": 1.86, "learning_rate": 4.866624921728241e-05, "loss": 0.5624, "step": 2201, "task_loss": 0.7956205010414124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3100022077560425, "epoch": 1.86, "learning_rate": 4.866311834690044e-05, "loss": 0.4288, "step": 2202, "task_loss": 0.3601055145263672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3589179515838623, "epoch": 1.86, "learning_rate": 4.865998747651848e-05, "loss": 0.4224, "step": 2203, "task_loss": 0.5323647856712341 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4949137568473816, "epoch": 1.86, "learning_rate": 4.865685660613651e-05, "loss": 0.5327, "step": 2204, "task_loss": 0.7203745245933533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5541373491287231, "epoch": 1.86, "learning_rate": 4.865372573575454e-05, "loss": 0.5169, "step": 2205, "task_loss": 0.5352662801742554 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.621067225933075, "epoch": 1.86, "learning_rate": 4.865059486537258e-05, "loss": 0.5533, "step": 2206, "task_loss": 0.3639921247959137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5573874711990356, "epoch": 1.87, "learning_rate": 4.864746399499061e-05, "loss": 0.6785, "step": 2207, "task_loss": 1.1045007705688477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3991885483264923, "epoch": 1.87, "learning_rate": 4.8644333124608645e-05, "loss": 0.5641, "step": 2208, "task_loss": 0.4841661751270294 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.48987460136413574, "epoch": 1.87, "learning_rate": 4.8641202254226676e-05, "loss": 0.4202, "step": 2209, "task_loss": 0.6078423261642456 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5457128286361694, "epoch": 1.87, "learning_rate": 4.8638071383844715e-05, "loss": 0.4337, "step": 2210, "task_loss": 0.867026150226593 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41692811250686646, "epoch": 1.87, "learning_rate": 4.863494051346275e-05, "loss": 0.4691, "step": 2211, "task_loss": 0.502586305141449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7726911306381226, "epoch": 1.87, "learning_rate": 4.863180964308078e-05, "loss": 0.5822, "step": 2212, "task_loss": 1.239492654800415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5839382410049438, "epoch": 1.87, "learning_rate": 4.862867877269881e-05, "loss": 0.6054, "step": 2213, "task_loss": 0.7306771278381348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6883859038352966, "epoch": 1.87, "learning_rate": 4.862554790231685e-05, "loss": 0.6591, "step": 2214, "task_loss": 0.8644038438796997 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6988998055458069, "epoch": 1.87, "learning_rate": 4.862241703193488e-05, "loss": 0.5352, "step": 2215, "task_loss": 0.7293692231178284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.847305417060852, "epoch": 1.87, "learning_rate": 4.861928616155291e-05, "loss": 0.5465, "step": 2216, "task_loss": 0.7182483673095703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5955390334129333, "epoch": 1.87, "learning_rate": 4.861615529117095e-05, "loss": 0.5721, "step": 2217, "task_loss": 0.8275623321533203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49777621030807495, "epoch": 1.87, "learning_rate": 4.861302442078898e-05, "loss": 0.5567, "step": 2218, "task_loss": 1.8063640594482422 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.36714044213294983, "epoch": 1.88, "learning_rate": 4.8609893550407015e-05, "loss": 0.3992, "step": 2219, "task_loss": 0.3525436818599701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2521466314792633, "epoch": 1.88, "learning_rate": 4.8606762680025046e-05, "loss": 0.5996, "step": 2220, "task_loss": 0.38019150495529175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4946924149990082, "epoch": 1.88, "learning_rate": 4.8603631809643085e-05, "loss": 0.6946, "step": 2221, "task_loss": 0.40845057368278503 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5118348002433777, "epoch": 1.88, "learning_rate": 4.860050093926112e-05, "loss": 0.4591, "step": 2222, "task_loss": 0.7039198875427246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7832226753234863, "epoch": 1.88, "learning_rate": 4.859737006887915e-05, "loss": 0.5257, "step": 2223, "task_loss": 1.1461080312728882 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5330616235733032, "epoch": 1.88, "learning_rate": 4.859423919849718e-05, "loss": 0.5571, "step": 2224, "task_loss": 0.9466378688812256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47596317529678345, "epoch": 1.88, "learning_rate": 4.859110832811522e-05, "loss": 0.6174, "step": 2225, "task_loss": 1.0317274332046509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3439880609512329, "epoch": 1.88, "learning_rate": 4.858797745773325e-05, "loss": 0.4797, "step": 2226, "task_loss": 0.14573565125465393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4096469283103943, "epoch": 1.88, "learning_rate": 4.858484658735128e-05, "loss": 0.5885, "step": 2227, "task_loss": 0.3415091931819916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6815884113311768, "epoch": 1.88, "learning_rate": 4.8581715716969314e-05, "loss": 0.5847, "step": 2228, "task_loss": 1.0233526229858398 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9058316946029663, "epoch": 1.88, "learning_rate": 4.857858484658735e-05, "loss": 0.6057, "step": 2229, "task_loss": 0.4832155704498291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7172305583953857, "epoch": 1.88, "learning_rate": 4.8575453976205384e-05, "loss": 0.7863, "step": 2230, "task_loss": 1.3079471588134766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6532737016677856, "epoch": 1.89, "learning_rate": 4.857232310582342e-05, "loss": 0.7322, "step": 2231, "task_loss": 1.1798009872436523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43863439559936523, "epoch": 1.89, "learning_rate": 4.8569192235441455e-05, "loss": 0.5499, "step": 2232, "task_loss": 0.3716825246810913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7116318941116333, "epoch": 1.89, "learning_rate": 4.8566061365059487e-05, "loss": 0.636, "step": 2233, "task_loss": 0.7725350856781006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.819441556930542, "epoch": 1.89, "learning_rate": 4.8562930494677525e-05, "loss": 0.5958, "step": 2234, "task_loss": 0.31311309337615967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8627134561538696, "epoch": 1.89, "learning_rate": 4.855979962429556e-05, "loss": 0.6031, "step": 2235, "task_loss": 0.8065341114997864 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5047589540481567, "epoch": 1.89, "learning_rate": 4.8556668753913595e-05, "loss": 0.3941, "step": 2236, "task_loss": 0.5051278471946716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8206089735031128, "epoch": 1.89, "learning_rate": 4.855353788353163e-05, "loss": 0.637, "step": 2237, "task_loss": 0.713137686252594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5062882304191589, "epoch": 1.89, "learning_rate": 4.855040701314966e-05, "loss": 0.5897, "step": 2238, "task_loss": 0.9989140033721924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45204123854637146, "epoch": 1.89, "learning_rate": 4.854727614276769e-05, "loss": 0.5192, "step": 2239, "task_loss": 0.4997401535511017 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4479646384716034, "epoch": 1.89, "learning_rate": 4.854414527238573e-05, "loss": 0.5449, "step": 2240, "task_loss": 0.5937856435775757 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4261937141418457, "epoch": 1.89, "learning_rate": 4.854101440200376e-05, "loss": 0.589, "step": 2241, "task_loss": 0.45445430278778076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4354429841041565, "epoch": 1.89, "learning_rate": 4.853788353162179e-05, "loss": 0.7142, "step": 2242, "task_loss": 1.290386438369751 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5802772045135498, "epoch": 1.9, "learning_rate": 4.853475266123983e-05, "loss": 0.4622, "step": 2243, "task_loss": 0.9056605696678162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6103945970535278, "epoch": 1.9, "learning_rate": 4.853162179085786e-05, "loss": 0.561, "step": 2244, "task_loss": 0.809261679649353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5052226781845093, "epoch": 1.9, "learning_rate": 4.8528490920475895e-05, "loss": 0.7168, "step": 2245, "task_loss": 0.6984712481498718 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.614409327507019, "epoch": 1.9, "learning_rate": 4.852536005009393e-05, "loss": 0.6911, "step": 2246, "task_loss": 0.6481942534446716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.825417697429657, "epoch": 1.9, "learning_rate": 4.8522229179711965e-05, "loss": 0.6651, "step": 2247, "task_loss": 0.6266444325447083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5685454607009888, "epoch": 1.9, "learning_rate": 4.851909830933e-05, "loss": 0.578, "step": 2248, "task_loss": 1.0769387483596802 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.470525860786438, "epoch": 1.9, "learning_rate": 4.851596743894803e-05, "loss": 0.5814, "step": 2249, "task_loss": 0.624211311340332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5710526704788208, "epoch": 1.9, "learning_rate": 4.851283656856606e-05, "loss": 0.525, "step": 2250, "task_loss": 1.6746357679367065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4028887152671814, "epoch": 1.9, "learning_rate": 4.85097056981841e-05, "loss": 0.4739, "step": 2251, "task_loss": 0.8210620880126953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.26171261072158813, "epoch": 1.9, "learning_rate": 4.850657482780213e-05, "loss": 0.5415, "step": 2252, "task_loss": 0.2836996018886566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4880768358707428, "epoch": 1.9, "learning_rate": 4.850344395742016e-05, "loss": 0.4706, "step": 2253, "task_loss": 0.49408388137817383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6781322360038757, "epoch": 1.9, "learning_rate": 4.85003130870382e-05, "loss": 0.4689, "step": 2254, "task_loss": 0.24423527717590332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4817344546318054, "epoch": 1.91, "learning_rate": 4.849718221665623e-05, "loss": 0.4238, "step": 2255, "task_loss": 1.0819272994995117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4278075695037842, "epoch": 1.91, "learning_rate": 4.8494051346274265e-05, "loss": 0.6338, "step": 2256, "task_loss": 0.5323134660720825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43255361914634705, "epoch": 1.91, "learning_rate": 4.84909204758923e-05, "loss": 0.6065, "step": 2257, "task_loss": 0.15960419178009033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.755547046661377, "epoch": 1.91, "learning_rate": 4.8487789605510335e-05, "loss": 0.4504, "step": 2258, "task_loss": 0.3668578863143921 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4487781524658203, "epoch": 1.91, "learning_rate": 4.848465873512837e-05, "loss": 0.5707, "step": 2259, "task_loss": 0.926218569278717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46880006790161133, "epoch": 1.91, "learning_rate": 4.84815278647464e-05, "loss": 0.502, "step": 2260, "task_loss": 1.4549139738082886 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8510129451751709, "epoch": 1.91, "learning_rate": 4.847839699436443e-05, "loss": 0.673, "step": 2261, "task_loss": 1.2251989841461182 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4598747193813324, "epoch": 1.91, "learning_rate": 4.847526612398247e-05, "loss": 0.4787, "step": 2262, "task_loss": 0.6540231704711914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4813276529312134, "epoch": 1.91, "learning_rate": 4.84721352536005e-05, "loss": 0.4172, "step": 2263, "task_loss": 0.1667739748954773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7407900094985962, "epoch": 1.91, "learning_rate": 4.846900438321853e-05, "loss": 0.6184, "step": 2264, "task_loss": 0.5356577634811401 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5079731941223145, "epoch": 1.91, "learning_rate": 4.846587351283657e-05, "loss": 0.52, "step": 2265, "task_loss": 0.5252299308776855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9221036434173584, "epoch": 1.91, "learning_rate": 4.84627426424546e-05, "loss": 0.6639, "step": 2266, "task_loss": 1.843277096748352 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5918920040130615, "epoch": 1.92, "learning_rate": 4.845961177207264e-05, "loss": 0.5695, "step": 2267, "task_loss": 0.4227873384952545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3924558460712433, "epoch": 1.92, "learning_rate": 4.8456480901690673e-05, "loss": 0.3948, "step": 2268, "task_loss": 0.5301018357276917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38882505893707275, "epoch": 1.92, "learning_rate": 4.8453350031308705e-05, "loss": 0.4935, "step": 2269, "task_loss": 0.4432561993598938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.46606874465942383, "epoch": 1.92, "learning_rate": 4.8450219160926744e-05, "loss": 0.5263, "step": 2270, "task_loss": 0.9365003108978271 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37876835465431213, "epoch": 1.92, "learning_rate": 4.8447088290544776e-05, "loss": 0.4668, "step": 2271, "task_loss": 0.4731386601924896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7847674489021301, "epoch": 1.92, "learning_rate": 4.844395742016281e-05, "loss": 0.478, "step": 2272, "task_loss": 0.6595020294189453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4343452751636505, "epoch": 1.92, "learning_rate": 4.8440826549780846e-05, "loss": 0.5318, "step": 2273, "task_loss": 0.07590820640325546 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.446866512298584, "epoch": 1.92, "learning_rate": 4.843769567939888e-05, "loss": 0.5559, "step": 2274, "task_loss": 0.5367981195449829 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5787578821182251, "epoch": 1.92, "learning_rate": 4.843456480901691e-05, "loss": 0.5417, "step": 2275, "task_loss": 0.6284345984458923 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.40120887756347656, "epoch": 1.92, "learning_rate": 4.843143393863494e-05, "loss": 0.6421, "step": 2276, "task_loss": 0.33334973454475403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6779735684394836, "epoch": 1.92, "learning_rate": 4.842830306825298e-05, "loss": 0.6171, "step": 2277, "task_loss": 0.293052077293396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4384811222553253, "epoch": 1.93, "learning_rate": 4.842517219787101e-05, "loss": 0.5597, "step": 2278, "task_loss": 0.6216214895248413 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7013087868690491, "epoch": 1.93, "learning_rate": 4.842204132748904e-05, "loss": 0.6151, "step": 2279, "task_loss": 0.6321526169776917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7947863936424255, "epoch": 1.93, "learning_rate": 4.841891045710708e-05, "loss": 0.5582, "step": 2280, "task_loss": 0.9204905033111572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.44094327092170715, "epoch": 1.93, "learning_rate": 4.8415779586725114e-05, "loss": 0.5425, "step": 2281, "task_loss": 0.9712821841239929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.47170448303222656, "epoch": 1.93, "learning_rate": 4.8412648716343145e-05, "loss": 0.6048, "step": 2282, "task_loss": 0.7364827394485474 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.860568642616272, "epoch": 1.93, "learning_rate": 4.840951784596118e-05, "loss": 0.6358, "step": 2283, "task_loss": 0.8913965821266174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3109630346298218, "epoch": 1.93, "learning_rate": 4.8406386975579216e-05, "loss": 0.4889, "step": 2284, "task_loss": 0.4300428032875061 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5326797962188721, "epoch": 1.93, "learning_rate": 4.840325610519725e-05, "loss": 0.6871, "step": 2285, "task_loss": 0.9036595821380615 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6360039710998535, "epoch": 1.93, "learning_rate": 4.840012523481528e-05, "loss": 0.5964, "step": 2286, "task_loss": 1.129393458366394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8353090286254883, "epoch": 1.93, "learning_rate": 4.839699436443331e-05, "loss": 0.6296, "step": 2287, "task_loss": 1.259513020515442 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4459793269634247, "epoch": 1.93, "learning_rate": 4.839386349405135e-05, "loss": 0.6134, "step": 2288, "task_loss": 0.7472996115684509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6030188202857971, "epoch": 1.93, "learning_rate": 4.839073262366938e-05, "loss": 0.4816, "step": 2289, "task_loss": 1.0895473957061768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2766847610473633, "epoch": 1.94, "learning_rate": 4.838760175328741e-05, "loss": 0.5977, "step": 2290, "task_loss": 0.44785985350608826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9258207082748413, "epoch": 1.94, "learning_rate": 4.838447088290545e-05, "loss": 0.5965, "step": 2291, "task_loss": 2.1939585208892822 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.580630362033844, "epoch": 1.94, "learning_rate": 4.8381340012523484e-05, "loss": 0.5953, "step": 2292, "task_loss": 0.8714878559112549 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.582234263420105, "epoch": 1.94, "learning_rate": 4.8378209142141515e-05, "loss": 0.5938, "step": 2293, "task_loss": 0.5279783010482788 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.37117767333984375, "epoch": 1.94, "learning_rate": 4.837507827175955e-05, "loss": 0.695, "step": 2294, "task_loss": 0.16425535082817078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7244715690612793, "epoch": 1.94, "learning_rate": 4.8371947401377586e-05, "loss": 0.4713, "step": 2295, "task_loss": 0.7118629217147827 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5657180547714233, "epoch": 1.94, "learning_rate": 4.836881653099562e-05, "loss": 0.4816, "step": 2296, "task_loss": 1.0887293815612793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4801936149597168, "epoch": 1.94, "learning_rate": 4.836568566061365e-05, "loss": 0.4649, "step": 2297, "task_loss": 0.8230075240135193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.36039698123931885, "epoch": 1.94, "learning_rate": 4.836255479023168e-05, "loss": 0.4639, "step": 2298, "task_loss": 0.31595292687416077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5434979200363159, "epoch": 1.94, "learning_rate": 4.835942391984972e-05, "loss": 0.4652, "step": 2299, "task_loss": 0.9178555607795715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5618191957473755, "epoch": 1.94, "learning_rate": 4.835629304946775e-05, "loss": 0.5013, "step": 2300, "task_loss": 1.772692084312439 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.20331023633480072, "epoch": 1.94, "learning_rate": 4.835316217908579e-05, "loss": 0.4754, "step": 2301, "task_loss": 0.47345834970474243 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5976462364196777, "epoch": 1.95, "learning_rate": 4.835003130870382e-05, "loss": 0.6929, "step": 2302, "task_loss": 0.6692781448364258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3999878168106079, "epoch": 1.95, "learning_rate": 4.834690043832186e-05, "loss": 0.4871, "step": 2303, "task_loss": 1.6149033308029175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6641921401023865, "epoch": 1.95, "learning_rate": 4.834376956793989e-05, "loss": 0.4804, "step": 2304, "task_loss": 0.757662296295166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.848527193069458, "epoch": 1.95, "learning_rate": 4.8340638697557924e-05, "loss": 0.6909, "step": 2305, "task_loss": 0.9269886016845703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.49475526809692383, "epoch": 1.95, "learning_rate": 4.833750782717596e-05, "loss": 0.4947, "step": 2306, "task_loss": 1.1016515493392944 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8003455400466919, "epoch": 1.95, "learning_rate": 4.8334376956793994e-05, "loss": 0.495, "step": 2307, "task_loss": 0.7892528176307678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7096338272094727, "epoch": 1.95, "learning_rate": 4.8331246086412026e-05, "loss": 0.6716, "step": 2308, "task_loss": 0.6020727753639221 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5426114201545715, "epoch": 1.95, "learning_rate": 4.832811521603006e-05, "loss": 0.6022, "step": 2309, "task_loss": 0.6870071887969971 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8560212850570679, "epoch": 1.95, "learning_rate": 4.8324984345648096e-05, "loss": 0.6392, "step": 2310, "task_loss": 0.7902010083198547 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6016613245010376, "epoch": 1.95, "learning_rate": 4.832185347526613e-05, "loss": 0.4684, "step": 2311, "task_loss": 0.6191394329071045 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.9783615469932556, "epoch": 1.95, "learning_rate": 4.831872260488416e-05, "loss": 0.6139, "step": 2312, "task_loss": 1.0745034217834473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6236714720726013, "epoch": 1.95, "learning_rate": 4.831559173450219e-05, "loss": 0.6124, "step": 2313, "task_loss": 1.1000510454177856 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5916001200675964, "epoch": 1.96, "learning_rate": 4.831246086412023e-05, "loss": 0.5616, "step": 2314, "task_loss": 1.220458984375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4502362906932831, "epoch": 1.96, "learning_rate": 4.830932999373826e-05, "loss": 0.6761, "step": 2315, "task_loss": 0.8882645964622498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4899022579193115, "epoch": 1.96, "learning_rate": 4.8306199123356294e-05, "loss": 0.4327, "step": 2316, "task_loss": 0.9517663717269897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6315869092941284, "epoch": 1.96, "learning_rate": 4.830306825297433e-05, "loss": 0.5883, "step": 2317, "task_loss": 0.7336822748184204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.41864582896232605, "epoch": 1.96, "learning_rate": 4.8299937382592364e-05, "loss": 0.3649, "step": 2318, "task_loss": 0.37236616015434265 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5211153030395508, "epoch": 1.96, "learning_rate": 4.8296806512210396e-05, "loss": 0.4577, "step": 2319, "task_loss": 0.42924952507019043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3662870526313782, "epoch": 1.96, "learning_rate": 4.829367564182843e-05, "loss": 0.531, "step": 2320, "task_loss": 0.4386194348335266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5030696392059326, "epoch": 1.96, "learning_rate": 4.8290544771446466e-05, "loss": 0.5211, "step": 2321, "task_loss": 0.3987022042274475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.613250732421875, "epoch": 1.96, "learning_rate": 4.82874139010645e-05, "loss": 0.4916, "step": 2322, "task_loss": 0.33156564831733704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5225822925567627, "epoch": 1.96, "learning_rate": 4.828428303068253e-05, "loss": 0.5046, "step": 2323, "task_loss": 0.37034887075424194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5736628770828247, "epoch": 1.96, "learning_rate": 4.828115216030056e-05, "loss": 0.5536, "step": 2324, "task_loss": 0.32722270488739014 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5345181226730347, "epoch": 1.96, "learning_rate": 4.82780212899186e-05, "loss": 0.6735, "step": 2325, "task_loss": 0.7939822673797607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4815193712711334, "epoch": 1.97, "learning_rate": 4.827489041953663e-05, "loss": 0.6578, "step": 2326, "task_loss": 0.896963894367218 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.42960891127586365, "epoch": 1.97, "learning_rate": 4.8271759549154664e-05, "loss": 0.4585, "step": 2327, "task_loss": 0.38021665811538696 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5771958827972412, "epoch": 1.97, "learning_rate": 4.82686286787727e-05, "loss": 0.5368, "step": 2328, "task_loss": 0.5883228182792664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.29872360825538635, "epoch": 1.97, "learning_rate": 4.8265497808390734e-05, "loss": 0.5615, "step": 2329, "task_loss": 0.561460018157959 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5133383274078369, "epoch": 1.97, "learning_rate": 4.8262366938008766e-05, "loss": 0.501, "step": 2330, "task_loss": 0.8558510541915894 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6714968085289001, "epoch": 1.97, "learning_rate": 4.82592360676268e-05, "loss": 0.5672, "step": 2331, "task_loss": 0.3734920918941498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.440443217754364, "epoch": 1.97, "learning_rate": 4.8256105197244836e-05, "loss": 0.5001, "step": 2332, "task_loss": 0.16747237741947174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45396584272384644, "epoch": 1.97, "learning_rate": 4.825297432686287e-05, "loss": 0.4829, "step": 2333, "task_loss": 0.7542091012001038 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5400270223617554, "epoch": 1.97, "learning_rate": 4.8249843456480907e-05, "loss": 0.6094, "step": 2334, "task_loss": 1.1040728092193604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5729333162307739, "epoch": 1.97, "learning_rate": 4.824671258609894e-05, "loss": 0.6343, "step": 2335, "task_loss": 1.1269257068634033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6913278102874756, "epoch": 1.97, "learning_rate": 4.824358171571697e-05, "loss": 0.834, "step": 2336, "task_loss": 1.6339163780212402 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4340217709541321, "epoch": 1.97, "learning_rate": 4.824045084533501e-05, "loss": 0.5217, "step": 2337, "task_loss": 1.2662758827209473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7026585340499878, "epoch": 1.98, "learning_rate": 4.823731997495304e-05, "loss": 0.5176, "step": 2338, "task_loss": 1.089275598526001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7056485414505005, "epoch": 1.98, "learning_rate": 4.823418910457107e-05, "loss": 0.6633, "step": 2339, "task_loss": 0.21117419004440308 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5564246773719788, "epoch": 1.98, "learning_rate": 4.823105823418911e-05, "loss": 0.484, "step": 2340, "task_loss": 0.1559027135372162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.43339788913726807, "epoch": 1.98, "learning_rate": 4.822792736380714e-05, "loss": 0.5432, "step": 2341, "task_loss": 0.518622636795044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2995416522026062, "epoch": 1.98, "learning_rate": 4.8224796493425174e-05, "loss": 0.3897, "step": 2342, "task_loss": 0.43075984716415405 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.50166916847229, "epoch": 1.98, "learning_rate": 4.822166562304321e-05, "loss": 0.4344, "step": 2343, "task_loss": 0.6882398724555969 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.4896463453769684, "epoch": 1.98, "learning_rate": 4.8218534752661245e-05, "loss": 0.5129, "step": 2344, "task_loss": 0.6826088428497314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6738490462303162, "epoch": 1.98, "learning_rate": 4.8215403882279276e-05, "loss": 0.5819, "step": 2345, "task_loss": 1.7720729112625122 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3711905777454376, "epoch": 1.98, "learning_rate": 4.821227301189731e-05, "loss": 0.4951, "step": 2346, "task_loss": 0.4412442445755005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45153188705444336, "epoch": 1.98, "learning_rate": 4.820914214151535e-05, "loss": 0.5616, "step": 2347, "task_loss": 0.46905645728111267 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.7163456678390503, "epoch": 1.98, "learning_rate": 4.820601127113338e-05, "loss": 0.716, "step": 2348, "task_loss": 1.5200878381729126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.21376165747642517, "epoch": 1.99, "learning_rate": 4.820288040075141e-05, "loss": 0.6132, "step": 2349, "task_loss": 0.03780439496040344 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6425676345825195, "epoch": 1.99, "learning_rate": 4.819974953036944e-05, "loss": 0.6212, "step": 2350, "task_loss": 0.530227780342102 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5389307141304016, "epoch": 1.99, "learning_rate": 4.819661865998748e-05, "loss": 0.5552, "step": 2351, "task_loss": 0.2800722122192383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8230087757110596, "epoch": 1.99, "learning_rate": 4.819348778960551e-05, "loss": 0.5647, "step": 2352, "task_loss": 0.3036193251609802 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.975246787071228, "epoch": 1.99, "learning_rate": 4.8190356919223544e-05, "loss": 0.733, "step": 2353, "task_loss": 1.2826430797576904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.3644910454750061, "epoch": 1.99, "learning_rate": 4.818722604884158e-05, "loss": 0.4565, "step": 2354, "task_loss": 0.12701237201690674 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5540529489517212, "epoch": 1.99, "learning_rate": 4.8184095178459615e-05, "loss": 0.4749, "step": 2355, "task_loss": 0.6275436878204346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6679797768592834, "epoch": 1.99, "learning_rate": 4.8180964308077646e-05, "loss": 0.58, "step": 2356, "task_loss": 0.8768760561943054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2897258400917053, "epoch": 1.99, "learning_rate": 4.817783343769568e-05, "loss": 0.3722, "step": 2357, "task_loss": 0.37791842222213745 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.38319751620292664, "epoch": 1.99, "learning_rate": 4.817470256731372e-05, "loss": 0.6613, "step": 2358, "task_loss": 0.1642351746559143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.5559956431388855, "epoch": 1.99, "learning_rate": 4.817157169693175e-05, "loss": 0.6609, "step": 2359, "task_loss": 0.9434230327606201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.2253756821155548, "epoch": 1.99, "learning_rate": 4.816844082654978e-05, "loss": 0.3833, "step": 2360, "task_loss": 0.6086814403533936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.23713171482086182, "epoch": 2.0, "learning_rate": 4.816530995616781e-05, "loss": 0.5745, "step": 2361, "task_loss": 0.4492885172367096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.45277753472328186, "epoch": 2.0, "learning_rate": 4.816217908578585e-05, "loss": 0.6587, "step": 2362, "task_loss": 0.19286037981510162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6741446852684021, "epoch": 2.0, "learning_rate": 4.815904821540388e-05, "loss": 0.6458, "step": 2363, "task_loss": 1.4836196899414062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.269355833530426, "epoch": 2.0, "learning_rate": 4.8155917345021914e-05, "loss": 0.5648, "step": 2364, "task_loss": 1.4011789560317993 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.8324846625328064, "epoch": 2.0, "learning_rate": 4.815278647463995e-05, "loss": 0.6006, "step": 2365, "task_loss": 1.1632333993911743 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -Infinity, "compression/movement_sparsity/linear_layer_sparsity": 0.0, "compression/movement_sparsity/model_sparsity": 0.0, "compression_loss": 0.0, "distillation_loss": 0.6602296829223633, "epoch": 2.0, "learning_rate": 4.8149655604257984e-05, "loss": 0.5425, "step": 2366, "task_loss": 0.9465316534042358 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0, "compression/movement_sparsity/importance_threshold": -0.00649014487862587, "compression/movement_sparsity/linear_layer_sparsity": 0.0010077591035773266, "compression/movement_sparsity/model_sparsity": 0.0009731394781398207, "compression_loss": 0.0, "distillation_loss": 0.6075857877731323, "epoch": 2.0, "learning_rate": 4.8146524733876016e-05, "loss": 1.0106, "step": 2367, "task_loss": 1.584323525428772 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0008443568368990961, "compression/movement_sparsity/importance_threshold": -0.006484664880425137, "compression/movement_sparsity/linear_layer_sparsity": 0.0010230220381512313, "compression/movement_sparsity/model_sparsity": 0.000987878083956833, "compression_loss": 0.09122346341609955, "distillation_loss": 0.7368004322052002, "epoch": 2.0, "learning_rate": 4.8143393863494055e-05, "loss": 0.7928, "step": 2368, "task_loss": 0.8754284381866455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0016882382476638425, "compression/movement_sparsity/importance_threshold": -0.006479187967808894, "compression/movement_sparsity/linear_layer_sparsity": 0.0010230220381512313, "compression/movement_sparsity/model_sparsity": 0.000987878083956833, "compression_loss": 0.18239536881446838, "distillation_loss": 0.5301963686943054, "epoch": 2.0, "learning_rate": 4.814026299311209e-05, "loss": 0.6769, "step": 2369, "task_loss": 0.7310242056846619 }, { "compression/movement_sparsity/importance_regularization_factor": 0.002531644366179364, "compression/movement_sparsity/importance_threshold": -0.0064737141399082085, "compression/movement_sparsity/linear_layer_sparsity": 0.001026086549233648, "compression/movement_sparsity/model_sparsity": 0.00099083731965603, "compression_loss": 0.2735157608985901, "distillation_loss": 0.6477317810058594, "epoch": 2.0, "learning_rate": 4.8137132122730125e-05, "loss": 0.909, "step": 2370, "task_loss": 0.729023814201355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0033745753263308975, "compression/movement_sparsity/importance_threshold": -0.006468243395854146, "compression/movement_sparsity/linear_layer_sparsity": 0.001023045886486503, "compression/movement_sparsity/model_sparsity": 0.0009879011130284221, "compression_loss": 0.3645848035812378, "distillation_loss": 0.42150405049324036, "epoch": 2.0, "learning_rate": 4.813400125234816e-05, "loss": 0.8189, "step": 2371, "task_loss": 0.7407252192497253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.004217031262003457, "compression/movement_sparsity/importance_threshold": -0.006462775734777773, "compression/movement_sparsity/linear_layer_sparsity": 0.0010291629844837006, "compression/movement_sparsity/model_sparsity": 0.0009938080698910215, "compression_loss": 0.45560222864151, "distillation_loss": 0.4441831111907959, "epoch": 2.01, "learning_rate": 4.813087038196619e-05, "loss": 1.0866, "step": 2372, "task_loss": 1.202262043952942 }, { "compression/movement_sparsity/importance_regularization_factor": 0.005059012307081834, "compression/movement_sparsity/importance_threshold": -0.006457311155810158, "compression/movement_sparsity/linear_layer_sparsity": 0.0010276247668586744, "compression/movement_sparsity/model_sparsity": 0.0009923226947735257, "compression_loss": 0.5465679168701172, "distillation_loss": 0.7204372882843018, "epoch": 2.01, "learning_rate": 4.812773951158423e-05, "loss": 1.1765, "step": 2373, "task_loss": 0.6989492774009705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.005900518595452042, "compression/movement_sparsity/importance_threshold": -0.00645184965808236, "compression/movement_sparsity/linear_layer_sparsity": 0.0010276247668586744, "compression/movement_sparsity/model_sparsity": 0.0009923226947735257, "compression_loss": 0.637482762336731, "distillation_loss": 0.31616097688674927, "epoch": 2.01, "learning_rate": 4.812460864120226e-05, "loss": 1.2026, "step": 2374, "task_loss": 0.11441697180271149 }, { "compression/movement_sparsity/importance_regularization_factor": 0.006741550260998652, "compression/movement_sparsity/importance_threshold": -0.00644639124072545, "compression/movement_sparsity/linear_layer_sparsity": 0.0010154144191995506, "compression/movement_sparsity/model_sparsity": 0.000980531810119916, "compression_loss": 0.7283458113670349, "distillation_loss": 0.5314382314682007, "epoch": 2.01, "learning_rate": 4.812147777082029e-05, "loss": 1.266, "step": 2375, "task_loss": 0.8511703610420227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.007582107437607011, "compression/movement_sparsity/importance_threshold": -0.006440935902870494, "compression/movement_sparsity/linear_layer_sparsity": 0.0010154263433671865, "compression/movement_sparsity/model_sparsity": 0.0009805433246557105, "compression_loss": 0.8191575407981873, "distillation_loss": 0.43945521116256714, "epoch": 2.01, "learning_rate": 4.811834690043832e-05, "loss": 1.296, "step": 2376, "task_loss": 0.6286712884902954 }, { "compression/movement_sparsity/importance_regularization_factor": 0.008422190259162132, "compression/movement_sparsity/importance_threshold": -0.006435483643648556, "compression/movement_sparsity/linear_layer_sparsity": 0.0010154263433671865, "compression/movement_sparsity/model_sparsity": 0.0009805433246557105, "compression_loss": 0.9099172353744507, "distillation_loss": 0.5057438611984253, "epoch": 2.01, "learning_rate": 4.811521603005636e-05, "loss": 1.5028, "step": 2377, "task_loss": 1.1899856328964233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.009261798859549142, "compression/movement_sparsity/importance_threshold": -0.006430034462190704, "compression/movement_sparsity/linear_layer_sparsity": 0.0010459522125149958, "compression/movement_sparsity/model_sparsity": 0.001010020536289735, "compression_loss": 1.000625729560852, "distillation_loss": 0.4951814115047455, "epoch": 2.01, "learning_rate": 4.811208515967439e-05, "loss": 1.5867, "step": 2378, "task_loss": 0.0590469054877758 }, { "compression/movement_sparsity/importance_regularization_factor": 0.010100933372653165, "compression/movement_sparsity/importance_threshold": -0.006424588357628004, "compression/movement_sparsity/linear_layer_sparsity": 0.0010444139948899696, "compression/movement_sparsity/model_sparsity": 0.0010085351611722393, "compression_loss": 1.0912823677062988, "distillation_loss": 0.49385765194892883, "epoch": 2.01, "learning_rate": 4.8108954289292425e-05, "loss": 1.7601, "step": 2379, "task_loss": 0.24165870249271393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.010939593932359548, "compression/movement_sparsity/importance_threshold": -0.00641914532909152, "compression/movement_sparsity/linear_layer_sparsity": 0.0010566124183814573, "compression/movement_sparsity/model_sparsity": 0.0010203145312900543, "compression_loss": 1.1818888187408447, "distillation_loss": 0.36515486240386963, "epoch": 2.01, "learning_rate": 4.810582341891046e-05, "loss": 1.5229, "step": 2380, "task_loss": 0.6111304759979248 }, { "compression/movement_sparsity/importance_regularization_factor": 0.011777780672553195, "compression/movement_sparsity/importance_threshold": -0.00641370537571232, "compression/movement_sparsity/linear_layer_sparsity": 0.001068822766040581, "compression/movement_sparsity/model_sparsity": 0.0010321054159436642, "compression_loss": 1.2724430561065674, "distillation_loss": 0.466522753238678, "epoch": 2.01, "learning_rate": 4.8102692548528495e-05, "loss": 1.8615, "step": 2381, "task_loss": 0.8660000562667847 }, { "compression/movement_sparsity/importance_regularization_factor": 0.012615493727119342, "compression/movement_sparsity/importance_threshold": -0.00640826849662147, "compression/movement_sparsity/linear_layer_sparsity": 0.0010871502116969025, "compression/movement_sparsity/model_sparsity": 0.0010498032574598733, "compression_loss": 1.3629459142684937, "distillation_loss": 0.45724570751190186, "epoch": 2.01, "learning_rate": 4.809956167814653e-05, "loss": 1.9242, "step": 2382, "task_loss": 0.9577873945236206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.013452733229942782, "compression/movement_sparsity/importance_threshold": -0.006402834690950036, "compression/movement_sparsity/linear_layer_sparsity": 0.0010993605593560263, "compression/movement_sparsity/model_sparsity": 0.0010615941421134832, "compression_loss": 1.4533970355987549, "distillation_loss": 0.5181804895401001, "epoch": 2.01, "learning_rate": 4.809643080776456e-05, "loss": 1.8853, "step": 2383, "task_loss": 0.2620887756347656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.014289499314909304, "compression/movement_sparsity/importance_threshold": -0.006397403957829083, "compression/movement_sparsity/linear_layer_sparsity": 0.0011298864285038356, "compression/movement_sparsity/model_sparsity": 0.0010910713537475075, "compression_loss": 1.5437968969345093, "distillation_loss": 0.9385466575622559, "epoch": 2.02, "learning_rate": 4.80932999373826e-05, "loss": 2.1159, "step": 2384, "task_loss": 1.0335040092468262 }, { "compression/movement_sparsity/importance_regularization_factor": 0.015125792115903702, "compression/movement_sparsity/importance_threshold": -0.006391976296389678, "compression/movement_sparsity/linear_layer_sparsity": 0.001102425070438443, "compression/movement_sparsity/model_sparsity": 0.0010645533778126802, "compression_loss": 1.6341437101364136, "distillation_loss": 0.37138110399246216, "epoch": 2.02, "learning_rate": 4.809016906700063e-05, "loss": 2.162, "step": 2385, "task_loss": 0.4602084755897522 }, { "compression/movement_sparsity/importance_regularization_factor": 0.015961611766811212, "compression/movement_sparsity/importance_threshold": -0.0063865517057628855, "compression/movement_sparsity/linear_layer_sparsity": 0.0010902266469469551, "compression/movement_sparsity/model_sparsity": 0.001052774007694865, "compression_loss": 1.7244389057159424, "distillation_loss": 0.6597011089324951, "epoch": 2.02, "learning_rate": 4.808703819661866e-05, "loss": 2.3298, "step": 2386, "task_loss": 0.7242158651351929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.01679695840151696, "compression/movement_sparsity/importance_threshold": -0.006381130185079772, "compression/movement_sparsity/linear_layer_sparsity": 0.001102436994606079, "compression/movement_sparsity/model_sparsity": 0.0010645648923484747, "compression_loss": 1.814683198928833, "distillation_loss": 0.47187626361846924, "epoch": 2.02, "learning_rate": 4.808390732623669e-05, "loss": 2.233, "step": 2387, "task_loss": 0.5295450687408447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.017631832153905957, "compression/movement_sparsity/importance_threshold": -0.006375711733471406, "compression/movement_sparsity/linear_layer_sparsity": 0.0011314246461288618, "compression/movement_sparsity/model_sparsity": 0.0010925567288650032, "compression_loss": 1.904875636100769, "distillation_loss": 0.3846927881240845, "epoch": 2.02, "learning_rate": 4.808077645585473e-05, "loss": 2.4401, "step": 2388, "task_loss": 0.429897278547287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.018466233157863443, "compression/movement_sparsity/importance_threshold": -0.006370296350068851, "compression/movement_sparsity/linear_layer_sparsity": 0.0011344891572112787, "compression/movement_sparsity/model_sparsity": 0.0010955159645642002, "compression_loss": 1.99501633644104, "distillation_loss": 0.30190205574035645, "epoch": 2.02, "learning_rate": 4.807764558547276e-05, "loss": 2.5459, "step": 2389, "task_loss": 0.7954344749450684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.01930016154727454, "compression/movement_sparsity/importance_threshold": -0.006364884034003174, "compression/movement_sparsity/linear_layer_sparsity": 0.0011482257983277929, "compression/movement_sparsity/model_sparsity": 0.0011087807097995112, "compression_loss": 2.0851047039031982, "distillation_loss": 0.47791004180908203, "epoch": 2.02, "learning_rate": 4.8074514715090795e-05, "loss": 2.6231, "step": 2390, "task_loss": 0.7972474098205566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.02013361745602449, "compression/movement_sparsity/importance_threshold": -0.00635947478440544, "compression/movement_sparsity/linear_layer_sparsity": 0.0011512903094102096, "compression/movement_sparsity/model_sparsity": 0.0011117399454987082, "compression_loss": 2.1751413345336914, "distillation_loss": 0.49011731147766113, "epoch": 2.02, "learning_rate": 4.807138384470883e-05, "loss": 2.7674, "step": 2391, "task_loss": 0.23868288099765778 }, { "compression/movement_sparsity/importance_regularization_factor": 0.020966601017998188, "compression/movement_sparsity/importance_threshold": -0.006354068600406717, "compression/movement_sparsity/linear_layer_sparsity": 0.0011545336830071644, "compression/movement_sparsity/model_sparsity": 0.0011148718992348235, "compression_loss": 2.265124797821045, "distillation_loss": 0.27632296085357666, "epoch": 2.02, "learning_rate": 4.8068252974326865e-05, "loss": 2.8008, "step": 2392, "task_loss": 0.5388728380203247 }, { "compression/movement_sparsity/importance_regularization_factor": 0.021799112367080653, "compression/movement_sparsity/importance_threshold": -0.006348665481138071, "compression/movement_sparsity/linear_layer_sparsity": 0.0011575862699219454, "compression/movement_sparsity/model_sparsity": 0.0011178196203982258, "compression_loss": 2.3550539016723633, "distillation_loss": 0.6467183232307434, "epoch": 2.02, "learning_rate": 4.80651221039449e-05, "loss": 3.0399, "step": 2393, "task_loss": 0.2825332283973694 }, { "compression/movement_sparsity/importance_regularization_factor": 0.022631151637157676, "compression/movement_sparsity/importance_threshold": -0.006343265425730565, "compression/movement_sparsity/linear_layer_sparsity": 0.001179717525054107, "compression/movement_sparsity/model_sparsity": 0.0011391905988328936, "compression_loss": 2.444932699203491, "distillation_loss": 0.6522690653800964, "epoch": 2.02, "learning_rate": 4.806199123356293e-05, "loss": 3.0138, "step": 2394, "task_loss": 1.2385218143463135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.023462718962113938, "compression/movement_sparsity/importance_threshold": -0.006337868433315268, "compression/movement_sparsity/linear_layer_sparsity": 0.0011801110225860905, "compression/movement_sparsity/model_sparsity": 0.0011395705785141135, "compression_loss": 2.5347588062286377, "distillation_loss": 0.732337474822998, "epoch": 2.02, "learning_rate": 4.805886036318097e-05, "loss": 3.1014, "step": 2395, "task_loss": 1.7141814231872559 }, { "compression/movement_sparsity/importance_regularization_factor": 0.024293814475834785, "compression/movement_sparsity/importance_threshold": -0.006332474503023244, "compression/movement_sparsity/linear_layer_sparsity": 0.0011947896729458379, "compression/movement_sparsity/model_sparsity": 0.0011537449720771932, "compression_loss": 2.624532699584961, "distillation_loss": 0.5605267286300659, "epoch": 2.03, "learning_rate": 4.8055729492799e-05, "loss": 3.1166, "step": 2396, "task_loss": 0.7844738364219666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.025124438312205122, "compression/movement_sparsity/importance_threshold": -0.00632708363398556, "compression/movement_sparsity/linear_layer_sparsity": 0.001201657993504095, "compression/movement_sparsity/model_sparsity": 0.0011603773446948486, "compression_loss": 2.714254140853882, "distillation_loss": 0.6501294374465942, "epoch": 2.03, "learning_rate": 4.805259862241703e-05, "loss": 3.2809, "step": 2397, "task_loss": 0.7354158759117126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.025954590605110184, "compression/movement_sparsity/importance_threshold": -0.006321695825333283, "compression/movement_sparsity/linear_layer_sparsity": 0.0012663466029286519, "compression/movement_sparsity/model_sparsity": 0.0012228437013802325, "compression_loss": 2.8039238452911377, "distillation_loss": 0.511442244052887, "epoch": 2.03, "learning_rate": 4.804946775203506e-05, "loss": 3.2274, "step": 2398, "task_loss": 1.3458755016326904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.026784271488435207, "compression/movement_sparsity/importance_threshold": -0.006316311076197477, "compression/movement_sparsity/linear_layer_sparsity": 0.001294201458526028, "compression/movement_sparsity/model_sparsity": 0.0012497416569962798, "compression_loss": 2.893541097640991, "distillation_loss": 0.8996846079826355, "epoch": 2.03, "learning_rate": 4.80463368816531e-05, "loss": 3.4432, "step": 2399, "task_loss": 1.267540693283081 }, { "compression/movement_sparsity/importance_regularization_factor": 0.027613481096065207, "compression/movement_sparsity/importance_threshold": -0.00631092938570921, "compression/movement_sparsity/linear_layer_sparsity": 0.0012924843783864636, "compression/movement_sparsity/model_sparsity": 0.0012480835638418659, "compression_loss": 2.983107566833496, "distillation_loss": 0.6550298929214478, "epoch": 2.03, "learning_rate": 4.804320601127113e-05, "loss": 3.6241, "step": 2400, "task_loss": 1.6606813669204712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.028442219561885307, "compression/movement_sparsity/importance_threshold": -0.006305550752999547, "compression/movement_sparsity/linear_layer_sparsity": 0.0013043131526812397, "compression/movement_sparsity/model_sparsity": 0.0012595059833500504, "compression_loss": 3.0726213455200195, "distillation_loss": 0.47700437903404236, "epoch": 2.03, "learning_rate": 4.804007514088917e-05, "loss": 3.6151, "step": 2401, "task_loss": 0.7379401326179504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.029270487019780744, "compression/movement_sparsity/importance_threshold": -0.006300175177199554, "compression/movement_sparsity/linear_layer_sparsity": 0.0013535361166820822, "compression/movement_sparsity/model_sparsity": 0.0013070379871099147, "compression_loss": 3.1620829105377197, "distillation_loss": 0.7665954828262329, "epoch": 2.03, "learning_rate": 4.80369442705072e-05, "loss": 3.7294, "step": 2402, "task_loss": 0.8929296135902405 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03009828360363631, "compression/movement_sparsity/importance_threshold": -0.0062948026574403005, "compression/movement_sparsity/linear_layer_sparsity": 0.0013716489273209581, "compression/movement_sparsity/model_sparsity": 0.0013245285669818221, "compression_loss": 3.2514920234680176, "distillation_loss": 0.3567678928375244, "epoch": 2.03, "learning_rate": 4.8033813400125235e-05, "loss": 3.689, "step": 2403, "task_loss": 0.3630422353744507 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03092560944733791, "compression/movement_sparsity/importance_threshold": -0.006289433192852846, "compression/movement_sparsity/linear_layer_sparsity": 0.0013989314228718126, "compression/movement_sparsity/model_sparsity": 0.0013508738248797316, "compression_loss": 3.3408493995666504, "distillation_loss": 0.6617258787155151, "epoch": 2.03, "learning_rate": 4.8030682529743273e-05, "loss": 3.9214, "step": 2404, "task_loss": 1.0481939315795898 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03175246468477011, "compression/movement_sparsity/importance_threshold": -0.00628406678256826, "compression/movement_sparsity/linear_layer_sparsity": 0.0014426334972572506, "compression/movement_sparsity/model_sparsity": 0.0013930745985667235, "compression_loss": 3.4301517009735107, "distillation_loss": 0.477468878030777, "epoch": 2.03, "learning_rate": 4.8027551659361305e-05, "loss": 3.8853, "step": 2405, "task_loss": 0.30540746450424194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03257884944981815, "compression/movement_sparsity/importance_threshold": -0.00627870342571761, "compression/movement_sparsity/linear_layer_sparsity": 0.0015202836769019905, "compression/movement_sparsity/model_sparsity": 0.0014680572556607732, "compression_loss": 3.519404649734497, "distillation_loss": 0.5837590098381042, "epoch": 2.03, "learning_rate": 4.8024420788979344e-05, "loss": 3.9875, "step": 2406, "task_loss": 0.4039652347564697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03340476387636726, "compression/movement_sparsity/importance_threshold": -0.006273343121431958, "compression/movement_sparsity/linear_layer_sparsity": 0.0015372636916154594, "compression/movement_sparsity/model_sparsity": 0.0014844539546321994, "compression_loss": 3.6086039543151855, "distillation_loss": 0.2445085644721985, "epoch": 2.03, "learning_rate": 4.8021289918597376e-05, "loss": 4.0699, "step": 2407, "task_loss": 0.9099907875061035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03423020809830246, "compression/movement_sparsity/importance_threshold": -0.006267985868842374, "compression/movement_sparsity/linear_layer_sparsity": 0.0015849603621589116, "compression/movement_sparsity/model_sparsity": 0.0015305120978103624, "compression_loss": 3.697749376296997, "distillation_loss": 0.6575204133987427, "epoch": 2.04, "learning_rate": 4.801815904821541e-05, "loss": 4.2396, "step": 2408, "task_loss": 1.1214654445648193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03505518224950899, "compression/movement_sparsity/importance_threshold": -0.006262631667079923, "compression/movement_sparsity/linear_layer_sparsity": 0.0016112889242988972, "compression/movement_sparsity/model_sparsity": 0.0015559361928447085, "compression_loss": 3.7868428230285645, "distillation_loss": 0.42375025153160095, "epoch": 2.04, "learning_rate": 4.801502817783344e-05, "loss": 4.2447, "step": 2409, "task_loss": 0.20064783096313477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03587968646387196, "compression/movement_sparsity/importance_threshold": -0.006257280515275669, "compression/movement_sparsity/linear_layer_sparsity": 0.001641063570885647, "compression/movement_sparsity/model_sparsity": 0.001584687988723677, "compression_loss": 3.875885248184204, "distillation_loss": 0.5622499585151672, "epoch": 2.04, "learning_rate": 4.801189730745148e-05, "loss": 4.3153, "step": 2410, "task_loss": 0.573811948299408 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0367037208752764, "compression/movement_sparsity/importance_threshold": -0.006251932412560681, "compression/movement_sparsity/linear_layer_sparsity": 0.0016633856126999826, "compression/movement_sparsity/model_sparsity": 0.0016062431997310573, "compression_loss": 3.9648735523223877, "distillation_loss": 0.3149120807647705, "epoch": 2.04, "learning_rate": 4.800876643706951e-05, "loss": 4.4999, "step": 2411, "task_loss": 1.0052573680877686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03752728561760765, "compression/movement_sparsity/importance_threshold": -0.0062465873580660235, "compression/movement_sparsity/linear_layer_sparsity": 0.0017370531203543442, "compression/movement_sparsity/model_sparsity": 0.0016773800018697303, "compression_loss": 4.053811550140381, "distillation_loss": 0.6441569328308105, "epoch": 2.04, "learning_rate": 4.800563556668754e-05, "loss": 4.5, "step": 2412, "task_loss": 0.4948197901248932 }, { "compression/movement_sparsity/importance_regularization_factor": 0.038350380824750396, "compression/movement_sparsity/importance_threshold": -0.006241245350922764, "compression/movement_sparsity/linear_layer_sparsity": 0.001787039231083882, "compression/movement_sparsity/model_sparsity": 0.0017256489359204453, "compression_loss": 4.142693996429443, "distillation_loss": 0.3789566159248352, "epoch": 2.04, "learning_rate": 4.800250469630558e-05, "loss": 4.7829, "step": 2413, "task_loss": 0.3423377573490143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03917300663059042, "compression/movement_sparsity/importance_threshold": -0.006235906390261966, "compression/movement_sparsity/linear_layer_sparsity": 0.0018003942988360487, "compression/movement_sparsity/model_sparsity": 0.0017385452160103311, "compression_loss": 4.231527805328369, "distillation_loss": 0.9483208656311035, "epoch": 2.04, "learning_rate": 4.799937382592361e-05, "loss": 4.7902, "step": 2414, "task_loss": 0.8575778007507324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.03999516316901264, "compression/movement_sparsity/importance_threshold": -0.006230570475214696, "compression/movement_sparsity/linear_layer_sparsity": 0.0018412107246536078, "compression/movement_sparsity/model_sparsity": 0.0017779594720350443, "compression_loss": 4.3203043937683105, "distillation_loss": 0.5988214612007141, "epoch": 2.04, "learning_rate": 4.7996242955541643e-05, "loss": 4.8452, "step": 2415, "task_loss": 0.4690380394458771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04081685057390205, "compression/movement_sparsity/importance_threshold": -0.006225237604912022, "compression/movement_sparsity/linear_layer_sparsity": 0.0019219015670454928, "compression/movement_sparsity/model_sparsity": 0.0018558783357567018, "compression_loss": 4.409032344818115, "distillation_loss": 0.471515417098999, "epoch": 2.04, "learning_rate": 4.7993112085159675e-05, "loss": 5.0725, "step": 2416, "task_loss": 0.7680541276931763 }, { "compression/movement_sparsity/importance_regularization_factor": 0.041638068979143905, "compression/movement_sparsity/importance_threshold": -0.0062199077784850084, "compression/movement_sparsity/linear_layer_sparsity": 0.001932394834565052, "compression/movement_sparsity/model_sparsity": 0.0018660111272558975, "compression_loss": 4.497708320617676, "distillation_loss": 0.4816495478153229, "epoch": 2.04, "learning_rate": 4.7989981214777714e-05, "loss": 5.0595, "step": 2417, "task_loss": 0.9911508560180664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04245881851862332, "compression/movement_sparsity/importance_threshold": -0.0062145809950647215, "compression/movement_sparsity/linear_layer_sparsity": 0.0019272435941463594, "compression/movement_sparsity/model_sparsity": 0.001861036847792656, "compression_loss": 4.58632755279541, "distillation_loss": 0.434975802898407, "epoch": 2.04, "learning_rate": 4.7986850344395746e-05, "loss": 5.2836, "step": 2418, "task_loss": 0.3383782207965851 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04327909932622542, "compression/movement_sparsity/importance_threshold": -0.006209257253782227, "compression/movement_sparsity/linear_layer_sparsity": 0.0019682865791489997, "compression/movement_sparsity/model_sparsity": 0.0019006698799974653, "compression_loss": 4.674900531768799, "distillation_loss": 0.5818958282470703, "epoch": 2.04, "learning_rate": 4.798371947401378e-05, "loss": 5.3117, "step": 2419, "task_loss": 0.2089730054140091 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04409891153583545, "compression/movement_sparsity/importance_threshold": -0.006203936553768592, "compression/movement_sparsity/linear_layer_sparsity": 0.0020392473007500205, "compression/movement_sparsity/model_sparsity": 0.001969192882510778, "compression_loss": 4.763415813446045, "distillation_loss": 0.6268411874771118, "epoch": 2.05, "learning_rate": 4.798058860363181e-05, "loss": 5.3723, "step": 2420, "task_loss": 0.8575180768966675 }, { "compression/movement_sparsity/importance_regularization_factor": 0.044918255281338304, "compression/movement_sparsity/importance_threshold": -0.0061986188941548824, "compression/movement_sparsity/linear_layer_sparsity": 0.002085429602003718, "compression/movement_sparsity/model_sparsity": 0.002013788679643034, "compression_loss": 4.8518805503845215, "distillation_loss": 0.4193408191204071, "epoch": 2.05, "learning_rate": 4.797745773324985e-05, "loss": 5.3346, "step": 2421, "task_loss": 0.7247567176818848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04573713069661933, "compression/movement_sparsity/importance_threshold": -0.006193304274072164, "compression/movement_sparsity/linear_layer_sparsity": 0.0021653692218345437, "compression/movement_sparsity/model_sparsity": 0.002090982127609636, "compression_loss": 4.940296173095703, "distillation_loss": 0.5909290313720703, "epoch": 2.05, "learning_rate": 4.797432686286788e-05, "loss": 5.3759, "step": 2422, "task_loss": 1.589556097984314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04655553791556333, "compression/movement_sparsity/importance_threshold": -0.006187992692651504, "compression/movement_sparsity/linear_layer_sparsity": 0.0021813953031371435, "compression/movement_sparsity/model_sparsity": 0.0021064576637174984, "compression_loss": 5.028654098510742, "distillation_loss": 0.6817412376403809, "epoch": 2.05, "learning_rate": 4.797119599248591e-05, "loss": 5.558, "step": 2423, "task_loss": 0.48237931728363037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04737347707205597, "compression/movement_sparsity/importance_threshold": -0.006182684149023966, "compression/movement_sparsity/linear_layer_sparsity": 0.0022193857012250033, "compression/movement_sparsity/model_sparsity": 0.0021431429747589053, "compression_loss": 5.116967678070068, "distillation_loss": 0.3739989995956421, "epoch": 2.05, "learning_rate": 4.796806512210394e-05, "loss": 5.7049, "step": 2424, "task_loss": 0.393751859664917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04819094829998227, "compression/movement_sparsity/importance_threshold": -0.006177378642320616, "compression/movement_sparsity/linear_layer_sparsity": 0.002269181025272367, "compression/movement_sparsity/model_sparsity": 0.0021912276762369076, "compression_loss": 5.205223083496094, "distillation_loss": 0.9563711881637573, "epoch": 2.05, "learning_rate": 4.796493425172198e-05, "loss": 5.9243, "step": 2425, "task_loss": 0.9983874559402466 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04900795173322714, "compression/movement_sparsity/importance_threshold": -0.006172076171672522, "compression/movement_sparsity/linear_layer_sparsity": 0.0023161383974223955, "compression/movement_sparsity/model_sparsity": 0.0022365719181958094, "compression_loss": 5.2934250831604, "distillation_loss": 0.7592760324478149, "epoch": 2.05, "learning_rate": 4.796180338134001e-05, "loss": 5.896, "step": 2426, "task_loss": 0.27847200632095337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.04982448750567581, "compression/movement_sparsity/importance_threshold": -0.006166776736210749, "compression/movement_sparsity/linear_layer_sparsity": 0.002358874614229329, "compression/movement_sparsity/model_sparsity": 0.0022778400144834437, "compression_loss": 5.381577968597412, "distillation_loss": 0.46952933073043823, "epoch": 2.05, "learning_rate": 4.7958672510958045e-05, "loss": 5.7777, "step": 2427, "task_loss": 0.46050283312797546 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05064055575121351, "compression/movement_sparsity/importance_threshold": -0.006161480335066364, "compression/movement_sparsity/linear_layer_sparsity": 0.002401646603539169, "compression/movement_sparsity/model_sparsity": 0.0023191426543784617, "compression_loss": 5.469667911529541, "distillation_loss": 0.517757773399353, "epoch": 2.05, "learning_rate": 4.7955541640576084e-05, "loss": 5.9116, "step": 2428, "task_loss": 0.23131200671195984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05145615660372527, "compression/movement_sparsity/importance_threshold": -0.006156186967370431, "compression/movement_sparsity/linear_layer_sparsity": 0.002518038403832828, "compression/movement_sparsity/model_sparsity": 0.002431536038268974, "compression_loss": 5.557706832885742, "distillation_loss": 0.5169776678085327, "epoch": 2.05, "learning_rate": 4.7952410770194115e-05, "loss": 6.0982, "step": 2429, "task_loss": 0.6633235216140747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05227129019709631, "compression/movement_sparsity/importance_threshold": -0.0061508966322540186, "compression/movement_sparsity/linear_layer_sparsity": 0.0026288854661758107, "compression/movement_sparsity/model_sparsity": 0.0025385751630150255, "compression_loss": 5.645697593688965, "distillation_loss": 0.5901542901992798, "epoch": 2.05, "learning_rate": 4.794927989981215e-05, "loss": 6.1831, "step": 2430, "task_loss": 1.0530812740325928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05308595666521165, "compression/movement_sparsity/importance_threshold": -0.0061456093288481915, "compression/movement_sparsity/linear_layer_sparsity": 0.0026899372044714293, "compression/movement_sparsity/model_sparsity": 0.0025975295862830745, "compression_loss": 5.733635425567627, "distillation_loss": 0.4503580927848816, "epoch": 2.05, "learning_rate": 4.794614902943018e-05, "loss": 6.2261, "step": 2431, "task_loss": 0.5380730032920837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.053900156141956646, "compression/movement_sparsity/importance_threshold": -0.006140325056284015, "compression/movement_sparsity/linear_layer_sparsity": 0.0027723689753381504, "compression/movement_sparsity/model_sparsity": 0.002677129572230735, "compression_loss": 5.821515083312988, "distillation_loss": 0.6026744842529297, "epoch": 2.06, "learning_rate": 4.794301815904822e-05, "loss": 6.3913, "step": 2432, "task_loss": 0.7269959449768066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.054713888761215856, "compression/movement_sparsity/importance_threshold": -0.006135043813692559, "compression/movement_sparsity/linear_layer_sparsity": 0.0028618479292776664, "compression/movement_sparsity/model_sparsity": 0.002763534648832969, "compression_loss": 5.909345626831055, "distillation_loss": 0.5396760106086731, "epoch": 2.06, "learning_rate": 4.793988728866625e-05, "loss": 6.4972, "step": 2433, "task_loss": 1.1610075235366821 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0555271546568753, "compression/movement_sparsity/importance_threshold": -0.006129765600204884, "compression/movement_sparsity/linear_layer_sparsity": 0.0029629767949974207, "compression/movement_sparsity/model_sparsity": 0.0028611894269064695, "compression_loss": 5.997123718261719, "distillation_loss": 0.5538237690925598, "epoch": 2.06, "learning_rate": 4.793675641828428e-05, "loss": 6.4959, "step": 2434, "task_loss": 0.7483657002449036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.056339953962819544, "compression/movement_sparsity/importance_threshold": -0.0061244904149520595, "compression/movement_sparsity/linear_layer_sparsity": 0.00305685576679457, "compression/movement_sparsity/model_sparsity": 0.0029518433672168893, "compression_loss": 6.084847927093506, "distillation_loss": 0.5863285660743713, "epoch": 2.06, "learning_rate": 4.793362554790232e-05, "loss": 6.6904, "step": 2435, "task_loss": 1.1088011264801025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05715228681293394, "compression/movement_sparsity/importance_threshold": -0.00611921825706515, "compression/movement_sparsity/linear_layer_sparsity": 0.003111969269607529, "compression/movement_sparsity/model_sparsity": 0.0030050635516592566, "compression_loss": 6.172518730163574, "distillation_loss": 0.8408394455909729, "epoch": 2.06, "learning_rate": 4.793049467752035e-05, "loss": 6.7382, "step": 2436, "task_loss": 0.8827014565467834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05796415334110361, "compression/movement_sparsity/importance_threshold": -0.0061139491256752216, "compression/movement_sparsity/linear_layer_sparsity": 0.0031522252595462025, "compression/movement_sparsity/model_sparsity": 0.0030439366245016263, "compression_loss": 6.26013708114624, "distillation_loss": 0.7559710741043091, "epoch": 2.06, "learning_rate": 4.792736380713839e-05, "loss": 6.8958, "step": 2437, "task_loss": 1.2365788221359253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.058775553681213566, "compression/movement_sparsity/importance_threshold": -0.006108683019913342, "compression/movement_sparsity/linear_layer_sparsity": 0.0032914995375330825, "compression/movement_sparsity/model_sparsity": 0.003178426402581863, "compression_loss": 6.347705364227295, "distillation_loss": 0.5755919218063354, "epoch": 2.06, "learning_rate": 4.792423293675642e-05, "loss": 6.9174, "step": 2438, "task_loss": 0.7407458424568176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.05958648796714905, "compression/movement_sparsity/importance_threshold": -0.006103419938910576, "compression/movement_sparsity/linear_layer_sparsity": 0.0033758153268862697, "compression/movement_sparsity/model_sparsity": 0.003259845685185061, "compression_loss": 6.435220718383789, "distillation_loss": 0.60747230052948, "epoch": 2.06, "learning_rate": 4.7921102066374454e-05, "loss": 7.0963, "step": 2439, "task_loss": 0.6079419851303101 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06039695633279518, "compression/movement_sparsity/importance_threshold": -0.0060981598817979885, "compression/movement_sparsity/linear_layer_sparsity": 0.003479794068670995, "compression/movement_sparsity/model_sparsity": 0.0033602524373134565, "compression_loss": 6.5226874351501465, "distillation_loss": 0.3900158107280731, "epoch": 2.06, "learning_rate": 4.791797119599249e-05, "loss": 7.0396, "step": 2440, "task_loss": 0.34926968812942505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06120695891203709, "compression/movement_sparsity/importance_threshold": -0.006092902847706648, "compression/movement_sparsity/linear_layer_sparsity": 0.0036194618441898587, "compression/movement_sparsity/model_sparsity": 0.003495122195074913, "compression_loss": 6.610100746154785, "distillation_loss": 0.2940852642059326, "epoch": 2.06, "learning_rate": 4.7914840325610524e-05, "loss": 6.9432, "step": 2441, "task_loss": 0.11611046642065048 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06201649583876001, "compression/movement_sparsity/importance_threshold": -0.006087648835767619, "compression/movement_sparsity/linear_layer_sparsity": 0.003754550739336551, "compression/movement_sparsity/model_sparsity": 0.0036255703710912656, "compression_loss": 6.697459697723389, "distillation_loss": 0.4436652660369873, "epoch": 2.06, "learning_rate": 4.7911709455228556e-05, "loss": 7.2697, "step": 2442, "task_loss": 0.6699572205543518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06282556724684862, "compression/movement_sparsity/importance_threshold": -0.00608239784511197, "compression/movement_sparsity/linear_layer_sparsity": 0.0038858239008397666, "compression/movement_sparsity/model_sparsity": 0.0037523338956533653, "compression_loss": 6.7847676277160645, "distillation_loss": 0.5506840944290161, "epoch": 2.07, "learning_rate": 4.7908578584846594e-05, "loss": 7.2007, "step": 2443, "task_loss": 0.25510552525520325 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06363417327018872, "compression/movement_sparsity/importance_threshold": -0.006077149874870763, "compression/movement_sparsity/linear_layer_sparsity": 0.004041696620175768, "compression/movement_sparsity/model_sparsity": 0.0039028519075596027, "compression_loss": 6.872019290924072, "distillation_loss": 0.8852063417434692, "epoch": 2.07, "learning_rate": 4.7905447714464626e-05, "loss": 7.5125, "step": 2444, "task_loss": 1.067788004875183 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0644423140426652, "compression/movement_sparsity/importance_threshold": -0.006071904924175066, "compression/movement_sparsity/linear_layer_sparsity": 0.004155035833554646, "compression/movement_sparsity/model_sparsity": 0.0040122975702867126, "compression_loss": 6.959228515625, "distillation_loss": 0.5758681297302246, "epoch": 2.07, "learning_rate": 4.790231684408266e-05, "loss": 7.5146, "step": 2445, "task_loss": 0.21737156808376312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0652499896981632, "compression/movement_sparsity/importance_threshold": -0.006066662992155945, "compression/movement_sparsity/linear_layer_sparsity": 0.004265501322533281, "compression/movement_sparsity/model_sparsity": 0.004118968229887339, "compression_loss": 7.046379566192627, "distillation_loss": 0.42602860927581787, "epoch": 2.07, "learning_rate": 4.789918597370069e-05, "loss": 7.6664, "step": 2446, "task_loss": 0.4365503191947937 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06605720037056773, "compression/movement_sparsity/importance_threshold": -0.006061424077944467, "compression/movement_sparsity/linear_layer_sparsity": 0.004424247766269525, "compression/movement_sparsity/model_sparsity": 0.00427226124492006, "compression_loss": 7.133483409881592, "distillation_loss": 0.5850775837898254, "epoch": 2.07, "learning_rate": 4.789605510331873e-05, "loss": 7.5976, "step": 2447, "task_loss": 0.3459823727607727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06686394619376412, "compression/movement_sparsity/importance_threshold": -0.006056188180671696, "compression/movement_sparsity/linear_layer_sparsity": 0.004601882091540976, "compression/movement_sparsity/model_sparsity": 0.004443793284651334, "compression_loss": 7.220536708831787, "distillation_loss": 0.8706289529800415, "epoch": 2.07, "learning_rate": 4.789292423293676e-05, "loss": 7.8351, "step": 2448, "task_loss": 1.0656594038009644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06767022730163741, "compression/movement_sparsity/importance_threshold": -0.006050955299468699, "compression/movement_sparsity/linear_layer_sparsity": 0.004706230482522414, "compression/movement_sparsity/model_sparsity": 0.004544556987389361, "compression_loss": 7.307534694671631, "distillation_loss": 0.7764495611190796, "epoch": 2.07, "learning_rate": 4.788979336255479e-05, "loss": 7.8267, "step": 2449, "task_loss": 1.154152274131775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06847604382807271, "compression/movement_sparsity/importance_threshold": -0.006045725433466543, "compression/movement_sparsity/linear_layer_sparsity": 0.004818019554108629, "compression/movement_sparsity/model_sparsity": 0.004652505760463181, "compression_loss": 7.394482135772705, "distillation_loss": 0.4880790114402771, "epoch": 2.07, "learning_rate": 4.788666249217283e-05, "loss": 7.9235, "step": 2450, "task_loss": 0.9652631282806396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.06928139590695526, "compression/movement_sparsity/importance_threshold": -0.006040498581796293, "compression/movement_sparsity/linear_layer_sparsity": 0.004928127318058188, "compression/movement_sparsity/model_sparsity": 0.004758830983989971, "compression_loss": 7.481370449066162, "distillation_loss": 0.520179808139801, "epoch": 2.07, "learning_rate": 4.788353162179086e-05, "loss": 7.8684, "step": 2451, "task_loss": 0.2602771818637848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07008628367217007, "compression/movement_sparsity/importance_threshold": -0.006035274743589015, "compression/movement_sparsity/linear_layer_sparsity": 0.005076547432621775, "compression/movement_sparsity/model_sparsity": 0.00490215241102462, "compression_loss": 7.568206787109375, "distillation_loss": 0.2904081344604492, "epoch": 2.07, "learning_rate": 4.7880400751408894e-05, "loss": 7.9069, "step": 2452, "task_loss": 0.6402894854545593 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07089070725760216, "compression/movement_sparsity/importance_threshold": -0.0060300539179757775, "compression/movement_sparsity/linear_layer_sparsity": 0.005272497279381912, "compression/movement_sparsity/model_sparsity": 0.005091370777736309, "compression_loss": 7.654984951019287, "distillation_loss": 0.41537243127822876, "epoch": 2.07, "learning_rate": 4.7877269881026926e-05, "loss": 8.1827, "step": 2453, "task_loss": 0.8969612121582031 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0716946667971371, "compression/movement_sparsity/importance_threshold": -0.006024836104087643, "compression/movement_sparsity/linear_layer_sparsity": 0.005337352827153371, "compression/movement_sparsity/model_sparsity": 0.005153998337922816, "compression_loss": 7.7417097091674805, "distillation_loss": 0.44362667202949524, "epoch": 2.07, "learning_rate": 4.7874139010644964e-05, "loss": 8.2086, "step": 2454, "task_loss": 0.43033918738365173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07249816242465978, "compression/movement_sparsity/importance_threshold": -0.006019621301055678, "compression/movement_sparsity/linear_layer_sparsity": 0.005521414278780553, "compression/movement_sparsity/model_sparsity": 0.005331736712447348, "compression_loss": 7.828378200531006, "distillation_loss": 0.3020409643650055, "epoch": 2.08, "learning_rate": 4.7871008140262996e-05, "loss": 8.2991, "step": 2455, "task_loss": 0.5633275508880615 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07330119427405535, "compression/movement_sparsity/importance_threshold": -0.006014409508010949, "compression/movement_sparsity/linear_layer_sparsity": 0.0057202020774380245, "compression/movement_sparsity/model_sparsity": 0.005523695538678137, "compression_loss": 7.914995193481445, "distillation_loss": 0.5233095288276672, "epoch": 2.08, "learning_rate": 4.786787726988103e-05, "loss": 8.4755, "step": 2456, "task_loss": 1.0775245428085327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07410376247920902, "compression/movement_sparsity/importance_threshold": -0.006009200724084523, "compression/movement_sparsity/linear_layer_sparsity": 0.005863864449114902, "compression/movement_sparsity/model_sparsity": 0.005662422665930765, "compression_loss": 8.001562118530273, "distillation_loss": 0.7105423212051392, "epoch": 2.08, "learning_rate": 4.786474639949906e-05, "loss": 8.5267, "step": 2457, "task_loss": 1.055181622505188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07490586717400582, "compression/movement_sparsity/importance_threshold": -0.006003994948407466, "compression/movement_sparsity/linear_layer_sparsity": 0.005977012875811607, "compression/movement_sparsity/model_sparsity": 0.005771684096085162, "compression_loss": 8.088078498840332, "distillation_loss": 0.488961398601532, "epoch": 2.08, "learning_rate": 4.78616155291171e-05, "loss": 8.6397, "step": 2458, "task_loss": 0.45086681842803955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07570750849233099, "compression/movement_sparsity/importance_threshold": -0.005998792180110843, "compression/movement_sparsity/linear_layer_sparsity": 0.006235147256792769, "compression/movement_sparsity/model_sparsity": 0.006020950766965382, "compression_loss": 8.174543380737305, "distillation_loss": 0.37834522128105164, "epoch": 2.08, "learning_rate": 4.785848465873513e-05, "loss": 8.6698, "step": 2459, "task_loss": 0.5697919726371765 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07650868656806953, "compression/movement_sparsity/importance_threshold": -0.005993592418325721, "compression/movement_sparsity/linear_layer_sparsity": 0.006426112801481115, "compression/movement_sparsity/model_sparsity": 0.006205356057714952, "compression_loss": 8.260957717895508, "distillation_loss": 0.5034921765327454, "epoch": 2.08, "learning_rate": 4.785535378835316e-05, "loss": 8.8814, "step": 2460, "task_loss": 1.0568251609802246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07730940153510668, "compression/movement_sparsity/importance_threshold": -0.0059883956621831665, "compression/movement_sparsity/linear_layer_sparsity": 0.006621669150709268, "compression/movement_sparsity/model_sparsity": 0.006394194444745422, "compression_loss": 8.34732437133789, "distillation_loss": 0.556475043296814, "epoch": 2.08, "learning_rate": 4.7852222917971193e-05, "loss": 8.9458, "step": 2461, "task_loss": 0.39032644033432007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07810965352732757, "compression/movement_sparsity/importance_threshold": -0.005983201910814244, "compression/movement_sparsity/linear_layer_sparsity": 0.0068326792211935005, "compression/movement_sparsity/model_sparsity": 0.006597955670165616, "compression_loss": 8.433637619018555, "distillation_loss": 0.3479404151439667, "epoch": 2.08, "learning_rate": 4.784909204758923e-05, "loss": 9.0909, "step": 2462, "task_loss": 0.27851831912994385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0789094426786171, "compression/movement_sparsity/importance_threshold": -0.005978011163350021, "compression/movement_sparsity/linear_layer_sparsity": 0.006991807238294092, "compression/movement_sparsity/model_sparsity": 0.006751617150343762, "compression_loss": 8.519902229309082, "distillation_loss": 0.48571208119392395, "epoch": 2.08, "learning_rate": 4.7845961177207264e-05, "loss": 9.1818, "step": 2463, "task_loss": 0.5385258197784424 }, { "compression/movement_sparsity/importance_regularization_factor": 0.07970876912286096, "compression/movement_sparsity/importance_threshold": -0.005972823418921562, "compression/movement_sparsity/linear_layer_sparsity": 0.007201314863656205, "compression/movement_sparsity/model_sparsity": 0.006953927544253845, "compression_loss": 8.606112480163574, "distillation_loss": 0.8172193765640259, "epoch": 2.08, "learning_rate": 4.7842830306825296e-05, "loss": 9.0918, "step": 2464, "task_loss": 0.22411011159420013 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08050763299394392, "compression/movement_sparsity/importance_threshold": -0.005967638676659934, "compression/movement_sparsity/linear_layer_sparsity": 0.0074367694777939565, "compression/movement_sparsity/model_sparsity": 0.007181293568052847, "compression_loss": 8.692278861999512, "distillation_loss": 0.3886789083480835, "epoch": 2.08, "learning_rate": 4.7839699436443334e-05, "loss": 9.2494, "step": 2465, "task_loss": 0.11984727531671524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08130603442575124, "compression/movement_sparsity/importance_threshold": -0.005962456935696202, "compression/movement_sparsity/linear_layer_sparsity": 0.007652179566135822, "compression/movement_sparsity/model_sparsity": 0.007389303657181226, "compression_loss": 8.778374671936035, "distillation_loss": 0.799217939376831, "epoch": 2.08, "learning_rate": 4.7836568566061366e-05, "loss": 9.2798, "step": 2466, "task_loss": 0.7339923977851868 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08210397355216792, "compression/movement_sparsity/importance_threshold": -0.005957278195161433, "compression/movement_sparsity/linear_layer_sparsity": 0.007971580320430048, "compression/movement_sparsity/model_sparsity": 0.007697732012973796, "compression_loss": 8.864422798156738, "distillation_loss": 0.41805076599121094, "epoch": 2.09, "learning_rate": 4.78334376956794e-05, "loss": 9.4919, "step": 2467, "task_loss": 0.8373850584030151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0829014505070792, "compression/movement_sparsity/importance_threshold": -0.005952102454186693, "compression/movement_sparsity/linear_layer_sparsity": 0.008266333820220946, "compression/movement_sparsity/model_sparsity": 0.00798235982327905, "compression_loss": 8.95042610168457, "distillation_loss": 0.5397064685821533, "epoch": 2.09, "learning_rate": 4.7830306825297436e-05, "loss": 9.5097, "step": 2468, "task_loss": 1.4559286832809448 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0836984654243702, "compression/movement_sparsity/importance_threshold": -0.005946929711903049, "compression/movement_sparsity/linear_layer_sparsity": 0.008385957069943923, "compression/movement_sparsity/model_sparsity": 0.008097873646369883, "compression_loss": 9.036380767822266, "distillation_loss": 0.5260220766067505, "epoch": 2.09, "learning_rate": 4.782717595491547e-05, "loss": 9.5095, "step": 2469, "task_loss": 0.30859607458114624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08449501843792617, "compression/movement_sparsity/importance_threshold": -0.005941759967441565, "compression/movement_sparsity/linear_layer_sparsity": 0.008700552384680899, "compression/movement_sparsity/model_sparsity": 0.008401661644237253, "compression_loss": 9.12227725982666, "distillation_loss": 0.8130235075950623, "epoch": 2.09, "learning_rate": 4.78240450845335e-05, "loss": 9.7203, "step": 2470, "task_loss": 0.9196327328681946 }, { "compression/movement_sparsity/importance_regularization_factor": 0.085291109681632, "compression/movement_sparsity/importance_threshold": -0.005936593219933308, "compression/movement_sparsity/linear_layer_sparsity": 0.00899820345720731, "compression/movement_sparsity/model_sparsity": 0.00868908748674058, "compression_loss": 9.208130836486816, "distillation_loss": 0.49410247802734375, "epoch": 2.09, "learning_rate": 4.782091421415154e-05, "loss": 9.7591, "step": 2471, "task_loss": 1.1558340787887573 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08608673928937305, "compression/movement_sparsity/importance_threshold": -0.005931429468509344, "compression/movement_sparsity/linear_layer_sparsity": 0.009176767867554359, "compression/movement_sparsity/model_sparsity": 0.008861517660263829, "compression_loss": 9.29392147064209, "distillation_loss": 1.2711620330810547, "epoch": 2.09, "learning_rate": 4.781778334376957e-05, "loss": 10.1621, "step": 2472, "task_loss": 0.5127913951873779 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08688190739503443, "compression/movement_sparsity/importance_threshold": -0.00592626871230074, "compression/movement_sparsity/linear_layer_sparsity": 0.009390425103253752, "compression/movement_sparsity/model_sparsity": 0.00906783511263041, "compression_loss": 9.379671096801758, "distillation_loss": 0.4633873403072357, "epoch": 2.09, "learning_rate": 4.781465247338761e-05, "loss": 9.8982, "step": 2473, "task_loss": 0.8557053804397583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08767661413250083, "compression/movement_sparsity/importance_threshold": -0.005921110950438563, "compression/movement_sparsity/linear_layer_sparsity": 0.009641130727797773, "compression/movement_sparsity/model_sparsity": 0.00930992822771063, "compression_loss": 9.46536636352539, "distillation_loss": 0.35170578956604004, "epoch": 2.09, "learning_rate": 4.781152160300564e-05, "loss": 10.0273, "step": 2474, "task_loss": 0.6469836235046387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.08847085963565826, "compression/movement_sparsity/importance_threshold": -0.005915956182053874, "compression/movement_sparsity/linear_layer_sparsity": 0.009900779478068689, "compression/movement_sparsity/model_sparsity": 0.009560657244636757, "compression_loss": 9.551017761230469, "distillation_loss": 0.5214751958847046, "epoch": 2.09, "learning_rate": 4.780839073262367e-05, "loss": 10.3271, "step": 2475, "task_loss": 0.8135666847229004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.0892646440383913, "compression/movement_sparsity/importance_threshold": -0.005910804406277743, "compression/movement_sparsity/linear_layer_sparsity": 0.010139083968271411, "compression/movement_sparsity/model_sparsity": 0.009790775242490654, "compression_loss": 9.63662052154541, "distillation_loss": 0.47391343116760254, "epoch": 2.09, "learning_rate": 4.780525986224171e-05, "loss": 10.1311, "step": 2476, "task_loss": 0.4595968425273895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09005796747458517, "compression/movement_sparsity/importance_threshold": -0.005905655622241235, "compression/movement_sparsity/linear_layer_sparsity": 0.010407890479286672, "compression/movement_sparsity/model_sparsity": 0.010050347422906988, "compression_loss": 9.722159385681152, "distillation_loss": 0.7042484283447266, "epoch": 2.09, "learning_rate": 4.780212899185974e-05, "loss": 10.3074, "step": 2477, "task_loss": 1.041536808013916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.090850830078125, "compression/movement_sparsity/importance_threshold": -0.005900509829075418, "compression/movement_sparsity/linear_layer_sparsity": 0.01061622953622047, "compression/movement_sparsity/model_sparsity": 0.010251529392309205, "compression_loss": 9.807655334472656, "distillation_loss": 0.4340529441833496, "epoch": 2.09, "learning_rate": 4.7798998121477774e-05, "loss": 10.4186, "step": 2478, "task_loss": 1.127314805984497 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09164323198289603, "compression/movement_sparsity/importance_threshold": -0.005895367025911355, "compression/movement_sparsity/linear_layer_sparsity": 0.010908121235778761, "compression/movement_sparsity/model_sparsity": 0.01053339371402377, "compression_loss": 9.893097877502441, "distillation_loss": 0.5161107182502747, "epoch": 2.1, "learning_rate": 4.7795867251095806e-05, "loss": 10.4851, "step": 2479, "task_loss": 0.36080145835876465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09243517332278328, "compression/movement_sparsity/importance_threshold": -0.005890227211880113, "compression/movement_sparsity/linear_layer_sparsity": 0.011291244741919039, "compression/movement_sparsity/model_sparsity": 0.010903355749102363, "compression_loss": 9.978495597839355, "distillation_loss": 0.6847619414329529, "epoch": 2.1, "learning_rate": 4.7792736380713845e-05, "loss": 10.5747, "step": 2480, "task_loss": 1.3506616353988647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09322665423167198, "compression/movement_sparsity/importance_threshold": -0.005885090386112759, "compression/movement_sparsity/linear_layer_sparsity": 0.0115297280946363, "compression/movement_sparsity/model_sparsity": 0.01113364646499318, "compression_loss": 10.063837051391602, "distillation_loss": 0.48525765538215637, "epoch": 2.1, "learning_rate": 4.7789605510331877e-05, "loss": 10.5775, "step": 2481, "task_loss": 0.3752705752849579 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09401767484344714, "compression/movement_sparsity/importance_threshold": -0.005879956547740359, "compression/movement_sparsity/linear_layer_sparsity": 0.011770906309239266, "compression/movement_sparsity/model_sparsity": 0.011366539465973562, "compression_loss": 10.149129867553711, "distillation_loss": 0.3827081322669983, "epoch": 2.1, "learning_rate": 4.778647463994991e-05, "loss": 10.761, "step": 2482, "task_loss": 0.9920173287391663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09480823529199378, "compression/movement_sparsity/importance_threshold": -0.00587482569589398, "compression/movement_sparsity/linear_layer_sparsity": 0.012050826144491149, "compression/movement_sparsity/model_sparsity": 0.011636843193750407, "compression_loss": 10.234374046325684, "distillation_loss": 0.6438371539115906, "epoch": 2.1, "learning_rate": 4.778334376956794e-05, "loss": 10.7236, "step": 2483, "task_loss": 0.23454447090625763 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09559833571119758, "compression/movement_sparsity/importance_threshold": -0.005869697829704684, "compression/movement_sparsity/linear_layer_sparsity": 0.012334919438415585, "compression/movement_sparsity/model_sparsity": 0.011911177009055342, "compression_loss": 10.319570541381836, "distillation_loss": 0.42084020376205444, "epoch": 2.1, "learning_rate": 4.778021289918598e-05, "loss": 10.796, "step": 2484, "task_loss": 0.6355000138282776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09638797623494333, "compression/movement_sparsity/importance_threshold": -0.00586457294830354, "compression/movement_sparsity/linear_layer_sparsity": 0.012597418064751474, "compression/movement_sparsity/model_sparsity": 0.012164658000036363, "compression_loss": 10.40471076965332, "distillation_loss": 0.40922582149505615, "epoch": 2.1, "learning_rate": 4.777708202880401e-05, "loss": 10.8727, "step": 2485, "task_loss": 0.3642033040523529 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09717715699711615, "compression/movement_sparsity/importance_threshold": -0.005859451050821614, "compression/movement_sparsity/linear_layer_sparsity": 0.012962392987749968, "compression/movement_sparsity/model_sparsity": 0.012517094911635667, "compression_loss": 10.489790916442871, "distillation_loss": 0.5610826015472412, "epoch": 2.1, "learning_rate": 4.777395115842204e-05, "loss": 11.0904, "step": 2486, "task_loss": 0.9736693501472473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09796587813160129, "compression/movement_sparsity/importance_threshold": -0.005854332136389972, "compression/movement_sparsity/linear_layer_sparsity": 0.013272624057132216, "compression/movement_sparsity/model_sparsity": 0.012816668589402235, "compression_loss": 10.574828147888184, "distillation_loss": 0.8054144978523254, "epoch": 2.1, "learning_rate": 4.777082028804008e-05, "loss": 11.1799, "step": 2487, "task_loss": 0.8829770088195801 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09875413977228376, "compression/movement_sparsity/importance_threshold": -0.005849216204139679, "compression/movement_sparsity/linear_layer_sparsity": 0.013503094369198177, "compression/movement_sparsity/model_sparsity": 0.01303922153723912, "compression_loss": 10.659805297851562, "distillation_loss": 1.3927768468856812, "epoch": 2.1, "learning_rate": 4.776768941765811e-05, "loss": 11.5463, "step": 2488, "task_loss": 2.412458658218384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.09954194205304878, "compression/movement_sparsity/importance_threshold": -0.005844103253201802, "compression/movement_sparsity/linear_layer_sparsity": 0.013699008443455406, "compression/movement_sparsity/model_sparsity": 0.013228405360343426, "compression_loss": 10.744742393493652, "distillation_loss": 0.7488387823104858, "epoch": 2.1, "learning_rate": 4.7764558547276144e-05, "loss": 11.4226, "step": 2489, "task_loss": 1.9958994388580322 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1003292851077815, "compression/movement_sparsity/importance_threshold": -0.005838993282707407, "compression/movement_sparsity/linear_layer_sparsity": 0.013874699129402212, "compression/movement_sparsity/model_sparsity": 0.01339806053074019, "compression_loss": 10.82961654663086, "distillation_loss": 0.6004053354263306, "epoch": 2.1, "learning_rate": 4.7761427676894176e-05, "loss": 11.32, "step": 2490, "task_loss": 0.381400465965271 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10111616907036702, "compression/movement_sparsity/importance_threshold": -0.005833886291787559, "compression/movement_sparsity/linear_layer_sparsity": 0.01422176395261164, "compression/movement_sparsity/model_sparsity": 0.013733202609576094, "compression_loss": 10.914445877075195, "distillation_loss": 0.4100477397441864, "epoch": 2.11, "learning_rate": 4.7758296806512215e-05, "loss": 11.4417, "step": 2491, "task_loss": 1.1919161081314087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10190259407469049, "compression/movement_sparsity/importance_threshold": -0.0058287822795733265, "compression/movement_sparsity/linear_layer_sparsity": 0.014571297078521693, "compression/movement_sparsity/model_sparsity": 0.014070728197321468, "compression_loss": 10.999223709106445, "distillation_loss": 0.6430803537368774, "epoch": 2.11, "learning_rate": 4.7755165936130246e-05, "loss": 11.7333, "step": 2492, "task_loss": 1.7541719675064087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10268856025463713, "compression/movement_sparsity/importance_threshold": -0.005823681245195772, "compression/movement_sparsity/linear_layer_sparsity": 0.014901203024503115, "compression/movement_sparsity/model_sparsity": 0.014389300859149029, "compression_loss": 11.083954811096191, "distillation_loss": 0.46874165534973145, "epoch": 2.11, "learning_rate": 4.775203506574828e-05, "loss": 11.6788, "step": 2493, "task_loss": 0.3522428274154663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10347406774409162, "compression/movement_sparsity/importance_threshold": -0.005818583187785967, "compression/movement_sparsity/linear_layer_sparsity": 0.01521734848103275, "compression/movement_sparsity/model_sparsity": 0.014694585746669687, "compression_loss": 11.168622970581055, "distillation_loss": 0.5918667912483215, "epoch": 2.11, "learning_rate": 4.774890419536631e-05, "loss": 11.7992, "step": 2494, "task_loss": 1.0757381916046143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10425911667693977, "compression/movement_sparsity/importance_threshold": -0.005813488106474972, "compression/movement_sparsity/linear_layer_sparsity": 0.015600853560537378, "compression/movement_sparsity/model_sparsity": 0.01506491624689371, "compression_loss": 11.253251075744629, "distillation_loss": 0.41713857650756836, "epoch": 2.11, "learning_rate": 4.774577332498435e-05, "loss": 11.895, "step": 2495, "task_loss": 0.6001870632171631 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10504370718706657, "compression/movement_sparsity/importance_threshold": -0.005808396000393854, "compression/movement_sparsity/linear_layer_sparsity": 0.015897169126288573, "compression/movement_sparsity/model_sparsity": 0.015351052461388048, "compression_loss": 11.33781909942627, "distillation_loss": 0.4317503869533539, "epoch": 2.11, "learning_rate": 4.774264245460238e-05, "loss": 11.9093, "step": 2496, "task_loss": 0.22093573212623596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10582783940835694, "compression/movement_sparsity/importance_threshold": -0.005803306868673681, "compression/movement_sparsity/linear_layer_sparsity": 0.016309268359784, "compression/movement_sparsity/model_sparsity": 0.015748994818447378, "compression_loss": 11.422333717346191, "distillation_loss": 0.6259301900863647, "epoch": 2.11, "learning_rate": 4.773951158422041e-05, "loss": 11.9216, "step": 2497, "task_loss": 1.1824445724487305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1066115134746961, "compression/movement_sparsity/importance_threshold": -0.005798220710445518, "compression/movement_sparsity/linear_layer_sparsity": 0.016638935822412704, "compression/movement_sparsity/model_sparsity": 0.016067337189559047, "compression_loss": 11.506796836853027, "distillation_loss": 0.48445576429367065, "epoch": 2.11, "learning_rate": 4.7736380713838444e-05, "loss": 12.0243, "step": 2498, "task_loss": 0.7680969834327698 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1073947295199692, "compression/movement_sparsity/importance_threshold": -0.005793137524840431, "compression/movement_sparsity/linear_layer_sparsity": 0.01695046662606726, "compression/movement_sparsity/model_sparsity": 0.01636816595172722, "compression_loss": 11.591214179992676, "distillation_loss": 0.5935084819793701, "epoch": 2.11, "learning_rate": 4.773324984345648e-05, "loss": 12.0489, "step": 2499, "task_loss": 1.2045649290084839 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10817748767806146, "compression/movement_sparsity/importance_threshold": -0.005788057310989486, "compression/movement_sparsity/linear_layer_sparsity": 0.01743009242088458, "compression/movement_sparsity/model_sparsity": 0.016831315124991034, "compression_loss": 11.675570487976074, "distillation_loss": 0.8085423707962036, "epoch": 2.11, "learning_rate": 4.7730118973074514e-05, "loss": 12.3601, "step": 2500, "task_loss": 0.6951969265937805 }, { "epoch": 2.11, "eval_accuracy": 0.9023366336633664, "eval_loss": 12.032795906066895, "eval_runtime": 209.7114, "eval_samples_per_second": 120.404, "eval_steps_per_second": 0.944, "step": 2500 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10895978808285789, "compression/movement_sparsity/importance_threshold": -0.00578298006802375, "compression/movement_sparsity/linear_layer_sparsity": 0.017652955113998857, "compression/movement_sparsity/model_sparsity": 0.017046521798991, "compression_loss": 11.759878158569336, "distillation_loss": 0.278495728969574, "epoch": 2.11, "learning_rate": 4.7726988102692546e-05, "loss": 12.293, "step": 2501, "task_loss": 0.2632511854171753 }, { "compression/movement_sparsity/importance_regularization_factor": 0.10974163086824362, "compression/movement_sparsity/importance_threshold": -0.0057779057950742875, "compression/movement_sparsity/linear_layer_sparsity": 0.018011812939000154, "compression/movement_sparsity/model_sparsity": 0.01739305175372771, "compression_loss": 11.844137191772461, "distillation_loss": 0.5362043380737305, "epoch": 2.11, "learning_rate": 4.7723857232310585e-05, "loss": 12.3839, "step": 2502, "task_loss": 0.4176577627658844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.110523016168104, "compression/movement_sparsity/importance_threshold": -0.0057728344912721654, "compression/movement_sparsity/linear_layer_sparsity": 0.01838595754691063, "compression/movement_sparsity/model_sparsity": 0.017754343343353015, "compression_loss": 11.92834758758545, "distillation_loss": 0.43128153681755066, "epoch": 2.12, "learning_rate": 4.7720726361928616e-05, "loss": 12.4279, "step": 2503, "task_loss": 0.7632910013198853 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1113039441163236, "compression/movement_sparsity/importance_threshold": -0.005767766155748452, "compression/movement_sparsity/linear_layer_sparsity": 0.018758575861363713, "compression/movement_sparsity/model_sparsity": 0.01811416107239662, "compression_loss": 12.012508392333984, "distillation_loss": 0.613702118396759, "epoch": 2.12, "learning_rate": 4.7717595491546655e-05, "loss": 12.6604, "step": 2504, "task_loss": 1.2232531309127808 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11208441484678833, "compression/movement_sparsity/importance_threshold": -0.005762700787634209, "compression/movement_sparsity/linear_layer_sparsity": 0.019009078775057923, "compression/movement_sparsity/model_sparsity": 0.018356058440368334, "compression_loss": 12.096613883972168, "distillation_loss": 0.7630182504653931, "epoch": 2.12, "learning_rate": 4.771446462116469e-05, "loss": 12.7989, "step": 2505, "task_loss": 0.7728609442710876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11286442849338296, "compression/movement_sparsity/importance_threshold": -0.0057576383860605046, "compression/movement_sparsity/linear_layer_sparsity": 0.019288593188610186, "compression/movement_sparsity/model_sparsity": 0.018625970673928162, "compression_loss": 12.180685043334961, "distillation_loss": 0.6107282638549805, "epoch": 2.12, "learning_rate": 4.771133375078272e-05, "loss": 12.7492, "step": 2506, "task_loss": 0.9047357439994812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11364398518999275, "compression/movement_sparsity/importance_threshold": -0.005752578950158405, "compression/movement_sparsity/linear_layer_sparsity": 0.019633762069165513, "compression/movement_sparsity/model_sparsity": 0.018959281941572734, "compression_loss": 12.26469898223877, "distillation_loss": 0.4964275360107422, "epoch": 2.12, "learning_rate": 4.770820288040076e-05, "loss": 12.8756, "step": 2507, "task_loss": 0.4108199179172516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1144230850705027, "compression/movement_sparsity/importance_threshold": -0.0057475224790589745, "compression/movement_sparsity/linear_layer_sparsity": 0.019872066559368236, "compression/movement_sparsity/model_sparsity": 0.019189399939426633, "compression_loss": 12.348668098449707, "distillation_loss": 0.7940950989723206, "epoch": 2.12, "learning_rate": 4.770507201001879e-05, "loss": 13.042, "step": 2508, "task_loss": 0.9802160859107971 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11520172826879793, "compression/movement_sparsity/importance_threshold": -0.005742468971893282, "compression/movement_sparsity/linear_layer_sparsity": 0.020199098780949415, "compression/movement_sparsity/model_sparsity": 0.01950519759812771, "compression_loss": 12.43259048461914, "distillation_loss": 0.7256219387054443, "epoch": 2.12, "learning_rate": 4.770194113963682e-05, "loss": 13.004, "step": 2509, "task_loss": 0.7595838904380798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1159799149187637, "compression/movement_sparsity/importance_threshold": -0.005737418427792391, "compression/movement_sparsity/linear_layer_sparsity": 0.020454013636668895, "compression/movement_sparsity/model_sparsity": 0.019751355344343403, "compression_loss": 12.516448974609375, "distillation_loss": 0.40436020493507385, "epoch": 2.12, "learning_rate": 4.769881026925486e-05, "loss": 12.9443, "step": 2510, "task_loss": 0.2580111026763916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1167576451542851, "compression/movement_sparsity/importance_threshold": -0.005732370845887369, "compression/movement_sparsity/linear_layer_sparsity": 0.0208815665914204, "compression/movement_sparsity/model_sparsity": 0.020164220539792457, "compression_loss": 12.600245475769043, "distillation_loss": 0.5622445940971375, "epoch": 2.12, "learning_rate": 4.769567939887289e-05, "loss": 13.2169, "step": 2511, "task_loss": 0.8614364266395569 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1175349191092473, "compression/movement_sparsity/importance_threshold": -0.005727326225309283, "compression/movement_sparsity/linear_layer_sparsity": 0.02130433795494992, "compression/movement_sparsity/model_sparsity": 0.0205724684063879, "compression_loss": 12.683991432189941, "distillation_loss": 0.6842716932296753, "epoch": 2.12, "learning_rate": 4.769254852849092e-05, "loss": 13.3158, "step": 2512, "task_loss": 0.4576231837272644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11831173691753538, "compression/movement_sparsity/importance_threshold": -0.005722284565189196, "compression/movement_sparsity/linear_layer_sparsity": 0.02169049019966971, "compression/movement_sparsity/model_sparsity": 0.02094535513355831, "compression_loss": 12.767690658569336, "distillation_loss": 0.24844618141651154, "epoch": 2.12, "learning_rate": 4.768941765810896e-05, "loss": 13.1715, "step": 2513, "task_loss": 0.09964857250452042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11908809871303416, "compression/movement_sparsity/importance_threshold": -0.005717245864658179, "compression/movement_sparsity/linear_layer_sparsity": 0.02199593967782998, "compression/movement_sparsity/model_sparsity": 0.021240311482471266, "compression_loss": 12.851343154907227, "distillation_loss": 0.7388131618499756, "epoch": 2.13, "learning_rate": 4.768628678772699e-05, "loss": 13.5034, "step": 2514, "task_loss": 1.1730766296386719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.11986400462962954, "compression/movement_sparsity/importance_threshold": -0.005712210122847292, "compression/movement_sparsity/linear_layer_sparsity": 0.022307291618969997, "compression/movement_sparsity/model_sparsity": 0.021540967526602523, "compression_loss": 12.934932708740234, "distillation_loss": 0.8580704927444458, "epoch": 2.13, "learning_rate": 4.7683155917345025e-05, "loss": 13.5621, "step": 2515, "task_loss": 0.5532090067863464 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12063945480120619, "compression/movement_sparsity/importance_threshold": -0.0057071773388876045, "compression/movement_sparsity/linear_layer_sparsity": 0.022624963368957023, "compression/movement_sparsity/model_sparsity": 0.02184772627470488, "compression_loss": 13.018482208251953, "distillation_loss": 1.1399739980697632, "epoch": 2.13, "learning_rate": 4.7680025046963057e-05, "loss": 13.624, "step": 2516, "task_loss": 1.0813512802124023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12141444936164936, "compression/movement_sparsity/importance_threshold": -0.005702147511910181, "compression/movement_sparsity/linear_layer_sparsity": 0.02303420080221984, "compression/movement_sparsity/model_sparsity": 0.022242905143173522, "compression_loss": 13.101982116699219, "distillation_loss": 0.8479353785514832, "epoch": 2.13, "learning_rate": 4.7676894176581095e-05, "loss": 13.6719, "step": 2517, "task_loss": 1.492380976676941 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12218898844484405, "compression/movement_sparsity/importance_threshold": -0.0056971206410460894, "compression/movement_sparsity/linear_layer_sparsity": 0.023389815253624185, "compression/movement_sparsity/model_sparsity": 0.022586303144174112, "compression_loss": 13.185418128967285, "distillation_loss": 0.5313619375228882, "epoch": 2.13, "learning_rate": 4.767376330619913e-05, "loss": 13.6441, "step": 2518, "task_loss": 0.5065880417823792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12296307218467561, "compression/movement_sparsity/importance_threshold": -0.005692096725426394, "compression/movement_sparsity/linear_layer_sparsity": 0.023754802100790316, "compression/movement_sparsity/model_sparsity": 0.022938751570309213, "compression_loss": 13.268796920776367, "distillation_loss": 0.5685032606124878, "epoch": 2.13, "learning_rate": 4.767063243581716e-05, "loss": 13.8229, "step": 2519, "task_loss": 0.504263162612915 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12373670071502896, "compression/movement_sparsity/importance_threshold": -0.005687075764182163, "compression/movement_sparsity/linear_layer_sparsity": 0.02417338808147965, "compression/movement_sparsity/model_sparsity": 0.023342957834840774, "compression_loss": 13.352120399475098, "distillation_loss": 0.5057448148727417, "epoch": 2.13, "learning_rate": 4.766750156543519e-05, "loss": 14.0977, "step": 2520, "task_loss": 0.3477921485900879 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12450987416978943, "compression/movement_sparsity/importance_threshold": -0.00568205775644446, "compression/movement_sparsity/linear_layer_sparsity": 0.02459772151096947, "compression/movement_sparsity/model_sparsity": 0.0237527141056253, "compression_loss": 13.435402870178223, "distillation_loss": 0.4782823920249939, "epoch": 2.13, "learning_rate": 4.766437069505323e-05, "loss": 14.0854, "step": 2521, "task_loss": 0.7928575873374939 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12528259268284203, "compression/movement_sparsity/importance_threshold": -0.005677042701344352, "compression/movement_sparsity/linear_layer_sparsity": 0.024935032365052765, "compression/movement_sparsity/model_sparsity": 0.02407843729418127, "compression_loss": 13.518645286560059, "distillation_loss": 0.5356513261795044, "epoch": 2.13, "learning_rate": 4.766123982467126e-05, "loss": 14.0412, "step": 2522, "task_loss": 1.0778230428695679 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12605485638807157, "compression/movement_sparsity/importance_threshold": -0.005672030598012907, "compression/movement_sparsity/linear_layer_sparsity": 0.025243379415948546, "compression/movement_sparsity/model_sparsity": 0.0243761916752923, "compression_loss": 13.601839065551758, "distillation_loss": 0.528675377368927, "epoch": 2.13, "learning_rate": 4.765810895428929e-05, "loss": 14.1369, "step": 2523, "task_loss": 0.48589956760406494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12682666541936405, "compression/movement_sparsity/importance_threshold": -0.005667021445581188, "compression/movement_sparsity/linear_layer_sparsity": 0.02558088105671401, "compression/movement_sparsity/model_sparsity": 0.024702099096420984, "compression_loss": 13.684976577758789, "distillation_loss": 0.5117270350456238, "epoch": 2.13, "learning_rate": 4.765497808390733e-05, "loss": 14.2493, "step": 2524, "task_loss": 0.773658037185669 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12759801991060393, "compression/movement_sparsity/importance_threshold": -0.005662015243180262, "compression/movement_sparsity/linear_layer_sparsity": 0.02600978144240837, "compression/movement_sparsity/model_sparsity": 0.025116265434414822, "compression_loss": 13.768060684204102, "distillation_loss": 0.948596715927124, "epoch": 2.13, "learning_rate": 4.765184721352536e-05, "loss": 14.3856, "step": 2525, "task_loss": 1.4683077335357666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12836891999567668, "compression/movement_sparsity/importance_threshold": -0.005657011989941195, "compression/movement_sparsity/linear_layer_sparsity": 0.026354723763778613, "compression/movement_sparsity/model_sparsity": 0.0254493579258793, "compression_loss": 13.851090431213379, "distillation_loss": 0.5374099016189575, "epoch": 2.14, "learning_rate": 4.7648716343143395e-05, "loss": 14.3746, "step": 2526, "task_loss": 0.8020678758621216 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1291393658084673, "compression/movement_sparsity/importance_threshold": -0.005652011684995053, "compression/movement_sparsity/linear_layer_sparsity": 0.026741066795180575, "compression/movement_sparsity/model_sparsity": 0.02582242888562242, "compression_loss": 13.934062957763672, "distillation_loss": 0.5234959125518799, "epoch": 2.14, "learning_rate": 4.7645585472761427e-05, "loss": 14.4906, "step": 2527, "task_loss": 0.9672248363494873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.12990935748286092, "compression/movement_sparsity/importance_threshold": -0.005647014327472903, "compression/movement_sparsity/linear_layer_sparsity": 0.0270951788014628, "compression/movement_sparsity/model_sparsity": 0.026164376055112896, "compression_loss": 14.016979217529297, "distillation_loss": 0.7152500152587891, "epoch": 2.14, "learning_rate": 4.7642454602379465e-05, "loss": 14.5275, "step": 2528, "task_loss": 0.5796002149581909 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13067889515274267, "compression/movement_sparsity/importance_threshold": -0.00564201991650581, "compression/movement_sparsity/linear_layer_sparsity": 0.027475619369885008, "compression/movement_sparsity/model_sparsity": 0.02653174731963772, "compression_loss": 14.099839210510254, "distillation_loss": 0.318798303604126, "epoch": 2.14, "learning_rate": 4.76393237319975e-05, "loss": 14.733, "step": 2529, "task_loss": 0.46080684661865234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13144797895199778, "compression/movement_sparsity/importance_threshold": -0.00563702845122484, "compression/movement_sparsity/linear_layer_sparsity": 0.027977984552383917, "compression/movement_sparsity/model_sparsity": 0.027016854712661725, "compression_loss": 14.182652473449707, "distillation_loss": 0.6657871007919312, "epoch": 2.14, "learning_rate": 4.763619286161553e-05, "loss": 14.7417, "step": 2530, "task_loss": 0.9634329080581665 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13221660901451127, "compression/movement_sparsity/importance_threshold": -0.00563203993076106, "compression/movement_sparsity/linear_layer_sparsity": 0.028431675282593234, "compression/movement_sparsity/model_sparsity": 0.027454959770572412, "compression_loss": 14.26540470123291, "distillation_loss": 0.5511736869812012, "epoch": 2.14, "learning_rate": 4.763306199123356e-05, "loss": 14.8156, "step": 2531, "task_loss": 1.2524274587631226 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13298478547416837, "compression/movement_sparsity/importance_threshold": -0.005627054354245536, "compression/movement_sparsity/linear_layer_sparsity": 0.028824147336160028, "compression/movement_sparsity/model_sparsity": 0.02783394920171393, "compression_loss": 14.348104476928711, "distillation_loss": 0.7599971294403076, "epoch": 2.14, "learning_rate": 4.76299311208516e-05, "loss": 14.8604, "step": 2532, "task_loss": 0.833115816116333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1337525084648542, "compression/movement_sparsity/importance_threshold": -0.005622071720809333, "compression/movement_sparsity/linear_layer_sparsity": 0.029330876764013664, "compression/movement_sparsity/model_sparsity": 0.028323270914838734, "compression_loss": 14.430750846862793, "distillation_loss": 0.9440480470657349, "epoch": 2.14, "learning_rate": 4.762680025046963e-05, "loss": 15.0394, "step": 2533, "task_loss": 2.4352779388427734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13451977812045357, "compression/movement_sparsity/importance_threshold": -0.005617092029583519, "compression/movement_sparsity/linear_layer_sparsity": 0.029711138469921336, "compression/movement_sparsity/model_sparsity": 0.02869046946132664, "compression_loss": 14.513333320617676, "distillation_loss": 0.4248049855232239, "epoch": 2.14, "learning_rate": 4.762366938008766e-05, "loss": 15.023, "step": 2534, "task_loss": 1.3365730047225952 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13528659457485237, "compression/movement_sparsity/importance_threshold": -0.005612115279699157, "compression/movement_sparsity/linear_layer_sparsity": 0.030066752921325678, "compression/movement_sparsity/model_sparsity": 0.02903386746232723, "compression_loss": 14.595866203308105, "distillation_loss": 0.5043361783027649, "epoch": 2.14, "learning_rate": 4.76205385097057e-05, "loss": 15.2825, "step": 2535, "task_loss": 0.3752364218235016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13605295796193517, "compression/movement_sparsity/importance_threshold": -0.0056071414702873155, "compression/movement_sparsity/linear_layer_sparsity": 0.0305597457080628, "compression/movement_sparsity/model_sparsity": 0.029509924430216727, "compression_loss": 14.67834758758545, "distillation_loss": 0.8242826461791992, "epoch": 2.14, "learning_rate": 4.761740763932373e-05, "loss": 15.1913, "step": 2536, "task_loss": 1.1448078155517578 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13681886841558732, "compression/movement_sparsity/importance_threshold": -0.005602170600479059, "compression/movement_sparsity/linear_layer_sparsity": 0.030948950539697366, "compression/movement_sparsity/model_sparsity": 0.029885758878550536, "compression_loss": 14.760785102844238, "distillation_loss": 0.6137906312942505, "epoch": 2.14, "learning_rate": 4.7614276768941765e-05, "loss": 15.4015, "step": 2537, "task_loss": 0.4415515661239624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13758432606969384, "compression/movement_sparsity/importance_threshold": -0.005597202669405455, "compression/movement_sparsity/linear_layer_sparsity": 0.031384134961735824, "compression/movement_sparsity/model_sparsity": 0.0303059933769081, "compression_loss": 14.843162536621094, "distillation_loss": 0.37066197395324707, "epoch": 2.15, "learning_rate": 4.76111458985598e-05, "loss": 15.465, "step": 2538, "task_loss": 0.08855317533016205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.13834933105813996, "compression/movement_sparsity/importance_threshold": -0.0055922376761975675, "compression/movement_sparsity/linear_layer_sparsity": 0.032034526761266335, "compression/movement_sparsity/model_sparsity": 0.03093404221728553, "compression_loss": 14.925480842590332, "distillation_loss": 0.4489689767360687, "epoch": 2.15, "learning_rate": 4.7608015028177835e-05, "loss": 15.6577, "step": 2539, "task_loss": 0.41667982935905457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1391138835148108, "compression/movement_sparsity/importance_threshold": -0.005587275619986465, "compression/movement_sparsity/linear_layer_sparsity": 0.03251990000488414, "compression/movement_sparsity/model_sparsity": 0.03140274139680231, "compression_loss": 15.007755279541016, "distillation_loss": 0.41979971528053284, "epoch": 2.15, "learning_rate": 4.7604884157795874e-05, "loss": 15.5601, "step": 2540, "task_loss": 0.13476291298866272 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1398779835735915, "compression/movement_sparsity/importance_threshold": -0.005582316499903211, "compression/movement_sparsity/linear_layer_sparsity": 0.032996103563589965, "compression/movement_sparsity/model_sparsity": 0.0318625858982931, "compression_loss": 15.089981079101562, "distillation_loss": 0.9956264495849609, "epoch": 2.15, "learning_rate": 4.7601753287413905e-05, "loss": 15.6591, "step": 2541, "task_loss": 1.5913867950439453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14064163136836716, "compression/movement_sparsity/importance_threshold": -0.005577360315078874, "compression/movement_sparsity/linear_layer_sparsity": 0.033525953952489536, "compression/movement_sparsity/model_sparsity": 0.032374234296323516, "compression_loss": 15.172152519226074, "distillation_loss": 0.6104323267936707, "epoch": 2.15, "learning_rate": 4.759862241703194e-05, "loss": 15.745, "step": 2542, "task_loss": 1.356231689453125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14140482703302293, "compression/movement_sparsity/importance_threshold": -0.005572407064644519, "compression/movement_sparsity/linear_layer_sparsity": 0.03404185306525515, "compression/movement_sparsity/model_sparsity": 0.032872410687474325, "compression_loss": 15.25428295135498, "distillation_loss": 0.606002151966095, "epoch": 2.15, "learning_rate": 4.7595491546649976e-05, "loss": 15.7865, "step": 2543, "task_loss": 0.6194770336151123 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14216757070144403, "compression/movement_sparsity/importance_threshold": -0.005567456747731211, "compression/movement_sparsity/linear_layer_sparsity": 0.03456385735185032, "compression/movement_sparsity/model_sparsity": 0.03337648252095193, "compression_loss": 15.3363618850708, "distillation_loss": 0.48396044969558716, "epoch": 2.15, "learning_rate": 4.759236067626801e-05, "loss": 15.8203, "step": 2544, "task_loss": 0.4347563087940216 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14292986250751516, "compression/movement_sparsity/importance_threshold": -0.005562509363470021, "compression/movement_sparsity/linear_layer_sparsity": 0.03509368389241463, "compression/movement_sparsity/model_sparsity": 0.033888107889910764, "compression_loss": 15.41838550567627, "distillation_loss": 0.6619105339050293, "epoch": 2.15, "learning_rate": 4.758922980588604e-05, "loss": 16.0124, "step": 2545, "task_loss": 0.7195212244987488 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1436917025851222, "compression/movement_sparsity/importance_threshold": -0.005557564910992007, "compression/movement_sparsity/linear_layer_sparsity": 0.03571050916005018, "compression/movement_sparsity/model_sparsity": 0.03448374331202657, "compression_loss": 15.500356674194336, "distillation_loss": 0.43613138794898987, "epoch": 2.15, "learning_rate": 4.758609893550407e-05, "loss": 15.9575, "step": 2546, "task_loss": 0.6238259673118591 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14445309106814996, "compression/movement_sparsity/importance_threshold": -0.00555262338942824, "compression/movement_sparsity/linear_layer_sparsity": 0.03628018626885354, "compression/movement_sparsity/model_sparsity": 0.035033850259610755, "compression_loss": 15.582277297973633, "distillation_loss": 0.3563975691795349, "epoch": 2.15, "learning_rate": 4.758296806512211e-05, "loss": 16.0224, "step": 2547, "task_loss": 0.333992600440979 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14521402809048345, "compression/movement_sparsity/importance_threshold": -0.005547684797909786, "compression/movement_sparsity/linear_layer_sparsity": 0.036762673863903465, "compression/movement_sparsity/model_sparsity": 0.035499762921465255, "compression_loss": 15.664133071899414, "distillation_loss": 0.5010138750076294, "epoch": 2.15, "learning_rate": 4.757983719474014e-05, "loss": 16.1331, "step": 2548, "task_loss": 0.5144262909889221 }, { "compression/movement_sparsity/importance_regularization_factor": 0.145974513786008, "compression/movement_sparsity/importance_threshold": -0.0055427491355677086, "compression/movement_sparsity/linear_layer_sparsity": 0.03740068837742795, "compression/movement_sparsity/model_sparsity": 0.03611585967368796, "compression_loss": 15.745939254760742, "distillation_loss": 0.3559642434120178, "epoch": 2.15, "learning_rate": 4.757670632435817e-05, "loss": 16.2427, "step": 2549, "task_loss": 0.7619218230247498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14673454828860866, "compression/movement_sparsity/importance_threshold": -0.005537816401533076, "compression/movement_sparsity/linear_layer_sparsity": 0.0379184953570153, "compression/movement_sparsity/model_sparsity": 0.03661587839056589, "compression_loss": 15.827678680419922, "distillation_loss": 0.6284353733062744, "epoch": 2.16, "learning_rate": 4.757357545397621e-05, "loss": 16.3914, "step": 2550, "task_loss": 1.9819852113723755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14749413173217052, "compression/movement_sparsity/importance_threshold": -0.005532886594936954, "compression/movement_sparsity/linear_layer_sparsity": 0.03861620225372492, "compression/movement_sparsity/model_sparsity": 0.037289616908976066, "compression_loss": 15.909363746643066, "distillation_loss": 0.43171942234039307, "epoch": 2.16, "learning_rate": 4.7570444583594243e-05, "loss": 16.3498, "step": 2551, "task_loss": 0.42524290084838867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14825326425057883, "compression/movement_sparsity/importance_threshold": -0.005527959714910408, "compression/movement_sparsity/linear_layer_sparsity": 0.039180954681294664, "compression/movement_sparsity/model_sparsity": 0.037834968353277106, "compression_loss": 15.991000175476074, "distillation_loss": 0.4507143497467041, "epoch": 2.16, "learning_rate": 4.7567313713212275e-05, "loss": 16.567, "step": 2552, "task_loss": 0.22841984033584595 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1490119459777186, "compression/movement_sparsity/importance_threshold": -0.005523035760584504, "compression/movement_sparsity/linear_layer_sparsity": 0.039788419477336065, "compression/movement_sparsity/model_sparsity": 0.03842156486479419, "compression_loss": 16.07259750366211, "distillation_loss": 0.6524878144264221, "epoch": 2.16, "learning_rate": 4.756418284283031e-05, "loss": 16.6658, "step": 2553, "task_loss": 1.2741645574569702 }, { "compression/movement_sparsity/importance_regularization_factor": 0.14977017704747486, "compression/movement_sparsity/importance_threshold": -0.005518114731090311, "compression/movement_sparsity/linear_layer_sparsity": 0.04041574993665882, "compression/movement_sparsity/model_sparsity": 0.03902734459294498, "compression_loss": 16.154144287109375, "distillation_loss": 0.4062063694000244, "epoch": 2.16, "learning_rate": 4.7561051972448346e-05, "loss": 16.641, "step": 2554, "task_loss": 0.720470130443573 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15052795759373316, "compression/movement_sparsity/importance_threshold": -0.005513196625558891, "compression/movement_sparsity/linear_layer_sparsity": 0.0410445828411037, "compression/movement_sparsity/model_sparsity": 0.039634575152605885, "compression_loss": 16.235628128051758, "distillation_loss": 0.30214062333106995, "epoch": 2.16, "learning_rate": 4.755792110206638e-05, "loss": 16.7901, "step": 2555, "task_loss": 0.4137303829193115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15128528775037842, "compression/movement_sparsity/importance_threshold": -0.005508281443121311, "compression/movement_sparsity/linear_layer_sparsity": 0.041646145174165344, "compression/movement_sparsity/model_sparsity": 0.040215471968904674, "compression_loss": 16.317092895507812, "distillation_loss": 0.7011353373527527, "epoch": 2.16, "learning_rate": 4.755479023168441e-05, "loss": 16.9378, "step": 2556, "task_loss": 0.660033643245697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15204216765129575, "compression/movement_sparsity/importance_threshold": -0.005503369182908637, "compression/movement_sparsity/linear_layer_sparsity": 0.042146781352357054, "compression/movement_sparsity/model_sparsity": 0.04069890975423847, "compression_loss": 16.39850616455078, "distillation_loss": 0.6069098711013794, "epoch": 2.16, "learning_rate": 4.755165936130244e-05, "loss": 17.2013, "step": 2557, "task_loss": 2.1290268898010254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1527985974303704, "compression/movement_sparsity/importance_threshold": -0.005498459844051935, "compression/movement_sparsity/linear_layer_sparsity": 0.042672029012549184, "compression/movement_sparsity/model_sparsity": 0.0412061135414522, "compression_loss": 16.47986602783203, "distillation_loss": 0.712462306022644, "epoch": 2.16, "learning_rate": 4.754852849092048e-05, "loss": 17.0428, "step": 2558, "task_loss": 0.583937406539917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15355457722148735, "compression/movement_sparsity/importance_threshold": -0.005493553425682273, "compression/movement_sparsity/linear_layer_sparsity": 0.043306979014991254, "compression/movement_sparsity/model_sparsity": 0.0418192510579757, "compression_loss": 16.56117057800293, "distillation_loss": 0.6467421650886536, "epoch": 2.16, "learning_rate": 4.754539762053851e-05, "loss": 17.1761, "step": 2559, "task_loss": 1.1820670366287231 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15431010715853188, "compression/movement_sparsity/importance_threshold": -0.0054886499269307144, "compression/movement_sparsity/linear_layer_sparsity": 0.04376030009600386, "compression/movement_sparsity/model_sparsity": 0.042256999165276754, "compression_loss": 16.64242172241211, "distillation_loss": 0.6140848398208618, "epoch": 2.16, "learning_rate": 4.754226675015654e-05, "loss": 17.2849, "step": 2560, "task_loss": 0.6969770193099976 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1550651873753891, "compression/movement_sparsity/importance_threshold": -0.005483749346928328, "compression/movement_sparsity/linear_layer_sparsity": 0.044407853943637034, "compression/movement_sparsity/model_sparsity": 0.04288230754613509, "compression_loss": 16.723617553710938, "distillation_loss": 0.6086236834526062, "epoch": 2.16, "learning_rate": 4.753913587977458e-05, "loss": 17.3085, "step": 2561, "task_loss": 0.6385365128517151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15581981800594402, "compression/movement_sparsity/importance_threshold": -0.005478851684806177, "compression/movement_sparsity/linear_layer_sparsity": 0.04496648927320958, "compression/movement_sparsity/model_sparsity": 0.043421752033573534, "compression_loss": 16.804758071899414, "distillation_loss": 0.809333860874176, "epoch": 2.17, "learning_rate": 4.7536005009392613e-05, "loss": 17.4913, "step": 2562, "task_loss": 1.3450331687927246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.156573999184082, "compression/movement_sparsity/importance_threshold": -0.005473956939695329, "compression/movement_sparsity/linear_layer_sparsity": 0.04546004249582559, "compression/movement_sparsity/model_sparsity": 0.04389835018464537, "compression_loss": 16.885852813720703, "distillation_loss": 0.428792804479599, "epoch": 2.17, "learning_rate": 4.7532874139010645e-05, "loss": 17.3086, "step": 2563, "task_loss": 0.36649569869041443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1573277310436879, "compression/movement_sparsity/importance_threshold": -0.00546906511072685, "compression/movement_sparsity/linear_layer_sparsity": 0.045979387693037964, "compression/movement_sparsity/model_sparsity": 0.044399854276640796, "compression_loss": 16.96689224243164, "distillation_loss": 0.6158169507980347, "epoch": 2.17, "learning_rate": 4.752974326862868e-05, "loss": 17.4961, "step": 2564, "task_loss": 0.48326969146728516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1580810137186469, "compression/movement_sparsity/importance_threshold": -0.005464176197031808, "compression/movement_sparsity/linear_layer_sparsity": 0.04666793682900324, "compression/movement_sparsity/model_sparsity": 0.04506474963156076, "compression_loss": 17.04787254333496, "distillation_loss": 0.6933388710021973, "epoch": 2.17, "learning_rate": 4.7526612398246716e-05, "loss": 17.7725, "step": 2565, "task_loss": 0.12252967059612274 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15883384734284445, "compression/movement_sparsity/importance_threshold": -0.005459290197741265, "compression/movement_sparsity/linear_layer_sparsity": 0.04737785407337198, "compression/movement_sparsity/model_sparsity": 0.045750279034624544, "compression_loss": 17.12881851196289, "distillation_loss": 0.5452978610992432, "epoch": 2.17, "learning_rate": 4.752348152786475e-05, "loss": 17.7374, "step": 2566, "task_loss": 0.5267574191093445 }, { "compression/movement_sparsity/importance_regularization_factor": 0.15958623205016553, "compression/movement_sparsity/importance_threshold": -0.005454407111986288, "compression/movement_sparsity/linear_layer_sparsity": 0.04807689647685681, "compression/movement_sparsity/model_sparsity": 0.0464253071810437, "compression_loss": 17.2097110748291, "distillation_loss": 0.4271111488342285, "epoch": 2.17, "learning_rate": 4.752035065748278e-05, "loss": 17.7764, "step": 2567, "task_loss": 0.35401666164398193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1603381679744953, "compression/movement_sparsity/importance_threshold": -0.005449526938897945, "compression/movement_sparsity/linear_layer_sparsity": 0.048763752381017794, "compression/movement_sparsity/model_sparsity": 0.047088567471880845, "compression_loss": 17.29054832458496, "distillation_loss": 0.546153724193573, "epoch": 2.17, "learning_rate": 4.751721978710081e-05, "loss": 17.8335, "step": 2568, "task_loss": 0.35751181840896606 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16108965524971874, "compression/movement_sparsity/importance_threshold": -0.005444649677607301, "compression/movement_sparsity/linear_layer_sparsity": 0.04936529086574418, "compression/movement_sparsity/model_sparsity": 0.04766944125910804, "compression_loss": 17.3713321685791, "distillation_loss": 0.46392932534217834, "epoch": 2.17, "learning_rate": 4.751408891671885e-05, "loss": 17.8385, "step": 2569, "task_loss": 1.1553065776824951 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16184069400972123, "compression/movement_sparsity/importance_threshold": -0.005439775327245421, "compression/movement_sparsity/linear_layer_sparsity": 0.049960557238294095, "compression/movement_sparsity/model_sparsity": 0.04824425840050731, "compression_loss": 17.452062606811523, "distillation_loss": 0.21762919425964355, "epoch": 2.17, "learning_rate": 4.751095804633688e-05, "loss": 17.9281, "step": 2570, "task_loss": 1.5561257600784302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16259128438838777, "compression/movement_sparsity/importance_threshold": -0.0054349038869433725, "compression/movement_sparsity/linear_layer_sparsity": 0.05045699610947798, "compression/movement_sparsity/model_sparsity": 0.04872364306924143, "compression_loss": 17.53274154663086, "distillation_loss": 0.4284694492816925, "epoch": 2.17, "learning_rate": 4.750782717595492e-05, "loss": 18.0582, "step": 2571, "task_loss": 0.44271305203437805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16334142651960348, "compression/movement_sparsity/importance_threshold": -0.005430035355832221, "compression/movement_sparsity/linear_layer_sparsity": 0.05112399827452525, "compression/movement_sparsity/model_sparsity": 0.049367731657980654, "compression_loss": 17.613370895385742, "distillation_loss": 0.5635119080543518, "epoch": 2.17, "learning_rate": 4.750469630557295e-05, "loss": 18.1258, "step": 2572, "task_loss": 0.9482659101486206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1640911205372535, "compression/movement_sparsity/importance_threshold": -0.005425169733043034, "compression/movement_sparsity/linear_layer_sparsity": 0.0517978449117955, "compression/movement_sparsity/model_sparsity": 0.05001842959026595, "compression_loss": 17.69392967224121, "distillation_loss": 0.6442070007324219, "epoch": 2.17, "learning_rate": 4.750156543519098e-05, "loss": 18.3589, "step": 2573, "task_loss": 1.0752402544021606 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16484036657522272, "compression/movement_sparsity/importance_threshold": -0.005420307017706877, "compression/movement_sparsity/linear_layer_sparsity": 0.052426737437078555, "compression/movement_sparsity/model_sparsity": 0.05062571772260583, "compression_loss": 17.774463653564453, "distillation_loss": 0.6814181804656982, "epoch": 2.18, "learning_rate": 4.749843456480902e-05, "loss": 18.7106, "step": 2574, "task_loss": 1.0185085535049438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16558916476739693, "compression/movement_sparsity/importance_threshold": -0.005415447208954813, "compression/movement_sparsity/linear_layer_sparsity": 0.052959437702043095, "compression/movement_sparsity/model_sparsity": 0.05114011809469114, "compression_loss": 17.854948043823242, "distillation_loss": 0.5388513803482056, "epoch": 2.18, "learning_rate": 4.7495303694427054e-05, "loss": 18.4279, "step": 2575, "task_loss": 0.21930694580078125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16633751524766094, "compression/movement_sparsity/importance_threshold": -0.005410590305917911, "compression/movement_sparsity/linear_layer_sparsity": 0.05356386183533736, "compression/movement_sparsity/model_sparsity": 0.051723778399580624, "compression_loss": 17.93538475036621, "distillation_loss": 0.7340750098228455, "epoch": 2.18, "learning_rate": 4.749217282404509e-05, "loss": 18.6361, "step": 2576, "task_loss": 0.8354368805885315 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16708541814989974, "compression/movement_sparsity/importance_threshold": -0.005405736307727236, "compression/movement_sparsity/linear_layer_sparsity": 0.05421101026127091, "compression/movement_sparsity/model_sparsity": 0.05234869528622194, "compression_loss": 18.015769958496094, "distillation_loss": 0.5724314451217651, "epoch": 2.18, "learning_rate": 4.7489041953663124e-05, "loss": 18.6146, "step": 2577, "task_loss": 0.593174159526825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1678328736079987, "compression/movement_sparsity/importance_threshold": -0.005400885213513854, "compression/movement_sparsity/linear_layer_sparsity": 0.05470421768302548, "compression/movement_sparsity/model_sparsity": 0.052824959515755734, "compression_loss": 18.096105575561523, "distillation_loss": 0.7497145533561707, "epoch": 2.18, "learning_rate": 4.7485911083281156e-05, "loss": 18.7761, "step": 2578, "task_loss": 2.0371592044830322 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16857988175584282, "compression/movement_sparsity/importance_threshold": -0.005396037022408832, "compression/movement_sparsity/linear_layer_sparsity": 0.055507214979959765, "compression/movement_sparsity/model_sparsity": 0.0536003713852317, "compression_loss": 18.176387786865234, "distillation_loss": 0.506813645362854, "epoch": 2.18, "learning_rate": 4.748278021289919e-05, "loss": 18.9498, "step": 2579, "task_loss": 0.48924702405929565 }, { "compression/movement_sparsity/importance_regularization_factor": 0.16932644272731723, "compression/movement_sparsity/importance_threshold": -0.005391191733543235, "compression/movement_sparsity/linear_layer_sparsity": 0.05611163911325403, "compression/movement_sparsity/model_sparsity": 0.05418403169012118, "compression_loss": 18.25662612915039, "distillation_loss": 0.7935612201690674, "epoch": 2.18, "learning_rate": 4.7479649342517226e-05, "loss": 19.2499, "step": 2580, "task_loss": 0.48973962664604187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17007255665630716, "compression/movement_sparsity/importance_threshold": -0.00538634934604813, "compression/movement_sparsity/linear_layer_sparsity": 0.056707108196653755, "compression/movement_sparsity/model_sparsity": 0.05475904457862896, "compression_loss": 18.336841583251953, "distillation_loss": 0.7533597946166992, "epoch": 2.18, "learning_rate": 4.747651847213526e-05, "loss": 18.9592, "step": 2581, "task_loss": 0.6496530771255493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17081822367669763, "compression/movement_sparsity/importance_threshold": -0.005381509859054582, "compression/movement_sparsity/linear_layer_sparsity": 0.057253497406064266, "compression/movement_sparsity/model_sparsity": 0.055286663637806405, "compression_loss": 18.4169979095459, "distillation_loss": 0.8017264604568481, "epoch": 2.18, "learning_rate": 4.747338760175329e-05, "loss": 19.2127, "step": 2582, "task_loss": 0.39560580253601074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17156344392237388, "compression/movement_sparsity/importance_threshold": -0.005376673271693658, "compression/movement_sparsity/linear_layer_sparsity": 0.05787948043444417, "compression/movement_sparsity/model_sparsity": 0.05589114222341242, "compression_loss": 18.49711036682129, "distillation_loss": 0.6537675261497498, "epoch": 2.18, "learning_rate": 4.747025673137132e-05, "loss": 19.0345, "step": 2583, "task_loss": 0.848233699798584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17230821752722114, "compression/movement_sparsity/importance_threshold": -0.005371839583096423, "compression/movement_sparsity/linear_layer_sparsity": 0.058476249252116204, "compression/movement_sparsity/model_sparsity": 0.0564674101963218, "compression_loss": 18.577184677124023, "distillation_loss": 1.321580410003662, "epoch": 2.18, "learning_rate": 4.746712586098936e-05, "loss": 19.4039, "step": 2584, "task_loss": 0.9027097225189209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17305254462512398, "compression/movement_sparsity/importance_threshold": -0.005367008792393947, "compression/movement_sparsity/linear_layer_sparsity": 0.0590520553830844, "compression/movement_sparsity/model_sparsity": 0.05702343561530438, "compression_loss": 18.65719985961914, "distillation_loss": 0.4809798002243042, "epoch": 2.19, "learning_rate": 4.746399499060739e-05, "loss": 19.2038, "step": 2585, "task_loss": 0.7717567682266235 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17379642534996842, "compression/movement_sparsity/importance_threshold": -0.005362180898717289, "compression/movement_sparsity/linear_layer_sparsity": 0.05973891128724538, "compression/movement_sparsity/model_sparsity": 0.05768669590614152, "compression_loss": 18.737165451049805, "distillation_loss": 0.7324113845825195, "epoch": 2.19, "learning_rate": 4.7460864120225424e-05, "loss": 19.3647, "step": 2586, "task_loss": 1.06319260597229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17453985983563913, "compression/movement_sparsity/importance_threshold": -0.005357355901197519, "compression/movement_sparsity/linear_layer_sparsity": 0.06043342250702858, "compression/movement_sparsity/model_sparsity": 0.058357348528958755, "compression_loss": 18.817060470581055, "distillation_loss": 0.6199138760566711, "epoch": 2.19, "learning_rate": 4.745773324984346e-05, "loss": 19.3144, "step": 2587, "task_loss": 0.28767749667167664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17528284821602125, "compression/movement_sparsity/importance_threshold": -0.005352533798965704, "compression/movement_sparsity/linear_layer_sparsity": 0.06095274385590569, "compression/movement_sparsity/model_sparsity": 0.05885882959188259, "compression_loss": 18.896942138671875, "distillation_loss": 0.5891473889350891, "epoch": 2.19, "learning_rate": 4.7454602379461494e-05, "loss": 19.52, "step": 2588, "task_loss": 1.4223262071609497 }, { "compression/movement_sparsity/importance_regularization_factor": 0.176025390625, "compression/movement_sparsity/importance_threshold": -0.005347714591152908, "compression/movement_sparsity/linear_layer_sparsity": 0.06134366576767982, "compression/movement_sparsity/model_sparsity": 0.059236322133370815, "compression_loss": 18.97675323486328, "distillation_loss": 0.6089892983436584, "epoch": 2.19, "learning_rate": 4.7451471509079526e-05, "loss": 19.5565, "step": 2589, "task_loss": 0.9465068578720093 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17676748719646052, "compression/movement_sparsity/importance_threshold": -0.005342898276890198, "compression/movement_sparsity/linear_layer_sparsity": 0.0618961601509199, "compression/movement_sparsity/model_sparsity": 0.05976983663487507, "compression_loss": 19.056522369384766, "distillation_loss": 0.8925899863243103, "epoch": 2.19, "learning_rate": 4.744834063869756e-05, "loss": 19.7229, "step": 2590, "task_loss": 1.4510037899017334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17750913806428792, "compression/movement_sparsity/importance_threshold": -0.005338084855308639, "compression/movement_sparsity/linear_layer_sparsity": 0.06260128387989902, "compression/movement_sparsity/model_sparsity": 0.06045073719454945, "compression_loss": 19.136220932006836, "distillation_loss": 0.5736783742904663, "epoch": 2.19, "learning_rate": 4.7445209768315596e-05, "loss": 19.6297, "step": 2591, "task_loss": 1.009959101676941 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17825034336236723, "compression/movement_sparsity/importance_threshold": -0.005333274325539299, "compression/movement_sparsity/linear_layer_sparsity": 0.0633094840441282, "compression/movement_sparsity/model_sparsity": 0.061134608504458815, "compression_loss": 19.215866088867188, "distillation_loss": 0.7164064049720764, "epoch": 2.19, "learning_rate": 4.744207889793363e-05, "loss": 19.7297, "step": 2592, "task_loss": 0.37899401783943176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17899110322458378, "compression/movement_sparsity/importance_threshold": -0.005328466686713243, "compression/movement_sparsity/linear_layer_sparsity": 0.06392764481853896, "compression/movement_sparsity/model_sparsity": 0.061731533554583606, "compression_loss": 19.295448303222656, "distillation_loss": 0.5657089352607727, "epoch": 2.19, "learning_rate": 4.743894802755166e-05, "loss": 19.7436, "step": 2593, "task_loss": 0.3073692321777344 }, { "compression/movement_sparsity/importance_regularization_factor": 0.17973141778482238, "compression/movement_sparsity/importance_threshold": -0.005323661937961539, "compression/movement_sparsity/linear_layer_sparsity": 0.0646266872220238, "compression/movement_sparsity/model_sparsity": 0.06240656170100276, "compression_loss": 19.37498664855957, "distillation_loss": 0.48428961634635925, "epoch": 2.19, "learning_rate": 4.743581715716969e-05, "loss": 19.9469, "step": 2594, "task_loss": 0.42430898547172546 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1804712871769687, "compression/movement_sparsity/importance_threshold": -0.005318860078415248, "compression/movement_sparsity/linear_layer_sparsity": 0.06528604599561648, "compression/movement_sparsity/model_sparsity": 0.06304326947229769, "compression_loss": 19.454463958740234, "distillation_loss": 0.7414526343345642, "epoch": 2.19, "learning_rate": 4.743268628678773e-05, "loss": 20.0711, "step": 2595, "task_loss": 0.8540310263633728 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18121071153490753, "compression/movement_sparsity/importance_threshold": -0.00531406110720544, "compression/movement_sparsity/linear_layer_sparsity": 0.06591030001968919, "compression/movement_sparsity/model_sparsity": 0.06364607845021349, "compression_loss": 19.533884048461914, "distillation_loss": 0.54049152135849, "epoch": 2.19, "learning_rate": 4.742955541640576e-05, "loss": 20.164, "step": 2596, "task_loss": 1.4451738595962524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18194969099252412, "compression/movement_sparsity/importance_threshold": -0.00530926502346318, "compression/movement_sparsity/linear_layer_sparsity": 0.06654680016392392, "compression/movement_sparsity/model_sparsity": 0.06426071285639029, "compression_loss": 19.61327362060547, "distillation_loss": 0.7271435260772705, "epoch": 2.2, "learning_rate": 4.7426424546023794e-05, "loss": 20.3323, "step": 2597, "task_loss": 0.30670493841171265 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18268822568370358, "compression/movement_sparsity/importance_threshold": -0.0053044718263195335, "compression/movement_sparsity/linear_layer_sparsity": 0.06712072227640564, "compression/movement_sparsity/model_sparsity": 0.06481491897871733, "compression_loss": 19.69264793395996, "distillation_loss": 1.0440298318862915, "epoch": 2.2, "learning_rate": 4.742329367564183e-05, "loss": 20.7942, "step": 2598, "task_loss": 1.1466645002365112 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18342631574233104, "compression/movement_sparsity/importance_threshold": -0.005299681514905568, "compression/movement_sparsity/linear_layer_sparsity": 0.0678001255757942, "compression/movement_sparsity/model_sparsity": 0.06547098268468288, "compression_loss": 19.77198028564453, "distillation_loss": 0.9339715242385864, "epoch": 2.2, "learning_rate": 4.7420162805259864e-05, "loss": 20.6992, "step": 2599, "task_loss": 1.2223396301269531 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18416396130229173, "compression/movement_sparsity/importance_threshold": -0.005294894088352348, "compression/movement_sparsity/linear_layer_sparsity": 0.06842457038654907, "compression/movement_sparsity/model_sparsity": 0.06607397589517139, "compression_loss": 19.851266860961914, "distillation_loss": 0.330619752407074, "epoch": 2.2, "learning_rate": 4.7417031934877896e-05, "loss": 20.3393, "step": 2600, "task_loss": 0.25710442662239075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18490116249747057, "compression/movement_sparsity/importance_threshold": -0.005290109545790941, "compression/movement_sparsity/linear_layer_sparsity": 0.069041383730017, "compression/movement_sparsity/model_sparsity": 0.06666959980275139, "compression_loss": 19.930498123168945, "distillation_loss": 0.4047538638114929, "epoch": 2.2, "learning_rate": 4.741390106449593e-05, "loss": 20.5087, "step": 2601, "task_loss": 1.6334208250045776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1856379194617529, "compression/movement_sparsity/importance_threshold": -0.005285327886352413, "compression/movement_sparsity/linear_layer_sparsity": 0.06963360943981978, "compression/movement_sparsity/model_sparsity": 0.06724148073752306, "compression_loss": 20.00967025756836, "distillation_loss": 0.916823148727417, "epoch": 2.2, "learning_rate": 4.7410770194113966e-05, "loss": 20.7423, "step": 2602, "task_loss": 0.6348384618759155 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18637423232902384, "compression/movement_sparsity/importance_threshold": -0.005280549109167827, "compression/movement_sparsity/linear_layer_sparsity": 0.07039695887936481, "compression/movement_sparsity/model_sparsity": 0.06797860677548218, "compression_loss": 20.088804244995117, "distillation_loss": 0.5889856815338135, "epoch": 2.2, "learning_rate": 4.7407639323732e-05, "loss": 20.7093, "step": 2603, "task_loss": 0.17347808182239532 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1871101012331684, "compression/movement_sparsity/importance_threshold": -0.005275773213368254, "compression/movement_sparsity/linear_layer_sparsity": 0.07100443559957385, "compression/movement_sparsity/model_sparsity": 0.06856521480153506, "compression_loss": 20.167890548706055, "distillation_loss": 0.8831599950790405, "epoch": 2.2, "learning_rate": 4.740450845335003e-05, "loss": 20.8773, "step": 2604, "task_loss": 0.5865026712417603 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1878455263080716, "compression/movement_sparsity/importance_threshold": -0.0052710001980847574, "compression/movement_sparsity/linear_layer_sparsity": 0.07150812436468035, "compression/movement_sparsity/model_sparsity": 0.06905160030803226, "compression_loss": 20.246902465820312, "distillation_loss": 0.7560316920280457, "epoch": 2.2, "learning_rate": 4.740137758296807e-05, "loss": 20.8261, "step": 2605, "task_loss": 0.7945945262908936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18858050768761903, "compression/movement_sparsity/importance_threshold": -0.005266230062448403, "compression/movement_sparsity/linear_layer_sparsity": 0.07206674777008526, "compression/movement_sparsity/model_sparsity": 0.0695910332809349, "compression_loss": 20.32585906982422, "distillation_loss": 0.5205298662185669, "epoch": 2.2, "learning_rate": 4.73982467125861e-05, "loss": 20.8568, "step": 2606, "task_loss": 0.7792267799377441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.18931504550569556, "compression/movement_sparsity/importance_threshold": -0.0052614628055902566, "compression/movement_sparsity/linear_layer_sparsity": 0.07254007760439084, "compression/movement_sparsity/model_sparsity": 0.0700481027792992, "compression_loss": 20.404787063598633, "distillation_loss": 0.6214619278907776, "epoch": 2.2, "learning_rate": 4.739511584220414e-05, "loss": 20.8781, "step": 2607, "task_loss": 0.4333525002002716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19004913989618644, "compression/movement_sparsity/importance_threshold": -0.005256698426641384, "compression/movement_sparsity/linear_layer_sparsity": 0.07310826419223972, "compression/movement_sparsity/model_sparsity": 0.07059677040990907, "compression_loss": 20.483665466308594, "distillation_loss": 0.7767001986503601, "epoch": 2.2, "learning_rate": 4.739198497182217e-05, "loss": 21.2221, "step": 2608, "task_loss": 0.5853105783462524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19078279099297668, "compression/movement_sparsity/importance_threshold": -0.005251936924732852, "compression/movement_sparsity/linear_layer_sparsity": 0.07383802325155453, "compression/movement_sparsity/model_sparsity": 0.07130146000053497, "compression_loss": 20.562488555908203, "distillation_loss": 0.9071619510650635, "epoch": 2.21, "learning_rate": 4.73888541014402e-05, "loss": 21.2737, "step": 2609, "task_loss": 1.2055329084396362 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19151599892995141, "compression/movement_sparsity/importance_threshold": -0.005247178298995728, "compression/movement_sparsity/linear_layer_sparsity": 0.07456152212286052, "compression/movement_sparsity/model_sparsity": 0.07200010445986872, "compression_loss": 20.641273498535156, "distillation_loss": 0.23543834686279297, "epoch": 2.21, "learning_rate": 4.738572323105824e-05, "loss": 21.0178, "step": 2610, "task_loss": 0.4128836691379547 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19224876384099587, "compression/movement_sparsity/importance_threshold": -0.005242422548561076, "compression/movement_sparsity/linear_layer_sparsity": 0.07509552212209737, "compression/movement_sparsity/model_sparsity": 0.07251575991635564, "compression_loss": 20.71998405456543, "distillation_loss": 0.464539498090744, "epoch": 2.21, "learning_rate": 4.738259236067627e-05, "loss": 21.4524, "step": 2611, "task_loss": 0.9183220863342285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19298108585999518, "compression/movement_sparsity/importance_threshold": -0.005237669672559963, "compression/movement_sparsity/linear_layer_sparsity": 0.07557479019188561, "compression/movement_sparsity/model_sparsity": 0.07297856365354563, "compression_loss": 20.79865074157715, "distillation_loss": 0.5535117387771606, "epoch": 2.21, "learning_rate": 4.7379461490294304e-05, "loss": 21.3809, "step": 2612, "task_loss": 0.7829249501228333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19371296512083447, "compression/movement_sparsity/importance_threshold": -0.005232919670123454, "compression/movement_sparsity/linear_layer_sparsity": 0.07622211748033371, "compression/movement_sparsity/model_sparsity": 0.07360365325822386, "compression_loss": 20.877254486083984, "distillation_loss": 0.31471019983291626, "epoch": 2.21, "learning_rate": 4.737633061991234e-05, "loss": 21.507, "step": 2613, "task_loss": 0.044003866612911224 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19444440175739863, "compression/movement_sparsity/importance_threshold": -0.005228172540382618, "compression/movement_sparsity/linear_layer_sparsity": 0.07689712076186464, "compression/movement_sparsity/model_sparsity": 0.07425546810048123, "compression_loss": 20.9558162689209, "distillation_loss": 0.48466625809669495, "epoch": 2.21, "learning_rate": 4.7373199749530374e-05, "loss": 21.5129, "step": 2614, "task_loss": 0.42754027247428894 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19517539590357325, "compression/movement_sparsity/importance_threshold": -0.005223428282468518, "compression/movement_sparsity/linear_layer_sparsity": 0.07758395281769036, "compression/movement_sparsity/model_sparsity": 0.07491870536224678, "compression_loss": 21.03431510925293, "distillation_loss": 0.3695862591266632, "epoch": 2.21, "learning_rate": 4.7370068879148406e-05, "loss": 21.6682, "step": 2615, "task_loss": 0.589580237865448 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19590594769324332, "compression/movement_sparsity/importance_threshold": -0.005218686895512219, "compression/movement_sparsity/linear_layer_sparsity": 0.07816394433149873, "compression/movement_sparsity/model_sparsity": 0.07547877238329324, "compression_loss": 21.112768173217773, "distillation_loss": 0.4787243604660034, "epoch": 2.21, "learning_rate": 4.736693800876644e-05, "loss": 21.851, "step": 2616, "task_loss": 0.6775349378585815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19663605726029387, "compression/movement_sparsity/importance_threshold": -0.00521394837864479, "compression/movement_sparsity/linear_layer_sparsity": 0.07882336272592959, "compression/movement_sparsity/model_sparsity": 0.07611553772726715, "compression_loss": 21.191179275512695, "distillation_loss": 0.5062515139579773, "epoch": 2.21, "learning_rate": 4.7363807138384477e-05, "loss": 21.8698, "step": 2617, "task_loss": 1.243478536605835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19736572473861014, "compression/movement_sparsity/importance_threshold": -0.005209212730997296, "compression/movement_sparsity/linear_layer_sparsity": 0.07958503085783797, "compression/movement_sparsity/model_sparsity": 0.07685104021567923, "compression_loss": 21.26953125, "distillation_loss": 0.6589154601097107, "epoch": 2.21, "learning_rate": 4.736067626800251e-05, "loss": 21.9804, "step": 2618, "task_loss": 0.6725362539291382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19809495026207724, "compression/movement_sparsity/importance_threshold": -0.005204479951700803, "compression/movement_sparsity/linear_layer_sparsity": 0.08031437257128554, "compression/movement_sparsity/model_sparsity": 0.07755532679755232, "compression_loss": 21.347835540771484, "distillation_loss": 0.8615413904190063, "epoch": 2.21, "learning_rate": 4.735754539762054e-05, "loss": 21.9524, "step": 2619, "task_loss": 0.7157600522041321 }, { "compression/movement_sparsity/importance_regularization_factor": 0.1988237339645803, "compression/movement_sparsity/importance_threshold": -0.005199750039886377, "compression/movement_sparsity/linear_layer_sparsity": 0.08097392213156039, "compression/movement_sparsity/model_sparsity": 0.07819221880141995, "compression_loss": 21.426076889038086, "distillation_loss": 1.1033248901367188, "epoch": 2.21, "learning_rate": 4.735441452723857e-05, "loss": 22.2174, "step": 2620, "task_loss": 1.1565371751785278 }, { "compression/movement_sparsity/importance_regularization_factor": 0.19955207598000446, "compression/movement_sparsity/importance_threshold": -0.005195022994685083, "compression/movement_sparsity/linear_layer_sparsity": 0.08160599840960221, "compression/movement_sparsity/model_sparsity": 0.07880258131481697, "compression_loss": 21.504289627075195, "distillation_loss": 0.900945782661438, "epoch": 2.22, "learning_rate": 4.735128365685661e-05, "loss": 22.1445, "step": 2621, "task_loss": 0.3069593608379364 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20027997644223483, "compression/movement_sparsity/importance_threshold": -0.00519029881522799, "compression/movement_sparsity/linear_layer_sparsity": 0.08229763590498518, "compression/movement_sparsity/model_sparsity": 0.07947045893450773, "compression_loss": 21.58245277404785, "distillation_loss": 0.46861734986305237, "epoch": 2.22, "learning_rate": 4.734815278647464e-05, "loss": 22.1234, "step": 2622, "task_loss": 0.7127574682235718 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20100743548515654, "compression/movement_sparsity/importance_threshold": -0.005185577500646161, "compression/movement_sparsity/linear_layer_sparsity": 0.08287766319129647, "compression/movement_sparsity/model_sparsity": 0.08003056049916157, "compression_loss": 21.660573959350586, "distillation_loss": 0.5023425221443176, "epoch": 2.22, "learning_rate": 4.7345021916092674e-05, "loss": 22.334, "step": 2623, "task_loss": 1.18805992603302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20173445324265482, "compression/movement_sparsity/importance_threshold": -0.005180859050070664, "compression/movement_sparsity/linear_layer_sparsity": 0.0835799608925459, "compression/movement_sparsity/model_sparsity": 0.08070873211385264, "compression_loss": 21.7386474609375, "distillation_loss": 0.36884111166000366, "epoch": 2.22, "learning_rate": 4.734189104571071e-05, "loss": 22.4023, "step": 2624, "task_loss": 0.32678598165512085 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20246102984861447, "compression/movement_sparsity/importance_threshold": -0.005176143462632565, "compression/movement_sparsity/linear_layer_sparsity": 0.08429122556785748, "compression/movement_sparsity/model_sparsity": 0.08139556265946121, "compression_loss": 21.816675186157227, "distillation_loss": 0.5315065979957581, "epoch": 2.22, "learning_rate": 4.7338760175328744e-05, "loss": 22.4444, "step": 2625, "task_loss": 0.5775920748710632 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20318716543692117, "compression/movement_sparsity/importance_threshold": -0.005171430737462929, "compression/movement_sparsity/linear_layer_sparsity": 0.08508952474274324, "compression/movement_sparsity/model_sparsity": 0.08216643780183412, "compression_loss": 21.894657135009766, "distillation_loss": 0.41181349754333496, "epoch": 2.22, "learning_rate": 4.7335629304946776e-05, "loss": 22.6579, "step": 2626, "task_loss": 0.3277665376663208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20391286014145993, "compression/movement_sparsity/importance_threshold": -0.005166720873692821, "compression/movement_sparsity/linear_layer_sparsity": 0.08573363250592965, "compression/movement_sparsity/model_sparsity": 0.08278841848184784, "compression_loss": 21.97260093688965, "distillation_loss": 0.4927271008491516, "epoch": 2.22, "learning_rate": 4.733249843456481e-05, "loss": 22.6129, "step": 2627, "task_loss": 0.7031636238098145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20463811409611565, "compression/movement_sparsity/importance_threshold": -0.005162013870453308, "compression/movement_sparsity/linear_layer_sparsity": 0.08625546985417792, "compression/movement_sparsity/model_sparsity": 0.08329232911182433, "compression_loss": 22.050525665283203, "distillation_loss": 0.6874780654907227, "epoch": 2.22, "learning_rate": 4.7329367564182846e-05, "loss": 22.5714, "step": 2628, "task_loss": 1.1699211597442627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20536292743477358, "compression/movement_sparsity/importance_threshold": -0.005157309726875458, "compression/movement_sparsity/linear_layer_sparsity": 0.0869331799217622, "compression/movement_sparsity/model_sparsity": 0.08394675775370705, "compression_loss": 22.128366470336914, "distillation_loss": 0.47245094180107117, "epoch": 2.22, "learning_rate": 4.732623669380088e-05, "loss": 22.5769, "step": 2629, "task_loss": 0.7732880115509033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20608730029131894, "compression/movement_sparsity/importance_threshold": -0.005152608442090334, "compression/movement_sparsity/linear_layer_sparsity": 0.08746126553385167, "compression/movement_sparsity/model_sparsity": 0.08445670200043988, "compression_loss": 22.206144332885742, "distillation_loss": 0.521500825881958, "epoch": 2.22, "learning_rate": 4.732310582341891e-05, "loss": 22.7048, "step": 2630, "task_loss": 0.24414920806884766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20681123279963676, "compression/movement_sparsity/importance_threshold": -0.005147910015229005, "compression/movement_sparsity/linear_layer_sparsity": 0.08815151982578887, "compression/movement_sparsity/model_sparsity": 0.08512324393397847, "compression_loss": 22.283864974975586, "distillation_loss": 0.5158487558364868, "epoch": 2.22, "learning_rate": 4.731997495303694e-05, "loss": 22.8049, "step": 2631, "task_loss": 0.5783953070640564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20753472509361226, "compression/movement_sparsity/importance_threshold": -0.005143214445422534, "compression/movement_sparsity/linear_layer_sparsity": 0.08888112387092441, "compression/movement_sparsity/model_sparsity": 0.08582778383563903, "compression_loss": 22.36152458190918, "distillation_loss": 0.7815191745758057, "epoch": 2.22, "learning_rate": 4.731684408265498e-05, "loss": 23.1092, "step": 2632, "task_loss": 1.3155665397644043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.20825777730713058, "compression/movement_sparsity/importance_threshold": -0.005138521731801989, "compression/movement_sparsity/linear_layer_sparsity": 0.0896018921078418, "compression/movement_sparsity/model_sparsity": 0.08652379146627584, "compression_loss": 22.439125061035156, "distillation_loss": 0.4389807879924774, "epoch": 2.23, "learning_rate": 4.731371321227301e-05, "loss": 22.9286, "step": 2633, "task_loss": 0.3788175582885742 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2089803895740765, "compression/movement_sparsity/importance_threshold": -0.005133831873498438, "compression/movement_sparsity/linear_layer_sparsity": 0.09031942889532985, "compression/movement_sparsity/model_sparsity": 0.08721667865771233, "compression_loss": 22.51666831970215, "distillation_loss": 0.45241251587867737, "epoch": 2.23, "learning_rate": 4.7310582341891044e-05, "loss": 23.1542, "step": 2634, "task_loss": 0.39344698190689087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2097025620283357, "compression/movement_sparsity/importance_threshold": -0.005129144869642943, "compression/movement_sparsity/linear_layer_sparsity": 0.09111009660292865, "compression/movement_sparsity/model_sparsity": 0.08798018449717675, "compression_loss": 22.594154357910156, "distillation_loss": 0.42139220237731934, "epoch": 2.23, "learning_rate": 4.730745147150908e-05, "loss": 23.2301, "step": 2635, "task_loss": 1.0200010538101196 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2104242948037932, "compression/movement_sparsity/importance_threshold": -0.005124460719366571, "compression/movement_sparsity/linear_layer_sparsity": 0.09192838068277212, "compression/movement_sparsity/model_sparsity": 0.08877035800154132, "compression_loss": 22.67159080505371, "distillation_loss": 0.5375618934631348, "epoch": 2.23, "learning_rate": 4.7304320601127114e-05, "loss": 23.4298, "step": 2636, "task_loss": 1.094788908958435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21114558803433403, "compression/movement_sparsity/importance_threshold": -0.005119779421800389, "compression/movement_sparsity/linear_layer_sparsity": 0.09266711864031674, "compression/movement_sparsity/model_sparsity": 0.0894837180376205, "compression_loss": 22.748964309692383, "distillation_loss": 0.8708018064498901, "epoch": 2.23, "learning_rate": 4.7301189730745146e-05, "loss": 23.2834, "step": 2637, "task_loss": 0.7235084176063538 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2118664418538433, "compression/movement_sparsity/importance_threshold": -0.005115100976075463, "compression/movement_sparsity/linear_layer_sparsity": 0.09343049192819705, "compression/movement_sparsity/model_sparsity": 0.0902208671046512, "compression_loss": 22.82630157470703, "distillation_loss": 0.3416479825973511, "epoch": 2.23, "learning_rate": 4.7298058860363185e-05, "loss": 23.4102, "step": 2638, "task_loss": 0.1553429514169693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21258685639620623, "compression/movement_sparsity/importance_threshold": -0.005110425381322859, "compression/movement_sparsity/linear_layer_sparsity": 0.09412953433168188, "compression/movement_sparsity/model_sparsity": 0.09089589525107036, "compression_loss": 22.903593063354492, "distillation_loss": 0.7303729057312012, "epoch": 2.23, "learning_rate": 4.7294927989981216e-05, "loss": 23.6605, "step": 2639, "task_loss": 0.3490006625652313 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21330683179530796, "compression/movement_sparsity/importance_threshold": -0.0051057526366736425, "compression/movement_sparsity/linear_layer_sparsity": 0.0948044302957041, "compression/movement_sparsity/model_sparsity": 0.09154760646250558, "compression_loss": 22.980825424194336, "distillation_loss": 0.5052585601806641, "epoch": 2.23, "learning_rate": 4.729179711959925e-05, "loss": 23.6157, "step": 2640, "task_loss": 0.6333514451980591 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21402636818503362, "compression/movement_sparsity/importance_threshold": -0.005101082741258879, "compression/movement_sparsity/linear_layer_sparsity": 0.09559488336828545, "compression/movement_sparsity/model_sparsity": 0.09231090504032569, "compression_loss": 23.058002471923828, "distillation_loss": 0.7902531027793884, "epoch": 2.23, "learning_rate": 4.728866624921729e-05, "loss": 23.7985, "step": 2641, "task_loss": 1.7616405487060547 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21474546569926822, "compression/movement_sparsity/importance_threshold": -0.005096415694209637, "compression/movement_sparsity/linear_layer_sparsity": 0.09629112359237586, "compression/movement_sparsity/model_sparsity": 0.09298322727083314, "compression_loss": 23.135135650634766, "distillation_loss": 0.3821699917316437, "epoch": 2.23, "learning_rate": 4.728553537883532e-05, "loss": 23.6615, "step": 2642, "task_loss": 0.6751919984817505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2154641244718971, "compression/movement_sparsity/importance_threshold": -0.00509175149465698, "compression/movement_sparsity/linear_layer_sparsity": 0.09701763927809372, "compression/movement_sparsity/model_sparsity": 0.09368478490772292, "compression_loss": 23.21223258972168, "distillation_loss": 0.29339277744293213, "epoch": 2.23, "learning_rate": 4.728240450845336e-05, "loss": 23.7101, "step": 2643, "task_loss": 0.8954893350601196 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2161823446368053, "compression/movement_sparsity/importance_threshold": -0.005087090141731975, "compression/movement_sparsity/linear_layer_sparsity": 0.09765565379161821, "compression/movement_sparsity/model_sparsity": 0.09430088165994562, "compression_loss": 23.289270401000977, "distillation_loss": 0.6558669805526733, "epoch": 2.23, "learning_rate": 4.727927363807139e-05, "loss": 24.1616, "step": 2644, "task_loss": 0.9075151681900024 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2169001263278777, "compression/movement_sparsity/importance_threshold": -0.00508243163456569, "compression/movement_sparsity/linear_layer_sparsity": 0.09826921183732154, "compression/movement_sparsity/model_sparsity": 0.09489336209925371, "compression_loss": 23.36625862121582, "distillation_loss": 0.6486536264419556, "epoch": 2.24, "learning_rate": 4.727614276768942e-05, "loss": 24.0378, "step": 2645, "task_loss": 0.7238809466362 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21761746967899998, "compression/movement_sparsity/importance_threshold": -0.005077775972289188, "compression/movement_sparsity/linear_layer_sparsity": 0.09900488528378375, "compression/movement_sparsity/model_sparsity": 0.0956037628996337, "compression_loss": 23.443214416503906, "distillation_loss": 0.38805127143859863, "epoch": 2.24, "learning_rate": 4.727301189730746e-05, "loss": 24.0104, "step": 2646, "task_loss": 0.34521588683128357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21833437482405704, "compression/movement_sparsity/importance_threshold": -0.005073123154033535, "compression/movement_sparsity/linear_layer_sparsity": 0.0997314009695016, "compression/movement_sparsity/model_sparsity": 0.09630532053652348, "compression_loss": 23.520103454589844, "distillation_loss": 0.33291494846343994, "epoch": 2.24, "learning_rate": 4.726988102692549e-05, "loss": 24.0241, "step": 2647, "task_loss": 0.8195376396179199 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2190508418969339, "compression/movement_sparsity/importance_threshold": -0.0050684731789297985, "compression/movement_sparsity/linear_layer_sparsity": 0.10055768616582873, "compression/movement_sparsity/model_sparsity": 0.09710322029440618, "compression_loss": 23.596914291381836, "distillation_loss": 0.4051327705383301, "epoch": 2.24, "learning_rate": 4.726675015654352e-05, "loss": 23.9677, "step": 2648, "task_loss": 0.4771091043949127 }, { "compression/movement_sparsity/importance_regularization_factor": 0.21976687103151593, "compression/movement_sparsity/importance_threshold": -0.005063826046109044, "compression/movement_sparsity/linear_layer_sparsity": 0.10125366405823115, "compression/movement_sparsity/model_sparsity": 0.09777528920512615, "compression_loss": 23.6737003326416, "distillation_loss": 0.35507333278656006, "epoch": 2.24, "learning_rate": 4.7263619286161555e-05, "loss": 24.3562, "step": 2649, "task_loss": 0.6983992457389832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.220482462361688, "compression/movement_sparsity/importance_threshold": -0.005059181754702339, "compression/movement_sparsity/linear_layer_sparsity": 0.10209618997087833, "compression/movement_sparsity/model_sparsity": 0.09858887176076102, "compression_loss": 23.750442504882812, "distillation_loss": 0.8532113432884216, "epoch": 2.24, "learning_rate": 4.726048841577959e-05, "loss": 24.4177, "step": 2650, "task_loss": 0.49513962864875793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22119761602133547, "compression/movement_sparsity/importance_threshold": -0.005054540303840747, "compression/movement_sparsity/linear_layer_sparsity": 0.10285001200048231, "compression/movement_sparsity/model_sparsity": 0.09931679768462029, "compression_loss": 23.827123641967773, "distillation_loss": 0.3679310083389282, "epoch": 2.24, "learning_rate": 4.7257357545397625e-05, "loss": 24.4987, "step": 2651, "task_loss": 0.6169549822807312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22191233214434336, "compression/movement_sparsity/importance_threshold": -0.005049901692655337, "compression/movement_sparsity/linear_layer_sparsity": 0.10349130566010666, "compression/movement_sparsity/model_sparsity": 0.0999360609341865, "compression_loss": 23.903757095336914, "distillation_loss": 0.7356378436088562, "epoch": 2.24, "learning_rate": 4.725422667501566e-05, "loss": 24.6139, "step": 2652, "task_loss": 0.6439374685287476 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2226266108645969, "compression/movement_sparsity/importance_threshold": -0.0050452659202771715, "compression/movement_sparsity/linear_layer_sparsity": 0.10416616585162596, "compression/movement_sparsity/model_sparsity": 0.10058773760201432, "compression_loss": 23.98036003112793, "distillation_loss": 0.5498232841491699, "epoch": 2.24, "learning_rate": 4.725109580463369e-05, "loss": 24.5953, "step": 2653, "task_loss": 0.321719765663147 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2233404523159811, "compression/movement_sparsity/importance_threshold": -0.0050406329858373195, "compression/movement_sparsity/linear_layer_sparsity": 0.10491559978753995, "compression/movement_sparsity/model_sparsity": 0.10131142617670122, "compression_loss": 24.056884765625, "distillation_loss": 0.4804812967777252, "epoch": 2.24, "learning_rate": 4.724796493425173e-05, "loss": 24.6566, "step": 2654, "task_loss": 1.3845930099487305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2240538566323813, "compression/movement_sparsity/importance_threshold": -0.005036002888466846, "compression/movement_sparsity/linear_layer_sparsity": 0.10568504247674691, "compression/movement_sparsity/model_sparsity": 0.10205443614245134, "compression_loss": 24.13335609436035, "distillation_loss": 0.2904255986213684, "epoch": 2.24, "learning_rate": 4.724483406386976e-05, "loss": 24.6611, "step": 2655, "task_loss": 2.1085424423217773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2247668239476821, "compression/movement_sparsity/importance_threshold": -0.005031375627296818, "compression/movement_sparsity/linear_layer_sparsity": 0.10645449709012153, "compression/movement_sparsity/model_sparsity": 0.10279745762273726, "compression_loss": 24.20977783203125, "distillation_loss": 0.5884491205215454, "epoch": 2.24, "learning_rate": 4.724170319348779e-05, "loss": 25.0663, "step": 2656, "task_loss": 1.0367165803909302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22547935439576938, "compression/movement_sparsity/importance_threshold": -0.005026751201458299, "compression/movement_sparsity/linear_layer_sparsity": 0.10736742328849083, "compression/movement_sparsity/model_sparsity": 0.1036790219977031, "compression_loss": 24.286161422729492, "distillation_loss": 0.7065024375915527, "epoch": 2.25, "learning_rate": 4.723857232310582e-05, "loss": 24.8922, "step": 2657, "task_loss": 1.2362682819366455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22619144811052805, "compression/movement_sparsity/importance_threshold": -0.005022129610082357, "compression/movement_sparsity/linear_layer_sparsity": 0.1081063758810529, "compression/movement_sparsity/model_sparsity": 0.1043925892954266, "compression_loss": 24.36249351501465, "distillation_loss": 0.5661596059799194, "epoch": 2.25, "learning_rate": 4.723544145272386e-05, "loss": 25.0935, "step": 2658, "task_loss": 1.0125246047973633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22690310522584312, "compression/movement_sparsity/importance_threshold": -0.005017510852300057, "compression/movement_sparsity/linear_layer_sparsity": 0.10873221581942116, "compression/movement_sparsity/model_sparsity": 0.10499692970660307, "compression_loss": 24.438779830932617, "distillation_loss": 0.456632137298584, "epoch": 2.25, "learning_rate": 4.723231058234189e-05, "loss": 25.1371, "step": 2659, "task_loss": 1.001586675643921 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2276143258755997, "compression/movement_sparsity/importance_threshold": -0.005012894927242466, "compression/movement_sparsity/linear_layer_sparsity": 0.10953504617800855, "compression/movement_sparsity/model_sparsity": 0.1057721803725779, "compression_loss": 24.51500701904297, "distillation_loss": 0.44456180930137634, "epoch": 2.25, "learning_rate": 4.7229179711959924e-05, "loss": 25.2269, "step": 2660, "task_loss": 0.2866824269294739 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22832511019368307, "compression/movement_sparsity/importance_threshold": -0.005008281834040651, "compression/movement_sparsity/linear_layer_sparsity": 0.11024932766773202, "compression/movement_sparsity/model_sparsity": 0.1064619240957425, "compression_loss": 24.59119415283203, "distillation_loss": 0.6540637016296387, "epoch": 2.25, "learning_rate": 4.722604884157796e-05, "loss": 25.1216, "step": 2661, "task_loss": 0.14688332378864288 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2290354583139782, "compression/movement_sparsity/importance_threshold": -0.005003671571825675, "compression/movement_sparsity/linear_layer_sparsity": 0.11113156105860661, "compression/movement_sparsity/model_sparsity": 0.10731385005557317, "compression_loss": 24.667316436767578, "distillation_loss": 0.8222883343696594, "epoch": 2.25, "learning_rate": 4.7222917971195995e-05, "loss": 25.2521, "step": 2662, "task_loss": 1.1779440641403198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.22974537037037046, "compression/movement_sparsity/importance_threshold": -0.004999064139728606, "compression/movement_sparsity/linear_layer_sparsity": 0.1118764041899808, "compression/movement_sparsity/model_sparsity": 0.10803310553397917, "compression_loss": 24.743410110473633, "distillation_loss": 0.6208392977714539, "epoch": 2.25, "learning_rate": 4.7219787100814027e-05, "loss": 25.4578, "step": 2663, "task_loss": 1.0132321119308472 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23045484649674486, "compression/movement_sparsity/importance_threshold": -0.00499445953688051, "compression/movement_sparsity/linear_layer_sparsity": 0.11258462820254525, "compression/movement_sparsity/model_sparsity": 0.10871699987296013, "compression_loss": 24.81946563720703, "distillation_loss": 0.5902575254440308, "epoch": 2.25, "learning_rate": 4.721665623043206e-05, "loss": 25.3823, "step": 2664, "task_loss": 0.9076876640319824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2311638868269862, "compression/movement_sparsity/importance_threshold": -0.004989857762412455, "compression/movement_sparsity/linear_layer_sparsity": 0.11327146025837095, "compression/movement_sparsity/model_sparsity": 0.10938023713472568, "compression_loss": 24.89548110961914, "distillation_loss": 0.3179352879524231, "epoch": 2.25, "learning_rate": 4.72135253600501e-05, "loss": 25.4504, "step": 2665, "task_loss": 0.14397116005420685 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23187249149498024, "compression/movement_sparsity/importance_threshold": -0.004985258815455503, "compression/movement_sparsity/linear_layer_sparsity": 0.11383921757618494, "compression/movement_sparsity/model_sparsity": 0.10992849024204694, "compression_loss": 24.971446990966797, "distillation_loss": 0.8588826656341553, "epoch": 2.25, "learning_rate": 4.721039448966813e-05, "loss": 25.6283, "step": 2666, "task_loss": 1.2762912511825562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2325806606346119, "compression/movement_sparsity/importance_threshold": -0.004980662695140721, "compression/movement_sparsity/linear_layer_sparsity": 0.1147489361734602, "compression/movement_sparsity/model_sparsity": 0.11080695720688405, "compression_loss": 25.047380447387695, "distillation_loss": 0.8713862299919128, "epoch": 2.25, "learning_rate": 4.720726361928616e-05, "loss": 25.8066, "step": 2667, "task_loss": 1.0305534601211548 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23328839437976623, "compression/movement_sparsity/importance_threshold": -0.004976069400599178, "compression/movement_sparsity/linear_layer_sparsity": 0.11538695068698468, "compression/movement_sparsity/model_sparsity": 0.11142305395910675, "compression_loss": 25.123260498046875, "distillation_loss": 0.8315489292144775, "epoch": 2.26, "learning_rate": 4.720413274890419e-05, "loss": 25.8843, "step": 2668, "task_loss": 0.8052258491516113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23399569286432842, "compression/movement_sparsity/importance_threshold": -0.0049714789309619365, "compression/movement_sparsity/linear_layer_sparsity": 0.11618086176818045, "compression/movement_sparsity/model_sparsity": 0.11218969175230728, "compression_loss": 25.199073791503906, "distillation_loss": 0.7392635345458984, "epoch": 2.26, "learning_rate": 4.720100187852223e-05, "loss": 25.8108, "step": 2669, "task_loss": 0.6530494093894958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2347025562221836, "compression/movement_sparsity/importance_threshold": -0.004966891285360065, "compression/movement_sparsity/linear_layer_sparsity": 0.11684177068356578, "compression/movement_sparsity/model_sparsity": 0.1128278964132555, "compression_loss": 25.27483558654785, "distillation_loss": 0.28167724609375, "epoch": 2.26, "learning_rate": 4.719787100814026e-05, "loss": 25.7685, "step": 2670, "task_loss": 0.2562342584133148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23540898458721693, "compression/movement_sparsity/importance_threshold": -0.004962306462924627, "compression/movement_sparsity/linear_layer_sparsity": 0.11748302857068722, "compression/movement_sparsity/model_sparsity": 0.1134471251192143, "compression_loss": 25.3505802154541, "distillation_loss": 0.5110008716583252, "epoch": 2.26, "learning_rate": 4.7194740137758294e-05, "loss": 25.9117, "step": 2671, "task_loss": 0.4542040228843689 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2361149780933135, "compression/movement_sparsity/importance_threshold": -0.004957724462786692, "compression/movement_sparsity/linear_layer_sparsity": 0.11820652744199321, "compression/movement_sparsity/model_sparsity": 0.11414576957854806, "compression_loss": 25.4262638092041, "distillation_loss": 0.5463050603866577, "epoch": 2.26, "learning_rate": 4.719160926737633e-05, "loss": 25.9917, "step": 2672, "task_loss": 1.534857153892517 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23682053687435844, "compression/movement_sparsity/importance_threshold": -0.004953145284077323, "compression/movement_sparsity/linear_layer_sparsity": 0.11900173825746128, "compression/movement_sparsity/model_sparsity": 0.1149136624561502, "compression_loss": 25.501888275146484, "distillation_loss": 0.7513213753700256, "epoch": 2.26, "learning_rate": 4.7188478396994365e-05, "loss": 26.098, "step": 2673, "task_loss": 0.8110142350196838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23752566106423678, "compression/movement_sparsity/importance_threshold": -0.004948568925927589, "compression/movement_sparsity/linear_layer_sparsity": 0.11964892245589775, "compression/movement_sparsity/model_sparsity": 0.1155386138863989, "compression_loss": 25.577468872070312, "distillation_loss": 1.0473848581314087, "epoch": 2.26, "learning_rate": 4.71853475266124e-05, "loss": 26.2776, "step": 2674, "task_loss": 0.7302505373954773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2382303507968342, "compression/movement_sparsity/importance_threshold": -0.004943995387468552, "compression/movement_sparsity/linear_layer_sparsity": 0.12026574772353331, "compression/movement_sparsity/model_sparsity": 0.11613424930851471, "compression_loss": 25.65298843383789, "distillation_loss": 0.5484728217124939, "epoch": 2.26, "learning_rate": 4.7182216656230435e-05, "loss": 26.2101, "step": 2675, "task_loss": 0.9423106908798218 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23893460620603502, "compression/movement_sparsity/importance_threshold": -0.004939424667831283, "compression/movement_sparsity/linear_layer_sparsity": 0.12113724936353563, "compression/movement_sparsity/model_sparsity": 0.1169758121861303, "compression_loss": 25.72846794128418, "distillation_loss": 0.5987148880958557, "epoch": 2.26, "learning_rate": 4.7179085785848474e-05, "loss": 26.5761, "step": 2676, "task_loss": 0.67694091796875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.23963842742572505, "compression/movement_sparsity/importance_threshold": -0.0049348567661468435, "compression/movement_sparsity/linear_layer_sparsity": 0.12194618489595258, "compression/movement_sparsity/model_sparsity": 0.11775695829443196, "compression_loss": 25.80389404296875, "distillation_loss": 0.37201178073883057, "epoch": 2.26, "learning_rate": 4.7175954915466505e-05, "loss": 26.2325, "step": 2677, "task_loss": 0.10757792741060257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2403418145897892, "compression/movement_sparsity/importance_threshold": -0.004930291681546301, "compression/movement_sparsity/linear_layer_sparsity": 0.12272310403826724, "compression/movement_sparsity/model_sparsity": 0.11850718787412526, "compression_loss": 25.879234313964844, "distillation_loss": 0.8169753551483154, "epoch": 2.26, "learning_rate": 4.717282404508454e-05, "loss": 26.4643, "step": 2678, "task_loss": 0.484881192445755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2410447678321126, "compression/movement_sparsity/importance_threshold": -0.004925729413160723, "compression/movement_sparsity/linear_layer_sparsity": 0.12348623884279483, "compression/movement_sparsity/model_sparsity": 0.11924410665044008, "compression_loss": 25.95453643798828, "distillation_loss": 0.3653205633163452, "epoch": 2.26, "learning_rate": 4.716969317470257e-05, "loss": 26.6112, "step": 2679, "task_loss": 0.7734857201576233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24174728728658035, "compression/movement_sparsity/importance_threshold": -0.004921169960121173, "compression/movement_sparsity/linear_layer_sparsity": 0.12426460080939342, "compression/movement_sparsity/model_sparsity": 0.11999572948896452, "compression_loss": 26.029794692993164, "distillation_loss": 0.7885230779647827, "epoch": 2.27, "learning_rate": 4.716656230432061e-05, "loss": 26.7258, "step": 2680, "task_loss": 1.2691277265548706 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2424493730870777, "compression/movement_sparsity/importance_threshold": -0.004916613321558719, "compression/movement_sparsity/linear_layer_sparsity": 0.12511629640695257, "compression/movement_sparsity/model_sparsity": 0.1208181667226254, "compression_loss": 26.10498046875, "distillation_loss": 0.47384732961654663, "epoch": 2.27, "learning_rate": 4.716343143393864e-05, "loss": 26.6993, "step": 2681, "task_loss": 0.7312548756599426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24315102536748967, "compression/movement_sparsity/importance_threshold": -0.004912059496604428, "compression/movement_sparsity/linear_layer_sparsity": 0.12587945505981543, "compression/movement_sparsity/model_sparsity": 0.1215551085280118, "compression_loss": 26.18015480041504, "distillation_loss": 0.7642576098442078, "epoch": 2.27, "learning_rate": 4.716030056355667e-05, "loss": 26.9036, "step": 2682, "task_loss": 1.0068151950836182 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2438522442617015, "compression/movement_sparsity/importance_threshold": -0.004907508484389363, "compression/movement_sparsity/linear_layer_sparsity": 0.12664871888650786, "compression/movement_sparsity/model_sparsity": 0.122297945775725, "compression_loss": 26.255264282226562, "distillation_loss": 0.6757956743240356, "epoch": 2.27, "learning_rate": 4.715716969317471e-05, "loss": 26.7836, "step": 2683, "task_loss": 1.1640417575836182 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24455302990359828, "compression/movement_sparsity/importance_threshold": -0.004902960284044592, "compression/movement_sparsity/linear_layer_sparsity": 0.12739203572442467, "compression/movement_sparsity/model_sparsity": 0.1230157273935493, "compression_loss": 26.33033561706543, "distillation_loss": 1.3272367715835571, "epoch": 2.27, "learning_rate": 4.715403882279274e-05, "loss": 27.0489, "step": 2684, "task_loss": 1.1200578212738037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24525338242706485, "compression/movement_sparsity/importance_threshold": -0.004898414894701183, "compression/movement_sparsity/linear_layer_sparsity": 0.1281155465198983, "compression/movement_sparsity/model_sparsity": 0.12371438336741886, "compression_loss": 26.4053955078125, "distillation_loss": 0.5187952518463135, "epoch": 2.27, "learning_rate": 4.715090795241077e-05, "loss": 26.8766, "step": 2685, "task_loss": 1.5124976634979248 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24595330196598697, "compression/movement_sparsity/importance_threshold": -0.004893872315490197, "compression/movement_sparsity/linear_layer_sparsity": 0.12894911738265094, "compression/movement_sparsity/model_sparsity": 0.12451931850667203, "compression_loss": 26.48036766052246, "distillation_loss": 0.9749976396560669, "epoch": 2.27, "learning_rate": 4.7147777082028805e-05, "loss": 27.141, "step": 2686, "task_loss": 0.8362419009208679 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24665278865424944, "compression/movement_sparsity/importance_threshold": -0.004889332545542704, "compression/movement_sparsity/linear_layer_sparsity": 0.12967278319230383, "compression/movement_sparsity/model_sparsity": 0.12521812416950692, "compression_loss": 26.555299758911133, "distillation_loss": 0.6055502891540527, "epoch": 2.27, "learning_rate": 4.7144646211646844e-05, "loss": 27.2536, "step": 2687, "task_loss": 1.073548674583435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2473518426257375, "compression/movement_sparsity/importance_threshold": -0.004884795583989767, "compression/movement_sparsity/linear_layer_sparsity": 0.13035656266121476, "compression/movement_sparsity/model_sparsity": 0.12587841371010905, "compression_loss": 26.63018035888672, "distillation_loss": 0.6765737533569336, "epoch": 2.27, "learning_rate": 4.7141515341264875e-05, "loss": 27.382, "step": 2688, "task_loss": 1.345982551574707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24805046401433617, "compression/movement_sparsity/importance_threshold": -0.004880261429962455, "compression/movement_sparsity/linear_layer_sparsity": 0.13106497746046136, "compression/movement_sparsity/model_sparsity": 0.12656249228166272, "compression_loss": 26.70502471923828, "distillation_loss": 0.37891197204589844, "epoch": 2.27, "learning_rate": 4.713838447088291e-05, "loss": 27.222, "step": 2689, "task_loss": 0.19549186527729034 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24874865295393067, "compression/movement_sparsity/importance_threshold": -0.004875730082591833, "compression/movement_sparsity/linear_layer_sparsity": 0.13188639759639306, "compression/movement_sparsity/model_sparsity": 0.12735569410894126, "compression_loss": 26.779830932617188, "distillation_loss": 0.7147219777107239, "epoch": 2.27, "learning_rate": 4.713525360050094e-05, "loss": 27.3918, "step": 2690, "task_loss": 1.8517550230026245 }, { "compression/movement_sparsity/importance_regularization_factor": 0.24944640957840603, "compression/movement_sparsity/importance_threshold": -0.004871201541008967, "compression/movement_sparsity/linear_layer_sparsity": 0.13265278769868524, "compression/movement_sparsity/model_sparsity": 0.128095756353528, "compression_loss": 26.854557037353516, "distillation_loss": 0.7447193264961243, "epoch": 2.27, "learning_rate": 4.713212273011898e-05, "loss": 27.5517, "step": 2691, "task_loss": 1.0474928617477417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2501437340216476, "compression/movement_sparsity/importance_threshold": -0.004866675804344922, "compression/movement_sparsity/linear_layer_sparsity": 0.13336276456389218, "compression/movement_sparsity/model_sparsity": 0.12878134332927074, "compression_loss": 26.92926788330078, "distillation_loss": 0.4996473789215088, "epoch": 2.28, "learning_rate": 4.712899185973701e-05, "loss": 27.5575, "step": 2692, "task_loss": 0.5966014266014099 }, { "compression/movement_sparsity/importance_regularization_factor": 0.25084062641754024, "compression/movement_sparsity/importance_threshold": -0.0048621528717307655, "compression/movement_sparsity/linear_layer_sparsity": 0.1341137128690959, "compression/movement_sparsity/model_sparsity": 0.12950649425000355, "compression_loss": 27.0039119720459, "distillation_loss": 0.5940051674842834, "epoch": 2.28, "learning_rate": 4.712586098935504e-05, "loss": 27.5952, "step": 2693, "task_loss": 0.37660127878189087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.25153708689996934, "compression/movement_sparsity/importance_threshold": -0.0048576327422975634, "compression/movement_sparsity/linear_layer_sparsity": 0.13483296673672354, "compression/movement_sparsity/model_sparsity": 0.13020103953459444, "compression_loss": 27.078500747680664, "distillation_loss": 0.7594926357269287, "epoch": 2.28, "learning_rate": 4.712273011897307e-05, "loss": 27.7741, "step": 2694, "task_loss": 1.1937370300292969 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2522331156028199, "compression/movement_sparsity/importance_threshold": -0.004853115415176381, "compression/movement_sparsity/linear_layer_sparsity": 0.13558547710788757, "compression/movement_sparsity/model_sparsity": 0.13092769885951633, "compression_loss": 27.153047561645508, "distillation_loss": 0.6717749834060669, "epoch": 2.28, "learning_rate": 4.711959924859111e-05, "loss": 27.871, "step": 2695, "task_loss": 0.5170519351959229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.25292871265997685, "compression/movement_sparsity/importance_threshold": -0.004848600889498287, "compression/movement_sparsity/linear_layer_sparsity": 0.13631204049027598, "compression/movement_sparsity/model_sparsity": 0.13162930255454927, "compression_loss": 27.22756576538086, "distillation_loss": 0.2541246712207794, "epoch": 2.28, "learning_rate": 4.711646837820914e-05, "loss": 27.7618, "step": 2696, "task_loss": 1.2049568891525269 }, { "compression/movement_sparsity/importance_regularization_factor": 0.25362387820532584, "compression/movement_sparsity/importance_threshold": -0.004844089164394343, "compression/movement_sparsity/linear_layer_sparsity": 0.13708433305554793, "compression/movement_sparsity/model_sparsity": 0.1323750644943543, "compression_loss": 27.302026748657227, "distillation_loss": 0.7051365375518799, "epoch": 2.28, "learning_rate": 4.7113337507827175e-05, "loss": 28.0353, "step": 2697, "task_loss": 0.5899141430854797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2543186123727518, "compression/movement_sparsity/importance_threshold": -0.004839580238995617, "compression/movement_sparsity/linear_layer_sparsity": 0.1379760700080283, "compression/movement_sparsity/model_sparsity": 0.13323616753921325, "compression_loss": 27.376441955566406, "distillation_loss": 0.45815545320510864, "epoch": 2.28, "learning_rate": 4.7110206637445213e-05, "loss": 28.0013, "step": 2698, "task_loss": 0.17882615327835083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2550129152961398, "compression/movement_sparsity/importance_threshold": -0.004835074112433176, "compression/movement_sparsity/linear_layer_sparsity": 0.13866005218778904, "compression/movement_sparsity/model_sparsity": 0.1338966528269239, "compression_loss": 27.45081329345703, "distillation_loss": 1.0530192852020264, "epoch": 2.28, "learning_rate": 4.7107075767063245e-05, "loss": 28.2113, "step": 2699, "task_loss": 1.998399019241333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.255706787109375, "compression/movement_sparsity/importance_threshold": -0.004830570783838084, "compression/movement_sparsity/linear_layer_sparsity": 0.13944171714882278, "compression/movement_sparsity/model_sparsity": 0.13465146519186344, "compression_loss": 27.52513885498047, "distillation_loss": 0.49713635444641113, "epoch": 2.28, "learning_rate": 4.710394489668128e-05, "loss": 28.1921, "step": 2700, "task_loss": 1.3201836347579956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.25640022794634254, "compression/movement_sparsity/importance_threshold": -0.004826070252341409, "compression/movement_sparsity/linear_layer_sparsity": 0.14017451687088472, "compression/movement_sparsity/model_sparsity": 0.13535909098911694, "compression_loss": 27.599422454833984, "distillation_loss": 0.4871008098125458, "epoch": 2.28, "learning_rate": 4.710081402629931e-05, "loss": 28.074, "step": 2701, "task_loss": 0.35754257440567017 }, { "compression/movement_sparsity/importance_regularization_factor": 0.25709323794092764, "compression/movement_sparsity/importance_threshold": -0.0048215725170742164, "compression/movement_sparsity/linear_layer_sparsity": 0.14103076750048077, "compression/movement_sparsity/model_sparsity": 0.13618592677545133, "compression_loss": 27.67363739013672, "distillation_loss": 0.29010653495788574, "epoch": 2.28, "learning_rate": 4.709768315591735e-05, "loss": 28.0078, "step": 2702, "task_loss": 0.45912259817123413 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2577858172270152, "compression/movement_sparsity/importance_threshold": -0.004817077577167573, "compression/movement_sparsity/linear_layer_sparsity": 0.14183073605883556, "compression/movement_sparsity/model_sparsity": 0.13695841395283548, "compression_loss": 27.747814178466797, "distillation_loss": 0.5725622177124023, "epoch": 2.28, "learning_rate": 4.709455228553538e-05, "loss": 28.4403, "step": 2703, "task_loss": 0.7410875558853149 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2584779659384907, "compression/movement_sparsity/importance_threshold": -0.004812585431752542, "compression/movement_sparsity/linear_layer_sparsity": 0.14251798546052852, "compression/movement_sparsity/model_sparsity": 0.13762205422335383, "compression_loss": 27.821937561035156, "distillation_loss": 0.759935736656189, "epoch": 2.29, "learning_rate": 4.709142141515341e-05, "loss": 28.6182, "step": 2704, "task_loss": 1.484388828277588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2591696842092388, "compression/movement_sparsity/importance_threshold": -0.004808096079960194, "compression/movement_sparsity/linear_layer_sparsity": 0.14323686967895943, "compression/movement_sparsity/model_sparsity": 0.1383162425573351, "compression_loss": 27.896034240722656, "distillation_loss": 0.6316521763801575, "epoch": 2.29, "learning_rate": 4.708829054477145e-05, "loss": 28.6065, "step": 2705, "task_loss": 0.6817003488540649 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2598609721731452, "compression/movement_sparsity/importance_threshold": -0.004803609520921592, "compression/movement_sparsity/linear_layer_sparsity": 0.1439008073329243, "compression/movement_sparsity/model_sparsity": 0.13895737191037513, "compression_loss": 27.970090866088867, "distillation_loss": 0.42404451966285706, "epoch": 2.29, "learning_rate": 4.708515967438948e-05, "loss": 28.4612, "step": 2706, "task_loss": 0.5528674721717834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.26055182996409476, "compression/movement_sparsity/importance_threshold": -0.004799125753767801, "compression/movement_sparsity/linear_layer_sparsity": 0.14471128108296627, "compression/movement_sparsity/model_sparsity": 0.13974000339379428, "compression_loss": 28.044078826904297, "distillation_loss": 0.6900665760040283, "epoch": 2.29, "learning_rate": 4.708202880400751e-05, "loss": 28.7807, "step": 2707, "task_loss": 0.7807333469390869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.26124225771597265, "compression/movement_sparsity/importance_threshold": -0.00479464477762989, "compression/movement_sparsity/linear_layer_sparsity": 0.14555366390560182, "compression/movement_sparsity/model_sparsity": 0.14055344777499962, "compression_loss": 28.118030548095703, "distillation_loss": 0.7345662117004395, "epoch": 2.29, "learning_rate": 4.707889793362555e-05, "loss": 28.767, "step": 2708, "task_loss": 0.7363647222518921 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2619322555626641, "compression/movement_sparsity/importance_threshold": -0.004790166591638923, "compression/movement_sparsity/linear_layer_sparsity": 0.14624989220552456, "compression/movement_sparsity/model_sparsity": 0.14122575849097127, "compression_loss": 28.19194221496582, "distillation_loss": 0.42932015657424927, "epoch": 2.29, "learning_rate": 4.7075767063243583e-05, "loss": 28.7887, "step": 2709, "task_loss": 0.4497177004814148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2626218236380541, "compression/movement_sparsity/importance_threshold": -0.004785691194925967, "compression/movement_sparsity/linear_layer_sparsity": 0.14679056973863752, "compression/movement_sparsity/model_sparsity": 0.14174786208750312, "compression_loss": 28.26578712463379, "distillation_loss": 0.3798843324184418, "epoch": 2.29, "learning_rate": 4.707263619286162e-05, "loss": 28.7759, "step": 2710, "task_loss": 0.7166100144386292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2633109620760279, "compression/movement_sparsity/importance_threshold": -0.004781218586622086, "compression/movement_sparsity/linear_layer_sparsity": 0.14765464262220268, "compression/movement_sparsity/model_sparsity": 0.14258225140931874, "compression_loss": 28.33956527709961, "distillation_loss": 0.5922620892524719, "epoch": 2.29, "learning_rate": 4.7069505322479654e-05, "loss": 29.0221, "step": 2711, "task_loss": 1.0902985334396362 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2639996710104705, "compression/movement_sparsity/importance_threshold": -0.00477674876585835, "compression/movement_sparsity/linear_layer_sparsity": 0.14843306420963945, "compression/movement_sparsity/model_sparsity": 0.14333393182052215, "compression_loss": 28.413307189941406, "distillation_loss": 0.6019126176834106, "epoch": 2.29, "learning_rate": 4.7066374452097686e-05, "loss": 28.9435, "step": 2712, "task_loss": 0.7821654677391052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2646879505752672, "compression/movement_sparsity/importance_threshold": -0.004772281731765822, "compression/movement_sparsity/linear_layer_sparsity": 0.14919622286250234, "compression/movement_sparsity/model_sparsity": 0.14407087362590856, "compression_loss": 28.486982345581055, "distillation_loss": 0.573569655418396, "epoch": 2.29, "learning_rate": 4.7063243581715724e-05, "loss": 29.0846, "step": 2713, "task_loss": 0.26870492100715637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.265375800904303, "compression/movement_sparsity/importance_threshold": -0.004767817483475569, "compression/movement_sparsity/linear_layer_sparsity": 0.1500267769108431, "compression/movement_sparsity/model_sparsity": 0.14487289558760572, "compression_loss": 28.560609817504883, "distillation_loss": 0.4899322986602783, "epoch": 2.29, "learning_rate": 4.7060112711333756e-05, "loss": 29.2497, "step": 2714, "task_loss": 0.6840153336524963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2660632221314633, "compression/movement_sparsity/importance_threshold": -0.004763356020118657, "compression/movement_sparsity/linear_layer_sparsity": 0.15069683166280515, "compression/movement_sparsity/model_sparsity": 0.14551993189750834, "compression_loss": 28.634187698364258, "distillation_loss": 0.7508284449577332, "epoch": 2.29, "learning_rate": 4.705698184095179e-05, "loss": 29.3023, "step": 2715, "task_loss": 0.9499605298042297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2667502143906326, "compression/movement_sparsity/importance_threshold": -0.004758897340826153, "compression/movement_sparsity/linear_layer_sparsity": 0.15141876846815083, "compression/movement_sparsity/model_sparsity": 0.14621706795265302, "compression_loss": 28.707719802856445, "distillation_loss": 0.8868777751922607, "epoch": 2.3, "learning_rate": 4.705385097056982e-05, "loss": 29.5007, "step": 2716, "task_loss": 1.3008278608322144 }, { "compression/movement_sparsity/importance_regularization_factor": 0.26743677781569686, "compression/movement_sparsity/importance_threshold": -0.00475444144472912, "compression/movement_sparsity/linear_layer_sparsity": 0.15214205270443937, "compression/movement_sparsity/model_sparsity": 0.14691550515034246, "compression_loss": 28.781211853027344, "distillation_loss": 0.469014972448349, "epoch": 2.3, "learning_rate": 4.705072010018786e-05, "loss": 29.308, "step": 2717, "task_loss": 0.5354006290435791 }, { "compression/movement_sparsity/importance_regularization_factor": 0.26812291254054077, "compression/movement_sparsity/importance_threshold": -0.004749988330958627, "compression/movement_sparsity/linear_layer_sparsity": 0.15289944006016648, "compression/movement_sparsity/model_sparsity": 0.14764687392040432, "compression_loss": 28.854673385620117, "distillation_loss": 0.7657325863838196, "epoch": 2.3, "learning_rate": 4.704758922980589e-05, "loss": 29.4644, "step": 2718, "task_loss": 0.43019720911979675 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2688086186990496, "compression/movement_sparsity/importance_threshold": -0.004745537998645738, "compression/movement_sparsity/linear_layer_sparsity": 0.15376377527541965, "compression/movement_sparsity/model_sparsity": 0.1484815165620074, "compression_loss": 28.928071975708008, "distillation_loss": 0.9471403956413269, "epoch": 2.3, "learning_rate": 4.704445835942392e-05, "loss": 29.6058, "step": 2719, "task_loss": 1.2764431238174438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.26949389642510846, "compression/movement_sparsity/importance_threshold": -0.004741090446921521, "compression/movement_sparsity/linear_layer_sparsity": 0.15456262296201667, "compression/movement_sparsity/model_sparsity": 0.14925292137302687, "compression_loss": 29.001449584960938, "distillation_loss": 0.6095449924468994, "epoch": 2.3, "learning_rate": 4.704132748904196e-05, "loss": 29.6766, "step": 2720, "task_loss": 0.6696863770484924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2701787458526025, "compression/movement_sparsity/importance_threshold": -0.004736645674917042, "compression/movement_sparsity/linear_layer_sparsity": 0.15534105647362106, "compression/movement_sparsity/model_sparsity": 0.15000461329876608, "compression_loss": 29.074750900268555, "distillation_loss": 0.34912946820259094, "epoch": 2.3, "learning_rate": 4.703819661865999e-05, "loss": 29.5456, "step": 2721, "task_loss": 0.22652573883533478 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2708631671154169, "compression/movement_sparsity/importance_threshold": -0.004732203681763364, "compression/movement_sparsity/linear_layer_sparsity": 0.1561227572071577, "compression/movement_sparsity/model_sparsity": 0.150759460207313, "compression_loss": 29.147974014282227, "distillation_loss": 0.4300731420516968, "epoch": 2.3, "learning_rate": 4.7035065748278024e-05, "loss": 29.622, "step": 2722, "task_loss": 0.23775401711463928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27154716034743664, "compression/movement_sparsity/importance_threshold": -0.004727764466591556, "compression/movement_sparsity/linear_layer_sparsity": 0.15702937552084764, "compression/movement_sparsity/model_sparsity": 0.15163493339284354, "compression_loss": 29.22113609313965, "distillation_loss": 0.7582014799118042, "epoch": 2.3, "learning_rate": 4.7031934877896055e-05, "loss": 30.1755, "step": 2723, "task_loss": 0.523601770401001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.272230725682547, "compression/movement_sparsity/importance_threshold": -0.004723328028532683, "compression/movement_sparsity/linear_layer_sparsity": 0.157728596786847, "compression/movement_sparsity/model_sparsity": 0.1523101342572996, "compression_loss": 29.294261932373047, "distillation_loss": 0.8197702169418335, "epoch": 2.3, "learning_rate": 4.7028804007514094e-05, "loss": 30.4348, "step": 2724, "task_loss": 0.596432626247406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27291386325463274, "compression/movement_sparsity/importance_threshold": -0.004718894366717814, "compression/movement_sparsity/linear_layer_sparsity": 0.15859249080789764, "compression/movement_sparsity/model_sparsity": 0.15314435086107828, "compression_loss": 29.367353439331055, "distillation_loss": 1.323852777481079, "epoch": 2.3, "learning_rate": 4.7025673137132126e-05, "loss": 30.2297, "step": 2725, "task_loss": 0.902712881565094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27359657319758, "compression/movement_sparsity/importance_threshold": -0.004714463480278008, "compression/movement_sparsity/linear_layer_sparsity": 0.15931901841778315, "compression/movement_sparsity/model_sparsity": 0.15384592001250386, "compression_loss": 29.4404354095459, "distillation_loss": 0.3990781903266907, "epoch": 2.3, "learning_rate": 4.702254226675016e-05, "loss": 29.9856, "step": 2726, "task_loss": 0.4680297076702118 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27427885564527266, "compression/movement_sparsity/importance_threshold": -0.004710035368344339, "compression/movement_sparsity/linear_layer_sparsity": 0.16021306865878487, "compression/movement_sparsity/model_sparsity": 0.15470925687730694, "compression_loss": 29.51342010498047, "distillation_loss": 0.5918540954589844, "epoch": 2.3, "learning_rate": 4.701941139636819e-05, "loss": 30.1544, "step": 2727, "task_loss": 0.49882668256759644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2749607107315968, "compression/movement_sparsity/importance_threshold": -0.004705610030047868, "compression/movement_sparsity/linear_layer_sparsity": 0.16113174230595467, "compression/movement_sparsity/model_sparsity": 0.15559637125852574, "compression_loss": 29.586374282836914, "distillation_loss": 0.8039100170135498, "epoch": 2.31, "learning_rate": 4.701628052598623e-05, "loss": 30.3951, "step": 2728, "task_loss": 0.4539888799190521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27564213859043707, "compression/movement_sparsity/importance_threshold": -0.004701187464519662, "compression/movement_sparsity/linear_layer_sparsity": 0.16199867698975245, "compression/movement_sparsity/model_sparsity": 0.15643352406893204, "compression_loss": 29.659273147583008, "distillation_loss": 0.3955090045928955, "epoch": 2.31, "learning_rate": 4.701314965560426e-05, "loss": 30.3149, "step": 2729, "task_loss": 0.6456724405288696 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27632313935567887, "compression/movement_sparsity/importance_threshold": -0.004696767670890788, "compression/movement_sparsity/linear_layer_sparsity": 0.1628095442373264, "compression/movement_sparsity/model_sparsity": 0.1572165355320324, "compression_loss": 29.73211669921875, "distillation_loss": 0.5394034385681152, "epoch": 2.31, "learning_rate": 4.701001878522229e-05, "loss": 30.4722, "step": 2730, "task_loss": 0.4876680076122284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2770037131612072, "compression/movement_sparsity/importance_threshold": -0.004692350648292312, "compression/movement_sparsity/linear_layer_sparsity": 0.16352864309077478, "compression/movement_sparsity/model_sparsity": 0.15791093112765797, "compression_loss": 29.804920196533203, "distillation_loss": 1.187461256980896, "epoch": 2.31, "learning_rate": 4.700688791484032e-05, "loss": 30.9546, "step": 2731, "task_loss": 1.184269905090332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2776838601409073, "compression/movement_sparsity/importance_threshold": -0.004687936395855298, "compression/movement_sparsity/linear_layer_sparsity": 0.1642765268848961, "compression/movement_sparsity/model_sparsity": 0.1586331228126916, "compression_loss": 29.877683639526367, "distillation_loss": 0.8394446969032288, "epoch": 2.31, "learning_rate": 4.700375704445836e-05, "loss": 30.6821, "step": 2732, "task_loss": 0.520710825920105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27836358042866427, "compression/movement_sparsity/importance_threshold": -0.004683524912710814, "compression/movement_sparsity/linear_layer_sparsity": 0.16507325206965393, "compression/movement_sparsity/model_sparsity": 0.1594024780363396, "compression_loss": 29.95041275024414, "distillation_loss": 0.6270103454589844, "epoch": 2.31, "learning_rate": 4.7000626174076394e-05, "loss": 30.767, "step": 2733, "task_loss": 0.9088036417961121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.27904287415836315, "compression/movement_sparsity/importance_threshold": -0.004679116197989926, "compression/movement_sparsity/linear_layer_sparsity": 0.165703813978406, "compression/movement_sparsity/model_sparsity": 0.16001137820369074, "compression_loss": 30.023086547851562, "distillation_loss": 0.5488346815109253, "epoch": 2.31, "learning_rate": 4.6997495303694425e-05, "loss": 30.8171, "step": 2734, "task_loss": 0.5228575468063354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2797217414638892, "compression/movement_sparsity/importance_threshold": -0.0046747102508237, "compression/movement_sparsity/linear_layer_sparsity": 0.16643190365425178, "compression/movement_sparsity/model_sparsity": 0.1607144557593054, "compression_loss": 30.09574317932129, "distillation_loss": 0.30320632457733154, "epoch": 2.31, "learning_rate": 4.6994364433312464e-05, "loss": 30.6396, "step": 2735, "task_loss": 0.7227944135665894 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2804001824791271, "compression/movement_sparsity/importance_threshold": -0.004670307070343204, "compression/movement_sparsity/linear_layer_sparsity": 0.1671797874483731, "compression/movement_sparsity/model_sparsity": 0.161436647444339, "compression_loss": 30.168363571166992, "distillation_loss": 0.7067972421646118, "epoch": 2.31, "learning_rate": 4.6991233562930496e-05, "loss": 30.8268, "step": 2736, "task_loss": 1.083277702331543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2810781973379629, "compression/movement_sparsity/importance_threshold": -0.004665906655679498, "compression/movement_sparsity/linear_layer_sparsity": 0.16791547281900296, "compression/movement_sparsity/model_sparsity": 0.16214705975925478, "compression_loss": 30.240917205810547, "distillation_loss": 0.599014163017273, "epoch": 2.31, "learning_rate": 4.698810269254853e-05, "loss": 30.9492, "step": 2737, "task_loss": 0.8254029154777527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2817557861742811, "compression/movement_sparsity/importance_threshold": -0.0046615090059636535, "compression/movement_sparsity/linear_layer_sparsity": 0.16859605661098748, "compression/movement_sparsity/model_sparsity": 0.162804263404264, "compression_loss": 30.31342887878418, "distillation_loss": 0.7647791504859924, "epoch": 2.31, "learning_rate": 4.698497182216656e-05, "loss": 31.0584, "step": 2738, "task_loss": 1.2149444818496704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28243294912196704, "compression/movement_sparsity/importance_threshold": -0.004657114120326735, "compression/movement_sparsity/linear_layer_sparsity": 0.16929683994294714, "compression/movement_sparsity/model_sparsity": 0.16348097267290915, "compression_loss": 30.385908126831055, "distillation_loss": 0.7657783627510071, "epoch": 2.32, "learning_rate": 4.69818409517846e-05, "loss": 31.1449, "step": 2739, "task_loss": 1.025955319404602 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28310968631490585, "compression/movement_sparsity/importance_threshold": -0.004652721997899807, "compression/movement_sparsity/linear_layer_sparsity": 0.170157681377083, "compression/movement_sparsity/model_sparsity": 0.16431224155552443, "compression_loss": 30.458330154418945, "distillation_loss": 1.0408014059066772, "epoch": 2.32, "learning_rate": 4.697871008140263e-05, "loss": 31.2471, "step": 2740, "task_loss": 1.416770100593567 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28378599788698256, "compression/movement_sparsity/importance_threshold": -0.004648332637813938, "compression/movement_sparsity/linear_layer_sparsity": 0.17087846153816802, "compression/movement_sparsity/model_sparsity": 0.16500826070069705, "compression_loss": 30.53070640563965, "distillation_loss": 0.8738031387329102, "epoch": 2.32, "learning_rate": 4.697557921102067e-05, "loss": 31.4823, "step": 2741, "task_loss": 1.1227607727050781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2844618839720825, "compression/movement_sparsity/importance_threshold": -0.004643946039200192, "compression/movement_sparsity/linear_layer_sparsity": 0.17157032559273605, "compression/movement_sparsity/model_sparsity": 0.1656763570965679, "compression_loss": 30.603046417236328, "distillation_loss": 0.5981103777885437, "epoch": 2.32, "learning_rate": 4.69724483406387e-05, "loss": 31.1855, "step": 2742, "task_loss": 0.5233943462371826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2851373447040906, "compression/movement_sparsity/importance_threshold": -0.004639562201189637, "compression/movement_sparsity/linear_layer_sparsity": 0.1722999296378716, "compression/movement_sparsity/model_sparsity": 0.16638089699822844, "compression_loss": 30.675294876098633, "distillation_loss": 0.6121476888656616, "epoch": 2.32, "learning_rate": 4.696931747025674e-05, "loss": 31.2467, "step": 2743, "task_loss": 2.0060226917266846 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28581238021689204, "compression/movement_sparsity/importance_threshold": -0.004635181122913338, "compression/movement_sparsity/linear_layer_sparsity": 0.17317339876553386, "compression/movement_sparsity/model_sparsity": 0.16722435977425015, "compression_loss": 30.747495651245117, "distillation_loss": 0.6088132858276367, "epoch": 2.32, "learning_rate": 4.696618659987477e-05, "loss": 31.4186, "step": 2744, "task_loss": 0.4542213976383209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2864869906443719, "compression/movement_sparsity/importance_threshold": -0.004630802803502362, "compression/movement_sparsity/linear_layer_sparsity": 0.173927805079352, "compression/movement_sparsity/model_sparsity": 0.16795284991036336, "compression_loss": 30.819631576538086, "distillation_loss": 0.5590929985046387, "epoch": 2.32, "learning_rate": 4.69630557294928e-05, "loss": 31.3706, "step": 2745, "task_loss": 0.9159040451049805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2871611761204158, "compression/movement_sparsity/importance_threshold": -0.0046264272420877715, "compression/movement_sparsity/linear_layer_sparsity": 0.17473218557973205, "compression/movement_sparsity/model_sparsity": 0.1687295974659915, "compression_loss": 30.891727447509766, "distillation_loss": 0.6650780439376831, "epoch": 2.32, "learning_rate": 4.695992485911084e-05, "loss": 31.5778, "step": 2746, "task_loss": 0.5093753337860107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28783493677890815, "compression/movement_sparsity/importance_threshold": -0.004622054437800638, "compression/movement_sparsity/linear_layer_sparsity": 0.17565407875216352, "compression/movement_sparsity/model_sparsity": 0.16961982077187485, "compression_loss": 30.963783264160156, "distillation_loss": 0.6840406060218811, "epoch": 2.32, "learning_rate": 4.695679398872887e-05, "loss": 31.7772, "step": 2747, "task_loss": 0.7793703079223633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2885082727537346, "compression/movement_sparsity/importance_threshold": -0.004617684389772024, "compression/movement_sparsity/linear_layer_sparsity": 0.17655019187416626, "compression/movement_sparsity/model_sparsity": 0.17048514965137038, "compression_loss": 31.035776138305664, "distillation_loss": 0.7531603574752808, "epoch": 2.32, "learning_rate": 4.6953663118346904e-05, "loss": 31.6833, "step": 2748, "task_loss": 0.5740844011306763 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28918118417878014, "compression/movement_sparsity/importance_threshold": -0.004613317097132996, "compression/movement_sparsity/linear_layer_sparsity": 0.17733474248376788, "compression/movement_sparsity/model_sparsity": 0.17124274853397217, "compression_loss": 31.10773468017578, "distillation_loss": 1.2830471992492676, "epoch": 2.32, "learning_rate": 4.6950532247964936e-05, "loss": 32.0262, "step": 2749, "task_loss": 1.7059133052825928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.28985367118793004, "compression/movement_sparsity/importance_threshold": -0.004608952559014619, "compression/movement_sparsity/linear_layer_sparsity": 0.1780523388920941, "compression/movement_sparsity/model_sparsity": 0.17193569329808764, "compression_loss": 31.179645538330078, "distillation_loss": 0.9924700856208801, "epoch": 2.32, "learning_rate": 4.6947401377582975e-05, "loss": 32.0814, "step": 2750, "task_loss": 1.3663740158081055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2905257339150692, "compression/movement_sparsity/importance_threshold": -0.004604590774547961, "compression/movement_sparsity/linear_layer_sparsity": 0.1787696968170676, "compression/movement_sparsity/model_sparsity": 0.17262840777148722, "compression_loss": 31.251508712768555, "distillation_loss": 0.5187385678291321, "epoch": 2.33, "learning_rate": 4.6944270507201006e-05, "loss": 32.0518, "step": 2751, "task_loss": 0.5828070640563965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29119737249408295, "compression/movement_sparsity/importance_threshold": -0.004600231742864088, "compression/movement_sparsity/linear_layer_sparsity": 0.17949183633326313, "compression/movement_sparsity/model_sparsity": 0.1733257395737404, "compression_loss": 31.32333755493164, "distillation_loss": 0.9489725232124329, "epoch": 2.33, "learning_rate": 4.694113963681904e-05, "loss": 32.0973, "step": 2752, "task_loss": 0.9236429929733276 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2918685870588563, "compression/movement_sparsity/importance_threshold": -0.004595875463094064, "compression/movement_sparsity/linear_layer_sparsity": 0.1803530474165957, "compression/movement_sparsity/model_sparsity": 0.17415736540696533, "compression_loss": 31.395112991333008, "distillation_loss": 0.3054807782173157, "epoch": 2.33, "learning_rate": 4.693800876643707e-05, "loss": 32.0847, "step": 2753, "task_loss": 0.4341506063938141 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29253937774327454, "compression/movement_sparsity/importance_threshold": -0.0045915219343689575, "compression/movement_sparsity/linear_layer_sparsity": 0.18104753478804364, "compression/movement_sparsity/model_sparsity": 0.17482799500071097, "compression_loss": 31.46686553955078, "distillation_loss": 0.4498938322067261, "epoch": 2.33, "learning_rate": 4.693487789605511e-05, "loss": 32.1096, "step": 2754, "task_loss": 0.3062237501144409 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29320974468122263, "compression/movement_sparsity/importance_threshold": -0.004587171155819834, "compression/movement_sparsity/linear_layer_sparsity": 0.18177443204712584, "compression/movement_sparsity/model_sparsity": 0.17552992110274618, "compression_loss": 31.538545608520508, "distillation_loss": 0.8985080718994141, "epoch": 2.33, "learning_rate": 4.693174702567314e-05, "loss": 32.3468, "step": 2755, "task_loss": 1.0426867008209229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2938796880065856, "compression/movement_sparsity/importance_threshold": -0.00458282312657776, "compression/movement_sparsity/linear_layer_sparsity": 0.18255919729174488, "compression/movement_sparsity/model_sparsity": 0.1762877272469923, "compression_loss": 31.610183715820312, "distillation_loss": 0.5811997652053833, "epoch": 2.33, "learning_rate": 4.692861615529117e-05, "loss": 32.4081, "step": 2756, "task_loss": 1.0220046043395996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.294549207853249, "compression/movement_sparsity/importance_threshold": -0.004578477845773799, "compression/movement_sparsity/linear_layer_sparsity": 0.18337714749489456, "compression/movement_sparsity/model_sparsity": 0.1770775783443546, "compression_loss": 31.681785583496094, "distillation_loss": 0.7427621483802795, "epoch": 2.33, "learning_rate": 4.692548528490921e-05, "loss": 32.47, "step": 2757, "task_loss": 1.6808576583862305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29521830435509777, "compression/movement_sparsity/importance_threshold": -0.004574135312539019, "compression/movement_sparsity/linear_layer_sparsity": 0.18420765384656476, "compression/movement_sparsity/model_sparsity": 0.17787955424790858, "compression_loss": 31.75334930419922, "distillation_loss": 0.5990930199623108, "epoch": 2.33, "learning_rate": 4.692235441452724e-05, "loss": 32.5783, "step": 2758, "task_loss": 0.8903379440307617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.295886977646017, "compression/movement_sparsity/importance_threshold": -0.004569795526004485, "compression/movement_sparsity/linear_layer_sparsity": 0.1850246620404712, "compression/movement_sparsity/model_sparsity": 0.17866849569694312, "compression_loss": 31.82484245300293, "distillation_loss": 0.9743495583534241, "epoch": 2.33, "learning_rate": 4.6919223544145274e-05, "loss": 32.6205, "step": 2759, "task_loss": 1.5953030586242676 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29655522785989197, "compression/movement_sparsity/importance_threshold": -0.004565458485301264, "compression/movement_sparsity/linear_layer_sparsity": 0.18578648518655885, "compression/movement_sparsity/model_sparsity": 0.17940414787432055, "compression_loss": 31.896329879760742, "distillation_loss": 0.7787535190582275, "epoch": 2.33, "learning_rate": 4.6916092673763306e-05, "loss": 32.4714, "step": 2760, "task_loss": 1.5599355697631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29722305513060765, "compression/movement_sparsity/importance_threshold": -0.004561124189560422, "compression/movement_sparsity/linear_layer_sparsity": 0.1865649306223309, "compression/movement_sparsity/model_sparsity": 0.18015585131459555, "compression_loss": 31.967737197875977, "distillation_loss": 0.8798046112060547, "epoch": 2.33, "learning_rate": 4.6912961803381344e-05, "loss": 32.7404, "step": 2761, "task_loss": 0.37186360359191895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.2978904595920493, "compression/movement_sparsity/importance_threshold": -0.004556792637913024, "compression/movement_sparsity/linear_layer_sparsity": 0.1872212129606735, "compression/movement_sparsity/model_sparsity": 0.18078958833565548, "compression_loss": 32.03910446166992, "distillation_loss": 0.7092850804328918, "epoch": 2.33, "learning_rate": 4.6909830932999376e-05, "loss": 32.6973, "step": 2762, "task_loss": 1.6282918453216553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29855744137810203, "compression/movement_sparsity/importance_threshold": -0.004552463829490138, "compression/movement_sparsity/linear_layer_sparsity": 0.1880213484573752, "compression/movement_sparsity/model_sparsity": 0.18156223671654076, "compression_loss": 32.11042785644531, "distillation_loss": 0.7414035201072693, "epoch": 2.34, "learning_rate": 4.690670006261741e-05, "loss": 32.7268, "step": 2763, "task_loss": 0.5252096056938171 }, { "compression/movement_sparsity/importance_regularization_factor": 0.299224000622651, "compression/movement_sparsity/importance_threshold": -0.004548137763422827, "compression/movement_sparsity/linear_layer_sparsity": 0.1887589416948268, "compression/movement_sparsity/model_sparsity": 0.18227449135718368, "compression_loss": 32.18168640136719, "distillation_loss": 0.8627614974975586, "epoch": 2.34, "learning_rate": 4.690356919223544e-05, "loss": 32.8268, "step": 2764, "task_loss": 0.5220431089401245 }, { "compression/movement_sparsity/importance_regularization_factor": 0.29989013745958093, "compression/movement_sparsity/importance_threshold": -0.0045438144388421624, "compression/movement_sparsity/linear_layer_sparsity": 0.18955451023532394, "compression/movement_sparsity/model_sparsity": 0.18304272967085963, "compression_loss": 32.252899169921875, "distillation_loss": 0.5500897169113159, "epoch": 2.34, "learning_rate": 4.690043832185348e-05, "loss": 32.8813, "step": 2765, "task_loss": 0.5437158942222595 }, { "compression/movement_sparsity/importance_regularization_factor": 0.300555852022778, "compression/movement_sparsity/importance_threshold": -0.0045394938548792025, "compression/movement_sparsity/linear_layer_sparsity": 0.19042627420701425, "compression/movement_sparsity/model_sparsity": 0.18388454586826272, "compression_loss": 32.32405471801758, "distillation_loss": 0.6657775044441223, "epoch": 2.34, "learning_rate": 4.689730745147151e-05, "loss": 32.9995, "step": 2766, "task_loss": 0.6974325180053711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.30122114444612613, "compression/movement_sparsity/importance_threshold": -0.004535176010665021, "compression/movement_sparsity/linear_layer_sparsity": 0.1912294980631336, "compression/movement_sparsity/model_sparsity": 0.18466017651391878, "compression_loss": 32.395145416259766, "distillation_loss": 0.638074517250061, "epoch": 2.34, "learning_rate": 4.689417658108954e-05, "loss": 33.25, "step": 2767, "task_loss": 1.324385166168213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3018860148635112, "compression/movement_sparsity/importance_threshold": -0.00453086090533068, "compression/movement_sparsity/linear_layer_sparsity": 0.1920876923320543, "compression/movement_sparsity/model_sparsity": 0.1854888891695877, "compression_loss": 32.466209411621094, "distillation_loss": 0.9462876915931702, "epoch": 2.34, "learning_rate": 4.6891045710707574e-05, "loss": 33.1206, "step": 2768, "task_loss": 0.12003374099731445 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3025504634088183, "compression/movement_sparsity/importance_threshold": -0.004526548538007244, "compression/movement_sparsity/linear_layer_sparsity": 0.19289226361911652, "compression/movement_sparsity/model_sparsity": 0.18626582095778854, "compression_loss": 32.53721618652344, "distillation_loss": 0.7075235843658447, "epoch": 2.34, "learning_rate": 4.688791484032561e-05, "loss": 33.4506, "step": 2769, "task_loss": 0.6516581177711487 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3032144902159324, "compression/movement_sparsity/importance_threshold": -0.004522238907825783, "compression/movement_sparsity/linear_layer_sparsity": 0.19371071463730688, "compression/movement_sparsity/model_sparsity": 0.18705615566565423, "compression_loss": 32.608192443847656, "distillation_loss": 0.846467912197113, "epoch": 2.34, "learning_rate": 4.6884783969943644e-05, "loss": 33.4156, "step": 2770, "task_loss": 1.014008641242981 }, { "compression/movement_sparsity/importance_regularization_factor": 0.30387809541873867, "compression/movement_sparsity/importance_threshold": -0.004517932013917359, "compression/movement_sparsity/linear_layer_sparsity": 0.19464503279241493, "compression/movement_sparsity/model_sparsity": 0.18795837711783547, "compression_loss": 32.679107666015625, "distillation_loss": 0.6993773579597473, "epoch": 2.34, "learning_rate": 4.6881653099561676e-05, "loss": 33.4769, "step": 2771, "task_loss": 0.3808341920375824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3045412791511223, "compression/movement_sparsity/importance_threshold": -0.004513627855413042, "compression/movement_sparsity/linear_layer_sparsity": 0.19549691917665626, "compression/movement_sparsity/model_sparsity": 0.18878099858406905, "compression_loss": 32.74999237060547, "distillation_loss": 0.9787783622741699, "epoch": 2.34, "learning_rate": 4.6878522229179714e-05, "loss": 33.6346, "step": 2772, "task_loss": 0.5981093645095825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.30520404154696845, "compression/movement_sparsity/importance_threshold": -0.0045093264314438955, "compression/movement_sparsity/linear_layer_sparsity": 0.19621890367867248, "compression/movement_sparsity/model_sparsity": 0.1894781806973569, "compression_loss": 32.82085037231445, "distillation_loss": 0.4016333222389221, "epoch": 2.34, "learning_rate": 4.6875391358797746e-05, "loss": 33.4316, "step": 2773, "task_loss": 0.7964500188827515 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3058663827401621, "compression/movement_sparsity/importance_threshold": -0.004505027741140987, "compression/movement_sparsity/linear_layer_sparsity": 0.19692449245018942, "compression/movement_sparsity/model_sparsity": 0.19015953032392727, "compression_loss": 32.891666412353516, "distillation_loss": 0.8250312805175781, "epoch": 2.34, "learning_rate": 4.687226048841578e-05, "loss": 33.7039, "step": 2774, "task_loss": 0.4969243109226227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.30652830286458854, "compression/movement_sparsity/importance_threshold": -0.004500731783635381, "compression/movement_sparsity/linear_layer_sparsity": 0.19776092511317467, "compression/movement_sparsity/model_sparsity": 0.19096722895177112, "compression_loss": 32.96241760253906, "distillation_loss": 0.6218572854995728, "epoch": 2.35, "learning_rate": 4.6869129618033816e-05, "loss": 33.523, "step": 2775, "task_loss": 0.26783937215805054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3071898020541327, "compression/movement_sparsity/importance_threshold": -0.004496438558058146, "compression/movement_sparsity/linear_layer_sparsity": 0.19861585216016314, "compression/movement_sparsity/model_sparsity": 0.1917927866246323, "compression_loss": 33.03313064575195, "distillation_loss": 1.0333325862884521, "epoch": 2.35, "learning_rate": 4.686599874765185e-05, "loss": 33.7414, "step": 2776, "task_loss": 1.135291337966919 }, { "compression/movement_sparsity/importance_regularization_factor": 0.30785088044268005, "compression/movement_sparsity/importance_threshold": -0.0044921480635403445, "compression/movement_sparsity/linear_layer_sparsity": 0.19954249115131367, "compression/movement_sparsity/model_sparsity": 0.19268759271576189, "compression_loss": 33.10380172729492, "distillation_loss": 0.6956537961959839, "epoch": 2.35, "learning_rate": 4.686286787726989e-05, "loss": 33.7107, "step": 2777, "task_loss": 0.6610895991325378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3085115381641157, "compression/movement_sparsity/importance_threshold": -0.004487860299213045, "compression/movement_sparsity/linear_layer_sparsity": 0.2004277652049354, "compression/movement_sparsity/model_sparsity": 0.19354245488222016, "compression_loss": 33.174400329589844, "distillation_loss": 0.5103231072425842, "epoch": 2.35, "learning_rate": 4.685973700688792e-05, "loss": 33.6524, "step": 2778, "task_loss": 0.21128392219543457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.30917177535232454, "compression/movement_sparsity/importance_threshold": -0.004483575264207312, "compression/movement_sparsity/linear_layer_sparsity": 0.20131951408158344, "compression/movement_sparsity/model_sparsity": 0.1944035694416149, "compression_loss": 33.24491500854492, "distillation_loss": 0.6218981742858887, "epoch": 2.35, "learning_rate": 4.685660613650595e-05, "loss": 33.9578, "step": 2779, "task_loss": 0.6117700934410095 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3098315921411918, "compression/movement_sparsity/importance_threshold": -0.004479292957654214, "compression/movement_sparsity/linear_layer_sparsity": 0.20212568320710886, "compression/movement_sparsity/model_sparsity": 0.19518204417761223, "compression_loss": 33.31541061401367, "distillation_loss": 0.5278154611587524, "epoch": 2.35, "learning_rate": 4.685347526612399e-05, "loss": 33.9655, "step": 2780, "task_loss": 0.5826143026351929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3104909886646028, "compression/movement_sparsity/importance_threshold": -0.004475013378684815, "compression/movement_sparsity/linear_layer_sparsity": 0.20294075968585829, "compression/movement_sparsity/model_sparsity": 0.19596912027184807, "compression_loss": 33.385841369628906, "distillation_loss": 0.4080013632774353, "epoch": 2.35, "learning_rate": 4.685034439574202e-05, "loss": 33.9353, "step": 2781, "task_loss": 1.4821614027023315 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31114996505644243, "compression/movement_sparsity/importance_threshold": -0.004470736526430182, "compression/movement_sparsity/linear_layer_sparsity": 0.20386588430771907, "compression/movement_sparsity/model_sparsity": 0.19686246401693172, "compression_loss": 33.456214904785156, "distillation_loss": 0.46527233719825745, "epoch": 2.35, "learning_rate": 4.684721352536005e-05, "loss": 34.1553, "step": 2782, "task_loss": 1.2185508012771606 }, { "compression/movement_sparsity/importance_regularization_factor": 0.311808521450596, "compression/movement_sparsity/importance_threshold": -0.00446646240002138, "compression/movement_sparsity/linear_layer_sparsity": 0.20467522526183565, "compression/movement_sparsity/model_sparsity": 0.19764400161945037, "compression_loss": 33.526573181152344, "distillation_loss": 1.2856477499008179, "epoch": 2.35, "learning_rate": 4.684408265497809e-05, "loss": 34.326, "step": 2783, "task_loss": 1.0989196300506592 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31246665798094864, "compression/movement_sparsity/importance_threshold": -0.0044621909985894746, "compression/movement_sparsity/linear_layer_sparsity": 0.20569827114832215, "compression/movement_sparsity/model_sparsity": 0.1986319027324788, "compression_loss": 33.59687042236328, "distillation_loss": 0.6959792375564575, "epoch": 2.35, "learning_rate": 4.684095178459612e-05, "loss": 34.185, "step": 2784, "task_loss": 0.8747627139091492 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3131243747813851, "compression/movement_sparsity/importance_threshold": -0.004457922321265536, "compression/movement_sparsity/linear_layer_sparsity": 0.20658985308662325, "compression/movement_sparsity/model_sparsity": 0.19949285608837242, "compression_loss": 33.66714096069336, "distillation_loss": 0.7672760486602783, "epoch": 2.35, "learning_rate": 4.6837820914214155e-05, "loss": 34.5363, "step": 2785, "task_loss": 0.5969300270080566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3137816719857913, "compression/movement_sparsity/importance_threshold": -0.004453656367180624, "compression/movement_sparsity/linear_layer_sparsity": 0.20742367435689626, "compression/movement_sparsity/model_sparsity": 0.20029803303287727, "compression_loss": 33.73733901977539, "distillation_loss": 0.9064205884933472, "epoch": 2.35, "learning_rate": 4.6834690043832186e-05, "loss": 34.467, "step": 2786, "task_loss": 1.3746144771575928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3144385497280515, "compression/movement_sparsity/importance_threshold": -0.00444939313546581, "compression/movement_sparsity/linear_layer_sparsity": 0.2082836095542918, "compression/movement_sparsity/model_sparsity": 0.20112842681077217, "compression_loss": 33.80751037597656, "distillation_loss": 1.2024576663970947, "epoch": 2.36, "learning_rate": 4.6831559173450225e-05, "loss": 34.6611, "step": 2787, "task_loss": 1.168367624282837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3150950081420516, "compression/movement_sparsity/importance_threshold": -0.0044451326252521565, "compression/movement_sparsity/linear_layer_sparsity": 0.20913857237378317, "compression/movement_sparsity/model_sparsity": 0.20195401902724072, "compression_loss": 33.877628326416016, "distillation_loss": 0.608418881893158, "epoch": 2.36, "learning_rate": 4.682842830306826e-05, "loss": 34.6205, "step": 2788, "task_loss": 0.888451099395752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31575104736167636, "compression/movement_sparsity/importance_threshold": -0.0044408748356707314, "compression/movement_sparsity/linear_layer_sparsity": 0.20996006405472067, "compression/movement_sparsity/model_sparsity": 0.20274728994173402, "compression_loss": 33.947723388671875, "distillation_loss": 1.0934152603149414, "epoch": 2.36, "learning_rate": 4.682529743268629e-05, "loss": 34.7478, "step": 2789, "task_loss": 1.0978530645370483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31640666752081104, "compression/movement_sparsity/importance_threshold": -0.004436619765852599, "compression/movement_sparsity/linear_layer_sparsity": 0.210827177601033, "compression/movement_sparsity/model_sparsity": 0.20358461547017725, "compression_loss": 34.017784118652344, "distillation_loss": 0.40207523107528687, "epoch": 2.36, "learning_rate": 4.682216656230432e-05, "loss": 34.6597, "step": 2790, "task_loss": 0.6421834230422974 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31706186875334064, "compression/movement_sparsity/importance_threshold": -0.004432367414928829, "compression/movement_sparsity/linear_layer_sparsity": 0.21160716125443008, "compression/movement_sparsity/model_sparsity": 0.20433780428556975, "compression_loss": 34.087772369384766, "distillation_loss": 0.4820161759853363, "epoch": 2.36, "learning_rate": 4.681903569192236e-05, "loss": 34.6683, "step": 2791, "task_loss": 0.19914323091506958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3177166511931504, "compression/movement_sparsity/importance_threshold": -0.0044281177820304825, "compression/movement_sparsity/linear_layer_sparsity": 0.2125231877363847, "compression/movement_sparsity/model_sparsity": 0.20522236243984215, "compression_loss": 34.15770721435547, "distillation_loss": 0.5720499157905579, "epoch": 2.36, "learning_rate": 4.681590482154039e-05, "loss": 34.7568, "step": 2792, "task_loss": 0.24341435730457306 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31837101497412545, "compression/movement_sparsity/importance_threshold": -0.004423870866288629, "compression/movement_sparsity/linear_layer_sparsity": 0.21339932787759738, "compression/movement_sparsity/model_sparsity": 0.20606840447188185, "compression_loss": 34.22758865356445, "distillation_loss": 0.9194385409355164, "epoch": 2.36, "learning_rate": 4.681277395115842e-05, "loss": 34.9481, "step": 2793, "task_loss": 1.1347148418426514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3190249602301509, "compression/movement_sparsity/importance_threshold": -0.004419626666834334, "compression/movement_sparsity/linear_layer_sparsity": 0.21421785044079356, "compression/movement_sparsity/model_sparsity": 0.2068588082669623, "compression_loss": 34.29742431640625, "distillation_loss": 0.8019289374351501, "epoch": 2.36, "learning_rate": 4.680964308077646e-05, "loss": 35.0083, "step": 2794, "task_loss": 1.3380203247070312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.31967848709511204, "compression/movement_sparsity/importance_threshold": -0.0044153851827986625, "compression/movement_sparsity/linear_layer_sparsity": 0.21505425925544353, "compression/movement_sparsity/model_sparsity": 0.20766648386573458, "compression_loss": 34.367218017578125, "distillation_loss": 0.45591774582862854, "epoch": 2.36, "learning_rate": 4.680651221039449e-05, "loss": 34.9082, "step": 2795, "task_loss": 0.09135734289884567 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3203315957028935, "compression/movement_sparsity/importance_threshold": -0.004411146413312683, "compression/movement_sparsity/linear_layer_sparsity": 0.21581175392867935, "compression/movement_sparsity/model_sparsity": 0.20839795626661856, "compression_loss": 34.43696212768555, "distillation_loss": 0.954008936882019, "epoch": 2.36, "learning_rate": 4.6803381340012525e-05, "loss": 35.1674, "step": 2796, "task_loss": 0.6402214169502258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3209842861873812, "compression/movement_sparsity/importance_threshold": -0.004406910357507458, "compression/movement_sparsity/linear_layer_sparsity": 0.2167094291166424, "compression/movement_sparsity/model_sparsity": 0.2092647935503032, "compression_loss": 34.50669479370117, "distillation_loss": 0.5899990200996399, "epoch": 2.36, "learning_rate": 4.6800250469630556e-05, "loss": 35.1971, "step": 2797, "task_loss": 0.7693442106246948 }, { "compression/movement_sparsity/importance_regularization_factor": 0.32163655868245977, "compression/movement_sparsity/importance_threshold": -0.004402677014514054, "compression/movement_sparsity/linear_layer_sparsity": 0.2175462075804891, "compression/movement_sparsity/model_sparsity": 0.21007282609968508, "compression_loss": 34.57639694213867, "distillation_loss": 0.6711979508399963, "epoch": 2.36, "learning_rate": 4.6797119599248595e-05, "loss": 35.1043, "step": 2798, "task_loss": 0.2544258236885071 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3222884133220144, "compression/movement_sparsity/importance_threshold": -0.004398446383463541, "compression/movement_sparsity/linear_layer_sparsity": 0.21846962704637796, "compression/movement_sparsity/model_sparsity": 0.21096452326615012, "compression_loss": 34.64600372314453, "distillation_loss": 0.5910271406173706, "epoch": 2.37, "learning_rate": 4.679398872886663e-05, "loss": 35.1941, "step": 2799, "task_loss": 1.1743701696395874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.32293985023993044, "compression/movement_sparsity/importance_threshold": -0.00439421846348698, "compression/movement_sparsity/linear_layer_sparsity": 0.21949689408820755, "compression/movement_sparsity/model_sparsity": 0.21195650052484982, "compression_loss": 34.71552658081055, "distillation_loss": 0.819922924041748, "epoch": 2.37, "learning_rate": 4.679085785848466e-05, "loss": 35.4387, "step": 2800, "task_loss": 0.3998600244522095 }, { "compression/movement_sparsity/importance_regularization_factor": 0.32359086957009287, "compression/movement_sparsity/importance_threshold": -0.004389993253715439, "compression/movement_sparsity/linear_layer_sparsity": 0.22037016050501998, "compression/movement_sparsity/model_sparsity": 0.212799767553763, "compression_loss": 34.784976959228516, "distillation_loss": 0.35562777519226074, "epoch": 2.37, "learning_rate": 4.678772698810269e-05, "loss": 35.2764, "step": 2801, "task_loss": 1.1417474746704102 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3242414714463868, "compression/movement_sparsity/importance_threshold": -0.004385770753279986, "compression/movement_sparsity/linear_layer_sparsity": 0.22143532255159634, "compression/movement_sparsity/model_sparsity": 0.21382833800721776, "compression_loss": 34.854408264160156, "distillation_loss": 1.0053315162658691, "epoch": 2.37, "learning_rate": 4.678459611772073e-05, "loss": 35.5293, "step": 2802, "task_loss": 0.7591885328292847 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3248916560026974, "compression/movement_sparsity/importance_threshold": -0.004381550961311685, "compression/movement_sparsity/linear_layer_sparsity": 0.22240319531443173, "compression/movement_sparsity/model_sparsity": 0.21476296136312484, "compression_loss": 34.92377471923828, "distillation_loss": 0.8699702024459839, "epoch": 2.37, "learning_rate": 4.678146524733876e-05, "loss": 35.6001, "step": 2803, "task_loss": 0.7717055082321167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3255414233729099, "compression/movement_sparsity/importance_threshold": -0.004377333876941602, "compression/movement_sparsity/linear_layer_sparsity": 0.22351935702598666, "compression/movement_sparsity/model_sparsity": 0.21584077948617283, "compression_loss": 34.993125915527344, "distillation_loss": 0.5772320628166199, "epoch": 2.37, "learning_rate": 4.677833437695679e-05, "loss": 35.722, "step": 2804, "task_loss": 0.8272057175636292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3261907736909093, "compression/movement_sparsity/importance_threshold": -0.004373119499300805, "compression/movement_sparsity/linear_layer_sparsity": 0.2243514015952819, "compression/movement_sparsity/model_sparsity": 0.2166442407648443, "compression_loss": 35.062400817871094, "distillation_loss": 0.9692050218582153, "epoch": 2.37, "learning_rate": 4.6775203506574824e-05, "loss": 35.8433, "step": 2805, "task_loss": 1.8767869472503662 }, { "compression/movement_sparsity/importance_regularization_factor": 0.32683970709058086, "compression/movement_sparsity/importance_threshold": -0.004368907827520357, "compression/movement_sparsity/linear_layer_sparsity": 0.22530465532859573, "compression/movement_sparsity/model_sparsity": 0.21756474729986727, "compression_loss": 35.13166046142578, "distillation_loss": 0.9171130657196045, "epoch": 2.37, "learning_rate": 4.677207263619286e-05, "loss": 35.9384, "step": 2806, "task_loss": 0.42592653632164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.32748822370580943, "compression/movement_sparsity/importance_threshold": -0.0043646988607313275, "compression/movement_sparsity/linear_layer_sparsity": 0.22622999458547396, "compression/movement_sparsity/model_sparsity": 0.21845829830659522, "compression_loss": 35.20086669921875, "distillation_loss": 0.5807700157165527, "epoch": 2.37, "learning_rate": 4.6768941765810894e-05, "loss": 35.8018, "step": 2807, "task_loss": 0.7594570517539978 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3281363236704806, "compression/movement_sparsity/importance_threshold": -0.004360492598064779, "compression/movement_sparsity/linear_layer_sparsity": 0.227263927237012, "compression/movement_sparsity/model_sparsity": 0.21945671219080407, "compression_loss": 35.270023345947266, "distillation_loss": 0.6636044979095459, "epoch": 2.37, "learning_rate": 4.676581089542893e-05, "loss": 35.9512, "step": 2808, "task_loss": 1.0169124603271484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3287840071184793, "compression/movement_sparsity/importance_threshold": -0.00435628903865178, "compression/movement_sparsity/linear_layer_sparsity": 0.22812169223589782, "compression/movement_sparsity/model_sparsity": 0.22028501032318434, "compression_loss": 35.339195251464844, "distillation_loss": 0.9646652936935425, "epoch": 2.37, "learning_rate": 4.6762680025046965e-05, "loss": 36.1959, "step": 2809, "task_loss": 0.7792971134185791 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3294312741836908, "compression/movement_sparsity/importance_threshold": -0.004352088181623394, "compression/movement_sparsity/linear_layer_sparsity": 0.2291586059293448, "compression/movement_sparsity/model_sparsity": 0.22128630284134182, "compression_loss": 35.40829849243164, "distillation_loss": 0.9661957025527954, "epoch": 2.38, "learning_rate": 4.6759549154664997e-05, "loss": 36.3452, "step": 2810, "task_loss": 0.8791704177856445 }, { "compression/movement_sparsity/importance_regularization_factor": 0.330078125, "compression/movement_sparsity/importance_threshold": -0.00434789002611069, "compression/movement_sparsity/linear_layer_sparsity": 0.22992827517773687, "compression/movement_sparsity/model_sparsity": 0.22202953158327204, "compression_loss": 35.477359771728516, "distillation_loss": 0.45383894443511963, "epoch": 2.38, "learning_rate": 4.6756418284283035e-05, "loss": 36.2553, "step": 2811, "task_loss": 0.9472261071205139 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3307245597012922, "compression/movement_sparsity/importance_threshold": -0.0043436945712447324, "compression/movement_sparsity/linear_layer_sparsity": 0.23080451071229063, "compression/movement_sparsity/model_sparsity": 0.2228756657315981, "compression_loss": 35.54640579223633, "distillation_loss": 1.051933765411377, "epoch": 2.38, "learning_rate": 4.675328741390107e-05, "loss": 36.3503, "step": 2812, "task_loss": 0.9278545379638672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33137057842145246, "compression/movement_sparsity/importance_threshold": -0.004339501816156588, "compression/movement_sparsity/linear_layer_sparsity": 0.23150987292462247, "compression/movement_sparsity/model_sparsity": 0.22355679658198835, "compression_loss": 35.615379333496094, "distillation_loss": 0.41822201013565063, "epoch": 2.38, "learning_rate": 4.6750156543519105e-05, "loss": 36.1521, "step": 2813, "task_loss": 0.7499445080757141 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33201618129436594, "compression/movement_sparsity/importance_threshold": -0.004335311759977322, "compression/movement_sparsity/linear_layer_sparsity": 0.2322364840036814, "compression/movement_sparsity/model_sparsity": 0.22425844633516448, "compression_loss": 35.68429946899414, "distillation_loss": 0.420144259929657, "epoch": 2.38, "learning_rate": 4.674702567313714e-05, "loss": 36.3622, "step": 2814, "task_loss": 0.26806408166885376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3326613684539179, "compression/movement_sparsity/importance_threshold": -0.004331124401838001, "compression/movement_sparsity/linear_layer_sparsity": 0.23313854728533445, "compression/movement_sparsity/model_sparsity": 0.2251295209680215, "compression_loss": 35.75319290161133, "distillation_loss": 1.0627728700637817, "epoch": 2.38, "learning_rate": 4.674389480275517e-05, "loss": 36.5908, "step": 2815, "task_loss": 1.0578289031982422 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33330614003399306, "compression/movement_sparsity/importance_threshold": -0.004326939740869692, "compression/movement_sparsity/linear_layer_sparsity": 0.23399674155425515, "compression/movement_sparsity/model_sparsity": 0.22595823362369039, "compression_loss": 35.82202911376953, "distillation_loss": 0.7408710718154907, "epoch": 2.38, "learning_rate": 4.67407639323732e-05, "loss": 36.7179, "step": 2816, "task_loss": 0.9701938033103943 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33395049616847705, "compression/movement_sparsity/importance_threshold": -0.004322757776203461, "compression/movement_sparsity/linear_layer_sparsity": 0.23480283913477476, "compression/movement_sparsity/model_sparsity": 0.22673663927247295, "compression_loss": 35.89083480834961, "distillation_loss": 0.8855273127555847, "epoch": 2.38, "learning_rate": 4.673763306199124e-05, "loss": 36.4909, "step": 2817, "task_loss": 0.5000036954879761 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3345944369912549, "compression/movement_sparsity/importance_threshold": -0.00431857850697037, "compression/movement_sparsity/linear_layer_sparsity": 0.2356551905615539, "compression/movement_sparsity/model_sparsity": 0.2275597098056025, "compression_loss": 35.95958709716797, "distillation_loss": 0.4469972848892212, "epoch": 2.38, "learning_rate": 4.673450219160927e-05, "loss": 36.5131, "step": 2818, "task_loss": 0.14825467765331268 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33523796263621175, "compression/movement_sparsity/importance_threshold": -0.004314401932301489, "compression/movement_sparsity/linear_layer_sparsity": 0.23644581057248215, "compression/movement_sparsity/model_sparsity": 0.22832316958692375, "compression_loss": 36.02830505371094, "distillation_loss": 0.32069677114486694, "epoch": 2.38, "learning_rate": 4.67313713212273e-05, "loss": 36.5906, "step": 2819, "task_loss": 0.27602848410606384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3358810732372326, "compression/movement_sparsity/importance_threshold": -0.004310228051327884, "compression/movement_sparsity/linear_layer_sparsity": 0.23734653834735997, "compression/movement_sparsity/model_sparsity": 0.22919295459177177, "compression_loss": 36.096946716308594, "distillation_loss": 0.6055190563201904, "epoch": 2.38, "learning_rate": 4.672824045084534e-05, "loss": 37.0309, "step": 2820, "task_loss": 1.2245550155639648 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3365237689282028, "compression/movement_sparsity/importance_threshold": -0.004306056863180619, "compression/movement_sparsity/linear_layer_sparsity": 0.23813849386506344, "compression/movement_sparsity/model_sparsity": 0.229957704001102, "compression_loss": 36.16558837890625, "distillation_loss": 1.0037580728530884, "epoch": 2.38, "learning_rate": 4.672510958046337e-05, "loss": 36.9062, "step": 2821, "task_loss": 1.0853989124298096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33716604984300724, "compression/movement_sparsity/importance_threshold": -0.004301888366990761, "compression/movement_sparsity/linear_layer_sparsity": 0.23908503429199826, "compression/movement_sparsity/model_sparsity": 0.23087172785247265, "compression_loss": 36.23418045043945, "distillation_loss": 1.1358416080474854, "epoch": 2.39, "learning_rate": 4.6721978710081405e-05, "loss": 37.1708, "step": 2822, "task_loss": 0.7193024158477783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3378079161155312, "compression/movement_sparsity/importance_threshold": -0.004297722561889377, "compression/movement_sparsity/linear_layer_sparsity": 0.2397595129101532, "compression/movement_sparsity/model_sparsity": 0.23152303605515503, "compression_loss": 36.302734375, "distillation_loss": 0.6037541627883911, "epoch": 2.39, "learning_rate": 4.671884783969944e-05, "loss": 37.0054, "step": 2823, "task_loss": 0.9604467153549194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33844936787965985, "compression/movement_sparsity/importance_threshold": -0.004293559447007532, "compression/movement_sparsity/linear_layer_sparsity": 0.24057479209975244, "compression/movement_sparsity/model_sparsity": 0.2323103078964994, "compression_loss": 36.371219635009766, "distillation_loss": 0.8272299766540527, "epoch": 2.39, "learning_rate": 4.6715716969317475e-05, "loss": 37.1179, "step": 2824, "task_loss": 0.6271184086799622 }, { "compression/movement_sparsity/importance_regularization_factor": 0.33909040526927825, "compression/movement_sparsity/importance_threshold": -0.004289399021476292, "compression/movement_sparsity/linear_layer_sparsity": 0.2413139116306614, "compression/movement_sparsity/model_sparsity": 0.23302403639772398, "compression_loss": 36.439697265625, "distillation_loss": 0.45521846413612366, "epoch": 2.39, "learning_rate": 4.671258609893551e-05, "loss": 36.9652, "step": 2825, "task_loss": 0.5182157158851624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3397310284182715, "compression/movement_sparsity/importance_threshold": -0.004285241284426726, "compression/movement_sparsity/linear_layer_sparsity": 0.24209581507504785, "compression/movement_sparsity/model_sparsity": 0.2337790790533794, "compression_loss": 36.50811004638672, "distillation_loss": 0.5483400821685791, "epoch": 2.39, "learning_rate": 4.670945522855354e-05, "loss": 37.2186, "step": 2826, "task_loss": 0.0644339844584465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34037123746052467, "compression/movement_sparsity/importance_threshold": -0.0042810862349898955, "compression/movement_sparsity/linear_layer_sparsity": 0.2429704527711384, "compression/movement_sparsity/model_sparsity": 0.23462367025390898, "compression_loss": 36.57645034790039, "distillation_loss": 0.6850944757461548, "epoch": 2.39, "learning_rate": 4.670632435817157e-05, "loss": 37.3146, "step": 2827, "task_loss": 0.3326924741268158 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34101103252992326, "compression/movement_sparsity/importance_threshold": -0.0042769338722968685, "compression/movement_sparsity/linear_layer_sparsity": 0.2438221602928652, "compression/movement_sparsity/model_sparsity": 0.23544611900210566, "compression_loss": 36.644775390625, "distillation_loss": 0.4832216799259186, "epoch": 2.39, "learning_rate": 4.670319348778961e-05, "loss": 37.2572, "step": 2828, "task_loss": 0.4763545095920563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3416504137603522, "compression/movement_sparsity/importance_threshold": -0.00427278419547871, "compression/movement_sparsity/linear_layer_sparsity": 0.24460251359545898, "compression/movement_sparsity/model_sparsity": 0.23619966476810778, "compression_loss": 36.71302795410156, "distillation_loss": 1.4766987562179565, "epoch": 2.39, "learning_rate": 4.670006261740764e-05, "loss": 37.8993, "step": 2829, "task_loss": 2.3835268020629883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34228938128569664, "compression/movement_sparsity/importance_threshold": -0.004268637203666489, "compression/movement_sparsity/linear_layer_sparsity": 0.24554920903657304, "compression/movement_sparsity/model_sparsity": 0.23711383830844376, "compression_loss": 36.78125, "distillation_loss": 0.9806388020515442, "epoch": 2.39, "learning_rate": 4.669693174702567e-05, "loss": 37.6856, "step": 2830, "task_loss": 0.8835890889167786 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3429279352398418, "compression/movement_sparsity/importance_threshold": -0.004264492895991267, "compression/movement_sparsity/linear_layer_sparsity": 0.24638884930065233, "compression/movement_sparsity/model_sparsity": 0.23792463434641636, "compression_loss": 36.849430084228516, "distillation_loss": 0.5653359889984131, "epoch": 2.39, "learning_rate": 4.669380087664371e-05, "loss": 37.4069, "step": 2831, "task_loss": 0.041991543024778366 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34356607575667253, "compression/movement_sparsity/importance_threshold": -0.004260351271584114, "compression/movement_sparsity/linear_layer_sparsity": 0.24714474613542498, "compression/movement_sparsity/model_sparsity": 0.2386545637995039, "compression_loss": 36.917537689208984, "distillation_loss": 0.5075536966323853, "epoch": 2.39, "learning_rate": 4.669067000626174e-05, "loss": 37.6353, "step": 2832, "task_loss": 0.522309422492981 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34420380297007425, "compression/movement_sparsity/importance_threshold": -0.004256212329576094, "compression/movement_sparsity/linear_layer_sparsity": 0.24796609472635084, "compression/movement_sparsity/model_sparsity": 0.23944769653956766, "compression_loss": 36.98560333251953, "distillation_loss": 0.6266186833381653, "epoch": 2.39, "learning_rate": 4.6687539135879775e-05, "loss": 37.6607, "step": 2833, "task_loss": 1.0996663570404053 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34484111701393205, "compression/movement_sparsity/importance_threshold": -0.0042520760690982744, "compression/movement_sparsity/linear_layer_sparsity": 0.24884527553031066, "compression/movement_sparsity/model_sparsity": 0.24029667477823494, "compression_loss": 37.053619384765625, "distillation_loss": 0.7030753493309021, "epoch": 2.4, "learning_rate": 4.668440826549781e-05, "loss": 37.6866, "step": 2834, "task_loss": 0.8611600995063782 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3454780180221311, "compression/movement_sparsity/importance_threshold": -0.00424794248928172, "compression/movement_sparsity/linear_layer_sparsity": 0.24968933966058285, "compression/movement_sparsity/model_sparsity": 0.2411117427089873, "compression_loss": 37.121551513671875, "distillation_loss": 0.6225899457931519, "epoch": 2.4, "learning_rate": 4.6681277395115845e-05, "loss": 37.6735, "step": 2835, "task_loss": 0.6386725306510925 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3461145061285561, "compression/movement_sparsity/importance_threshold": -0.004243811589257499, "compression/movement_sparsity/linear_layer_sparsity": 0.25045539588618126, "compression/movement_sparsity/model_sparsity": 0.24185148254657177, "compression_loss": 37.189449310302734, "distillation_loss": 0.6571491956710815, "epoch": 2.4, "learning_rate": 4.667814652473388e-05, "loss": 37.8585, "step": 2836, "task_loss": 1.6863741874694824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34675058146709314, "compression/movement_sparsity/importance_threshold": -0.004239683368156673, "compression/movement_sparsity/linear_layer_sparsity": 0.25131629694115526, "compression/movement_sparsity/model_sparsity": 0.24268280900186603, "compression_loss": 37.257301330566406, "distillation_loss": 0.5725002288818359, "epoch": 2.4, "learning_rate": 4.667501565435191e-05, "loss": 37.9629, "step": 2837, "task_loss": 0.739901602268219 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34738624417162633, "compression/movement_sparsity/importance_threshold": -0.004235557825110313, "compression/movement_sparsity/linear_layer_sparsity": 0.2524180661823738, "compression/movement_sparsity/model_sparsity": 0.24374672908021003, "compression_loss": 37.325111389160156, "distillation_loss": 0.8591113090515137, "epoch": 2.4, "learning_rate": 4.667188478396994e-05, "loss": 38.2245, "step": 2838, "task_loss": 0.7088752388954163 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34802149437604146, "compression/movement_sparsity/importance_threshold": -0.004231434959249483, "compression/movement_sparsity/linear_layer_sparsity": 0.2531936498179132, "compression/movement_sparsity/model_sparsity": 0.24449566903189432, "compression_loss": 37.39289855957031, "distillation_loss": 0.7340876460075378, "epoch": 2.4, "learning_rate": 4.666875391358798e-05, "loss": 38.2346, "step": 2839, "task_loss": 0.864829421043396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34865633221422343, "compression/movement_sparsity/importance_threshold": -0.004227314769705248, "compression/movement_sparsity/linear_layer_sparsity": 0.2540610256959135, "compression/movement_sparsity/model_sparsity": 0.24533324788012503, "compression_loss": 37.46063232421875, "distillation_loss": 0.9722486734390259, "epoch": 2.4, "learning_rate": 4.666562304320601e-05, "loss": 38.2463, "step": 2840, "task_loss": 1.3590614795684814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3492907578200576, "compression/movement_sparsity/importance_threshold": -0.0042231972556086746, "compression/movement_sparsity/linear_layer_sparsity": 0.25481540816139636, "compression/movement_sparsity/model_sparsity": 0.24606171498716664, "compression_loss": 37.528350830078125, "distillation_loss": 0.8004908561706543, "epoch": 2.4, "learning_rate": 4.666249217282404e-05, "loss": 38.2403, "step": 2841, "task_loss": 0.7707799673080444 }, { "compression/movement_sparsity/importance_regularization_factor": 0.34992477132742883, "compression/movement_sparsity/importance_threshold": -0.004219082416090829, "compression/movement_sparsity/linear_layer_sparsity": 0.25570123072672934, "compression/movement_sparsity/model_sparsity": 0.2469171068222715, "compression_loss": 37.59602737426758, "distillation_loss": 1.4305315017700195, "epoch": 2.4, "learning_rate": 4.665936130244208e-05, "loss": 38.514, "step": 2842, "task_loss": 0.8881051540374756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3505583728702224, "compression/movement_sparsity/importance_threshold": -0.004214970250282777, "compression/movement_sparsity/linear_layer_sparsity": 0.25663383180169785, "compression/movement_sparsity/model_sparsity": 0.24781767018129833, "compression_loss": 37.663639068603516, "distillation_loss": 0.8491525650024414, "epoch": 2.4, "learning_rate": 4.665623043206011e-05, "loss": 38.4653, "step": 2843, "task_loss": 0.6041810512542725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3511915625823233, "compression/movement_sparsity/importance_threshold": -0.004210860757315587, "compression/movement_sparsity/linear_layer_sparsity": 0.2575059058017467, "compression/movement_sparsity/model_sparsity": 0.24865978575663206, "compression_loss": 37.731239318847656, "distillation_loss": 0.4570489823818207, "epoch": 2.4, "learning_rate": 4.665309956167815e-05, "loss": 38.6108, "step": 2844, "task_loss": 0.423117458820343 }, { "compression/movement_sparsity/importance_regularization_factor": 0.351824340597617, "compression/movement_sparsity/importance_threshold": -0.004206753936320322, "compression/movement_sparsity/linear_layer_sparsity": 0.2583820101704565, "compression/movement_sparsity/model_sparsity": 0.24950579324506436, "compression_loss": 37.79874038696289, "distillation_loss": 0.85274338722229, "epoch": 2.4, "learning_rate": 4.6649968691296183e-05, "loss": 38.5337, "step": 2845, "task_loss": 0.5056070685386658 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3524567070499883, "compression/movement_sparsity/importance_threshold": -0.0042026497864280496, "compression/movement_sparsity/linear_layer_sparsity": 0.25923371769218323, "compression/movement_sparsity/model_sparsity": 0.25032824199326104, "compression_loss": 37.86625289916992, "distillation_loss": 0.742142379283905, "epoch": 2.41, "learning_rate": 4.664683782091422e-05, "loss": 38.653, "step": 2846, "task_loss": 0.16927161812782288 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35308866207332235, "compression/movement_sparsity/importance_threshold": -0.004198548306769837, "compression/movement_sparsity/linear_layer_sparsity": 0.2601372595706231, "compression/movement_sparsity/model_sparsity": 0.25120074442855655, "compression_loss": 37.9337158203125, "distillation_loss": 0.651244044303894, "epoch": 2.41, "learning_rate": 4.6643706950532254e-05, "loss": 38.7141, "step": 2847, "task_loss": 0.5600191950798035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35372020580150454, "compression/movement_sparsity/importance_threshold": -0.004194449496476746, "compression/movement_sparsity/linear_layer_sparsity": 0.26101548644117206, "compression/movement_sparsity/model_sparsity": 0.25204880150436026, "compression_loss": 38.00112533569336, "distillation_loss": 0.8997259140014648, "epoch": 2.41, "learning_rate": 4.6640576080150286e-05, "loss": 38.8972, "step": 2848, "task_loss": 0.5887425541877747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3543513383684199, "compression/movement_sparsity/importance_threshold": -0.004190353354679847, "compression/movement_sparsity/linear_layer_sparsity": 0.26175142221932224, "compression/movement_sparsity/model_sparsity": 0.25275945562452773, "compression_loss": 38.0684814453125, "distillation_loss": 0.7795658111572266, "epoch": 2.41, "learning_rate": 4.663744520976832e-05, "loss": 38.9276, "step": 2849, "task_loss": 1.01128351688385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3549820599079536, "compression/movement_sparsity/importance_threshold": -0.004186259880510202, "compression/movement_sparsity/linear_layer_sparsity": 0.26254668073146087, "compression/movement_sparsity/model_sparsity": 0.25352739456027307, "compression_loss": 38.13583755493164, "distillation_loss": 0.8401428461074829, "epoch": 2.41, "learning_rate": 4.6634314339386356e-05, "loss": 38.9228, "step": 2850, "task_loss": 0.8236297965049744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35561237055399075, "compression/movement_sparsity/importance_threshold": -0.0041821690730988815, "compression/movement_sparsity/linear_layer_sparsity": 0.2633468520006655, "compression/movement_sparsity/model_sparsity": 0.25430007748476574, "compression_loss": 38.20309829711914, "distillation_loss": 1.179113745689392, "epoch": 2.41, "learning_rate": 4.663118346900439e-05, "loss": 38.8837, "step": 2851, "task_loss": 0.7600310444831848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35624227044041656, "compression/movement_sparsity/importance_threshold": -0.004178080931576948, "compression/movement_sparsity/linear_layer_sparsity": 0.26416518377717946, "compression/movement_sparsity/model_sparsity": 0.2550902970472735, "compression_loss": 38.27036666870117, "distillation_loss": 1.1879817247390747, "epoch": 2.41, "learning_rate": 4.662805259862242e-05, "loss": 39.0994, "step": 2852, "task_loss": 0.6705665588378906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35687175970111595, "compression/movement_sparsity/importance_threshold": -0.00417399545507547, "compression/movement_sparsity/linear_layer_sparsity": 0.2650281238648192, "compression/movement_sparsity/model_sparsity": 0.2559235924881886, "compression_loss": 38.337562561035156, "distillation_loss": 0.9156839847564697, "epoch": 2.41, "learning_rate": 4.662492172824045e-05, "loss": 39.209, "step": 2853, "task_loss": 0.7268403768539429 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35750083846997427, "compression/movement_sparsity/importance_threshold": -0.004169912642725512, "compression/movement_sparsity/linear_layer_sparsity": 0.26583273092438436, "compression/movement_sparsity/model_sparsity": 0.2567005588199968, "compression_loss": 38.40476989746094, "distillation_loss": 0.7204276323318481, "epoch": 2.41, "learning_rate": 4.662179085785849e-05, "loss": 39.1545, "step": 2854, "task_loss": 0.6732168197631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35812950688087664, "compression/movement_sparsity/importance_threshold": -0.00416583249365814, "compression/movement_sparsity/linear_layer_sparsity": 0.2665553116347824, "compression/movement_sparsity/model_sparsity": 0.2573983166600744, "compression_loss": 38.47188186645508, "distillation_loss": 0.793121337890625, "epoch": 2.41, "learning_rate": 4.661865998747652e-05, "loss": 39.3454, "step": 2855, "task_loss": 1.2277765274047852 }, { "compression/movement_sparsity/importance_regularization_factor": 0.35875776506770773, "compression/movement_sparsity/importance_threshold": -0.004161755007004424, "compression/movement_sparsity/linear_layer_sparsity": 0.2673980164099441, "compression/movement_sparsity/model_sparsity": 0.2582120719337462, "compression_loss": 38.5389518737793, "distillation_loss": 0.5183939933776855, "epoch": 2.41, "learning_rate": 4.6615529117094553e-05, "loss": 39.3446, "step": 2856, "task_loss": 0.17983664572238922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3593856131643538, "compression/movement_sparsity/importance_threshold": -0.004157680181895421, "compression/movement_sparsity/linear_layer_sparsity": 0.26820275463535326, "compression/movement_sparsity/model_sparsity": 0.25898916492544816, "compression_loss": 38.605979919433594, "distillation_loss": 1.1016957759857178, "epoch": 2.41, "learning_rate": 4.661239824671259e-05, "loss": 39.4334, "step": 2857, "task_loss": 0.8517361879348755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36001305130469874, "compression/movement_sparsity/importance_threshold": -0.004153608017462207, "compression/movement_sparsity/linear_layer_sparsity": 0.26896746343000877, "compression/movement_sparsity/model_sparsity": 0.25972760362048786, "compression_loss": 38.672969818115234, "distillation_loss": 0.542939305305481, "epoch": 2.42, "learning_rate": 4.6609267376330624e-05, "loss": 39.4956, "step": 2858, "task_loss": 0.9081098437309265 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36064007962262845, "compression/movement_sparsity/importance_threshold": -0.004149538512835842, "compression/movement_sparsity/linear_layer_sparsity": 0.26974902107353377, "compression/movement_sparsity/model_sparsity": 0.2604823123546052, "compression_loss": 38.73987579345703, "distillation_loss": 0.6673402190208435, "epoch": 2.42, "learning_rate": 4.6606136505948656e-05, "loss": 39.6256, "step": 2859, "task_loss": 0.5219282507896423 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36126669825202784, "compression/movement_sparsity/importance_threshold": -0.004145471667147394, "compression/movement_sparsity/linear_layer_sparsity": 0.27055557177242356, "compression/movement_sparsity/model_sparsity": 0.26126115555574797, "compression_loss": 38.806766510009766, "distillation_loss": 0.9192186594009399, "epoch": 2.42, "learning_rate": 4.660300563556669e-05, "loss": 39.6278, "step": 2860, "task_loss": 0.7315831184387207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36189290732678214, "compression/movement_sparsity/importance_threshold": -0.0041414074795279275, "compression/movement_sparsity/linear_layer_sparsity": 0.271282492879841, "compression/movement_sparsity/model_sparsity": 0.2619631046868548, "compression_loss": 38.873634338378906, "distillation_loss": 1.2364275455474854, "epoch": 2.42, "learning_rate": 4.6599874765184726e-05, "loss": 39.917, "step": 2861, "task_loss": 1.8810677528381348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3625187069807764, "compression/movement_sparsity/importance_threshold": -0.004137345949108511, "compression/movement_sparsity/linear_layer_sparsity": 0.2722156901631913, "compression/movement_sparsity/model_sparsity": 0.2628642437726713, "compression_loss": 38.9404296875, "distillation_loss": 1.0317611694335938, "epoch": 2.42, "learning_rate": 4.659674389480276e-05, "loss": 39.9435, "step": 2862, "task_loss": 1.2118535041809082 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36314409734789577, "compression/movement_sparsity/importance_threshold": -0.00413328707502021, "compression/movement_sparsity/linear_layer_sparsity": 0.2729521625288851, "compression/movement_sparsity/model_sparsity": 0.26357541604694956, "compression_loss": 39.007225036621094, "distillation_loss": 0.6381598114967346, "epoch": 2.42, "learning_rate": 4.659361302442079e-05, "loss": 39.7733, "step": 2863, "task_loss": 0.5040556192398071 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36376907856202545, "compression/movement_sparsity/importance_threshold": -0.004129230856394089, "compression/movement_sparsity/linear_layer_sparsity": 0.2736884918045673, "compression/movement_sparsity/model_sparsity": 0.26428645014679825, "compression_loss": 39.07393264770508, "distillation_loss": 0.5515130758285522, "epoch": 2.42, "learning_rate": 4.659048215403882e-05, "loss": 39.6584, "step": 2864, "task_loss": 0.15428057312965393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36439365075705055, "compression/movement_sparsity/importance_threshold": -0.004125177292361214, "compression/movement_sparsity/linear_layer_sparsity": 0.27456634902591953, "compression/movement_sparsity/model_sparsity": 0.26513415027199233, "compression_loss": 39.14055633544922, "distillation_loss": 0.6442832350730896, "epoch": 2.42, "learning_rate": 4.658735128365686e-05, "loss": 39.8977, "step": 2865, "task_loss": 0.5284631252288818 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36501781406685607, "compression/movement_sparsity/importance_threshold": -0.004121126382052654, "compression/movement_sparsity/linear_layer_sparsity": 0.2754894465392822, "compression/movement_sparsity/model_sparsity": 0.2660255365459909, "compression_loss": 39.2071647644043, "distillation_loss": 0.7689279317855835, "epoch": 2.42, "learning_rate": 4.658422041327489e-05, "loss": 39.9215, "step": 2866, "task_loss": 1.6232861280441284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36564156862532715, "compression/movement_sparsity/importance_threshold": -0.004117078124599473, "compression/movement_sparsity/linear_layer_sparsity": 0.2762450929665345, "compression/movement_sparsity/model_sparsity": 0.26675522419382675, "compression_loss": 39.273681640625, "distillation_loss": 0.21237990260124207, "epoch": 2.42, "learning_rate": 4.658108954289292e-05, "loss": 39.9412, "step": 2867, "task_loss": 0.06179613992571831 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36626491456634935, "compression/movement_sparsity/importance_threshold": -0.004113032519132736, "compression/movement_sparsity/linear_layer_sparsity": 0.2771154975831143, "compression/movement_sparsity/model_sparsity": 0.26759572773414925, "compression_loss": 39.340171813964844, "distillation_loss": 0.8785097599029541, "epoch": 2.42, "learning_rate": 4.657795867251096e-05, "loss": 40.143, "step": 2868, "task_loss": 0.35905545949935913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36688785202380747, "compression/movement_sparsity/importance_threshold": -0.00410898956478351, "compression/movement_sparsity/linear_layer_sparsity": 0.2781108913248532, "compression/movement_sparsity/model_sparsity": 0.26855692663867015, "compression_loss": 39.40658950805664, "distillation_loss": 0.6887563467025757, "epoch": 2.42, "learning_rate": 4.6574827802128994e-05, "loss": 40.206, "step": 2869, "task_loss": 0.4455450475215912 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3675103811315866, "compression/movement_sparsity/importance_threshold": -0.004104949260682861, "compression/movement_sparsity/linear_layer_sparsity": 0.27897536963011804, "compression/movement_sparsity/model_sparsity": 0.26939170745470276, "compression_loss": 39.472991943359375, "distillation_loss": 0.6854485273361206, "epoch": 2.43, "learning_rate": 4.6571696931747025e-05, "loss": 40.2766, "step": 2870, "task_loss": 0.3987427353858948 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3681325020235722, "compression/movement_sparsity/importance_threshold": -0.004100911605961856, "compression/movement_sparsity/linear_layer_sparsity": 0.2798457623225302, "compression/movement_sparsity/model_sparsity": 0.2702321994804895, "compression_loss": 39.539344787597656, "distillation_loss": 0.8364598751068115, "epoch": 2.43, "learning_rate": 4.656856606136506e-05, "loss": 40.6591, "step": 2871, "task_loss": 0.3235260844230652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.368754214833649, "compression/movement_sparsity/importance_threshold": -0.0040968765997515585, "compression/movement_sparsity/linear_layer_sparsity": 0.28072531277568674, "compression/movement_sparsity/model_sparsity": 0.2710815346697664, "compression_loss": 39.605628967285156, "distillation_loss": 0.7317798733711243, "epoch": 2.43, "learning_rate": 4.6565435190983096e-05, "loss": 40.3236, "step": 2872, "task_loss": 0.5763554573059082 }, { "compression/movement_sparsity/importance_regularization_factor": 0.36937551969570237, "compression/movement_sparsity/importance_threshold": -0.004092844241183038, "compression/movement_sparsity/linear_layer_sparsity": 0.2814827478280844, "compression/movement_sparsity/model_sparsity": 0.2718129494979714, "compression_loss": 39.67190933227539, "distillation_loss": 0.6446233987808228, "epoch": 2.43, "learning_rate": 4.656230432060113e-05, "loss": 40.3812, "step": 2873, "task_loss": 0.3211532235145569 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3699964167436175, "compression/movement_sparsity/importance_threshold": -0.004088814529387358, "compression/movement_sparsity/linear_layer_sparsity": 0.2823440066080875, "compression/movement_sparsity/model_sparsity": 0.27264462138933954, "compression_loss": 39.73811721801758, "distillation_loss": 0.6134083271026611, "epoch": 2.43, "learning_rate": 4.655917345021916e-05, "loss": 40.4381, "step": 2874, "task_loss": 0.6651320457458496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3706169061112794, "compression/movement_sparsity/importance_threshold": -0.004084787463495585, "compression/movement_sparsity/linear_layer_sparsity": 0.2832735789444765, "compression/movement_sparsity/model_sparsity": 0.2735422600562745, "compression_loss": 39.80424499511719, "distillation_loss": 1.0629452466964722, "epoch": 2.43, "learning_rate": 4.65560425798372e-05, "loss": 40.7133, "step": 2875, "task_loss": 1.7438074350357056 }, { "compression/movement_sparsity/importance_regularization_factor": 0.371236987932573, "compression/movement_sparsity/importance_threshold": -0.004080763042638788, "compression/movement_sparsity/linear_layer_sparsity": 0.2841550134161195, "compression/movement_sparsity/model_sparsity": 0.274393414542207, "compression_loss": 39.870384216308594, "distillation_loss": 1.0434119701385498, "epoch": 2.43, "learning_rate": 4.655291170945523e-05, "loss": 40.7595, "step": 2876, "task_loss": 1.18073308467865 }, { "compression/movement_sparsity/importance_regularization_factor": 0.37185666234138415, "compression/movement_sparsity/importance_threshold": -0.004076741265948026, "compression/movement_sparsity/linear_layer_sparsity": 0.28501499631018556, "compression/movement_sparsity/model_sparsity": 0.27522385437824504, "compression_loss": 39.93647766113281, "distillation_loss": 1.1186249256134033, "epoch": 2.43, "learning_rate": 4.654978083907326e-05, "loss": 40.7084, "step": 2877, "task_loss": 0.698478639125824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.37247592947159713, "compression/movement_sparsity/importance_threshold": -0.004072722132554373, "compression/movement_sparsity/linear_layer_sparsity": 0.28585144089733844, "compression/movement_sparsity/model_sparsity": 0.2760315645206247, "compression_loss": 40.00252151489258, "distillation_loss": 0.7777044773101807, "epoch": 2.43, "learning_rate": 4.65466499686913e-05, "loss": 40.771, "step": 2878, "task_loss": 0.6971895694732666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3730947894570976, "compression/movement_sparsity/importance_threshold": -0.00406870564158889, "compression/movement_sparsity/linear_layer_sparsity": 0.28660719464209944, "compression/movement_sparsity/model_sparsity": 0.27676135579928274, "compression_loss": 40.06852340698242, "distillation_loss": 0.9045848250389099, "epoch": 2.43, "learning_rate": 4.654351909830933e-05, "loss": 40.7685, "step": 2879, "task_loss": 0.4576095938682556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.37371324243177073, "compression/movement_sparsity/importance_threshold": -0.004064691792182645, "compression/movement_sparsity/linear_layer_sparsity": 0.28753692199266767, "compression/movement_sparsity/model_sparsity": 0.27765914415518306, "compression_loss": 40.13444900512695, "distillation_loss": 1.1698685884475708, "epoch": 2.43, "learning_rate": 4.654038822792737e-05, "loss": 40.9807, "step": 2880, "task_loss": 0.7804399132728577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3743312885295015, "compression/movement_sparsity/importance_threshold": -0.004060680583466703, "compression/movement_sparsity/linear_layer_sparsity": 0.28831608287849786, "compression/movement_sparsity/model_sparsity": 0.27841153846760575, "compression_loss": 40.2003059387207, "distillation_loss": 0.5959482789039612, "epoch": 2.44, "learning_rate": 4.65372573575454e-05, "loss": 40.8497, "step": 2881, "task_loss": 0.23670023679733276 }, { "compression/movement_sparsity/importance_regularization_factor": 0.374948927884175, "compression/movement_sparsity/importance_threshold": -0.004056672014572131, "compression/movement_sparsity/linear_layer_sparsity": 0.2892685019200772, "compression/movement_sparsity/model_sparsity": 0.2793312389851231, "compression_loss": 40.26615905761719, "distillation_loss": 1.637770414352417, "epoch": 2.44, "learning_rate": 4.6534126487163434e-05, "loss": 41.1578, "step": 2882, "task_loss": 1.3308818340301514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3755661606296764, "compression/movement_sparsity/importance_threshold": -0.004052666084629995, "compression/movement_sparsity/linear_layer_sparsity": 0.2900410568170371, "compression/movement_sparsity/model_sparsity": 0.2800772542447156, "compression_loss": 40.331947326660156, "distillation_loss": 0.5301905274391174, "epoch": 2.44, "learning_rate": 4.653099561678147e-05, "loss": 41.0565, "step": 2883, "task_loss": 0.35828590393066406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.37618298689989105, "compression/movement_sparsity/importance_threshold": -0.004048662792771359, "compression/movement_sparsity/linear_layer_sparsity": 0.2908367088267077, "compression/movement_sparsity/model_sparsity": 0.28084557316014214, "compression_loss": 40.39767837524414, "distillation_loss": 0.6095954179763794, "epoch": 2.44, "learning_rate": 4.6527864746399504e-05, "loss": 40.9782, "step": 2884, "task_loss": 0.22338418662548065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.37679940682870383, "compression/movement_sparsity/importance_threshold": -0.004044662138127292, "compression/movement_sparsity/linear_layer_sparsity": 0.29176351475620516, "compression/movement_sparsity/model_sparsity": 0.2817405404547728, "compression_loss": 40.46339416503906, "distillation_loss": 1.1423096656799316, "epoch": 2.44, "learning_rate": 4.6524733876017536e-05, "loss": 41.3661, "step": 2885, "task_loss": 1.4694613218307495 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3774154205499999, "compression/movement_sparsity/importance_threshold": -0.00404066411982886, "compression/movement_sparsity/linear_layer_sparsity": 0.2927893747462537, "compression/movement_sparsity/model_sparsity": 0.28273115899824874, "compression_loss": 40.529052734375, "distillation_loss": 0.8341879844665527, "epoch": 2.44, "learning_rate": 4.652160300563557e-05, "loss": 41.6401, "step": 2886, "task_loss": 1.1159257888793945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3780310281976643, "compression/movement_sparsity/importance_threshold": -0.004036668737007127, "compression/movement_sparsity/linear_layer_sparsity": 0.29372080725279387, "compression/movement_sparsity/model_sparsity": 0.2836305939327677, "compression_loss": 40.59469223022461, "distillation_loss": 0.25385504961013794, "epoch": 2.44, "learning_rate": 4.6518472135253606e-05, "loss": 41.2741, "step": 2887, "task_loss": 0.2923908829689026 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3786462299055826, "compression/movement_sparsity/importance_threshold": -0.00403267598879316, "compression/movement_sparsity/linear_layer_sparsity": 0.2945209427494956, "compression/movement_sparsity/model_sparsity": 0.284403242313653, "compression_loss": 40.66029357910156, "distillation_loss": 1.0036131143569946, "epoch": 2.44, "learning_rate": 4.651534126487164e-05, "loss": 41.5371, "step": 2888, "task_loss": 0.8333678245544434 }, { "compression/movement_sparsity/importance_regularization_factor": 0.37926102580763965, "compression/movement_sparsity/importance_threshold": -0.004028685874318023, "compression/movement_sparsity/linear_layer_sparsity": 0.2954278233948735, "compression/movement_sparsity/model_sparsity": 0.28527896881897097, "compression_loss": 40.72584533691406, "distillation_loss": 0.6177372932434082, "epoch": 2.44, "learning_rate": 4.651221039448967e-05, "loss": 41.5322, "step": 2889, "task_loss": 0.8472769856452942 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3798754160377207, "compression/movement_sparsity/importance_threshold": -0.004024698392712785, "compression/movement_sparsity/linear_layer_sparsity": 0.296301447536715, "compression/movement_sparsity/model_sparsity": 0.286122581283958, "compression_loss": 40.79136657714844, "distillation_loss": 0.4667612314224243, "epoch": 2.44, "learning_rate": 4.65090795241077e-05, "loss": 41.693, "step": 2890, "task_loss": 0.52254319190979 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3804894007297107, "compression/movement_sparsity/importance_threshold": -0.004020713543108512, "compression/movement_sparsity/linear_layer_sparsity": 0.29715943909478587, "compression/movement_sparsity/model_sparsity": 0.2869510981925184, "compression_loss": 40.856842041015625, "distillation_loss": 0.6757433414459229, "epoch": 2.44, "learning_rate": 4.650594865372574e-05, "loss": 41.6772, "step": 2891, "task_loss": 0.2354496568441391 }, { "compression/movement_sparsity/importance_regularization_factor": 0.381102980017495, "compression/movement_sparsity/importance_threshold": -0.0040167313246362675, "compression/movement_sparsity/linear_layer_sparsity": 0.29789081984089916, "compression/movement_sparsity/model_sparsity": 0.2876573537600124, "compression_loss": 40.92229080200195, "distillation_loss": 0.5831711292266846, "epoch": 2.44, "learning_rate": 4.650281778334377e-05, "loss": 41.8361, "step": 2892, "task_loss": 0.2888459265232086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3817161540349585, "compression/movement_sparsity/importance_threshold": -0.00401275173642712, "compression/movement_sparsity/linear_layer_sparsity": 0.2987747941602486, "compression/movement_sparsity/model_sparsity": 0.28851096084206906, "compression_loss": 40.987693786621094, "distillation_loss": 0.5599726438522339, "epoch": 2.45, "learning_rate": 4.6499686912961804e-05, "loss": 41.7619, "step": 2893, "task_loss": 0.5428087115287781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.38232892291598664, "compression/movement_sparsity/importance_threshold": -0.004008774777612134, "compression/movement_sparsity/linear_layer_sparsity": 0.29967263628655855, "compression/movement_sparsity/model_sparsity": 0.2893779593292548, "compression_loss": 41.05303192138672, "distillation_loss": 0.6351958513259888, "epoch": 2.45, "learning_rate": 4.649655604257984e-05, "loss": 41.676, "step": 2894, "task_loss": 1.06733238697052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3829412867944644, "compression/movement_sparsity/importance_threshold": -0.004004800447322376, "compression/movement_sparsity/linear_layer_sparsity": 0.3005325237872835, "compression/movement_sparsity/model_sparsity": 0.2902083070490065, "compression_loss": 41.11833190917969, "distillation_loss": 1.173446774482727, "epoch": 2.45, "learning_rate": 4.6493425172197874e-05, "loss": 41.9153, "step": 2895, "task_loss": 0.44803524017333984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.38355324580427663, "compression/movement_sparsity/importance_threshold": -0.004000828744688914, "compression/movement_sparsity/linear_layer_sparsity": 0.3014793623184092, "compression/movement_sparsity/model_sparsity": 0.29112261876377205, "compression_loss": 41.18359375, "distillation_loss": 0.7239701747894287, "epoch": 2.45, "learning_rate": 4.6490294301815906e-05, "loss": 42.1057, "step": 2896, "task_loss": 0.4284662902355194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3841648000793093, "compression/movement_sparsity/importance_threshold": -0.0039968596688428094, "compression/movement_sparsity/linear_layer_sparsity": 0.3024134658384998, "compression/movement_sparsity/model_sparsity": 0.29202463295430897, "compression_loss": 41.248756408691406, "distillation_loss": 0.702984094619751, "epoch": 2.45, "learning_rate": 4.648716343143394e-05, "loss": 42.1409, "step": 2897, "task_loss": 1.0311353206634521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3847759497534464, "compression/movement_sparsity/importance_threshold": -0.003992893218915134, "compression/movement_sparsity/linear_layer_sparsity": 0.30326291969254343, "compression/movement_sparsity/model_sparsity": 0.29284490545524045, "compression_loss": 41.313968658447266, "distillation_loss": 0.8392283916473389, "epoch": 2.45, "learning_rate": 4.6484032561051976e-05, "loss": 42.1862, "step": 2898, "task_loss": 0.767379105091095 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3853866949605741, "compression/movement_sparsity/importance_threshold": -0.003988929394036949, "compression/movement_sparsity/linear_layer_sparsity": 0.3040476133921567, "compression/movement_sparsity/model_sparsity": 0.2936026425122718, "compression_loss": 41.37907028198242, "distillation_loss": 1.2725660800933838, "epoch": 2.45, "learning_rate": 4.648090169067001e-05, "loss": 42.359, "step": 2899, "task_loss": 0.3382527828216553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.385997035834577, "compression/movement_sparsity/importance_threshold": -0.003984968193339324, "compression/movement_sparsity/linear_layer_sparsity": 0.30498456678831226, "compression/movement_sparsity/model_sparsity": 0.29450740867686365, "compression_loss": 41.44411849975586, "distillation_loss": 1.0480101108551025, "epoch": 2.45, "learning_rate": 4.647777082028804e-05, "loss": 42.5172, "step": 2900, "task_loss": 1.4711451530456543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3866069725093405, "compression/movement_sparsity/importance_threshold": -0.003981009615953321, "compression/movement_sparsity/linear_layer_sparsity": 0.3059421966911484, "compression/movement_sparsity/model_sparsity": 0.2954321410465232, "compression_loss": 41.50917053222656, "distillation_loss": 0.6651859283447266, "epoch": 2.45, "learning_rate": 4.647463994990607e-05, "loss": 42.4328, "step": 2901, "task_loss": 0.9531567096710205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3872165051187495, "compression/movement_sparsity/importance_threshold": -0.00397705366101001, "compression/movement_sparsity/linear_layer_sparsity": 0.30678335132451745, "compression/movement_sparsity/model_sparsity": 0.2962443994305417, "compression_loss": 41.57416534423828, "distillation_loss": 1.171783685684204, "epoch": 2.45, "learning_rate": 4.647150907952411e-05, "loss": 42.4145, "step": 2902, "task_loss": 0.9167967438697815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.38782563379668933, "compression/movement_sparsity/importance_threshold": -0.003973100327640455, "compression/movement_sparsity/linear_layer_sparsity": 0.3077512717840234, "compression/movement_sparsity/model_sparsity": 0.297179068844592, "compression_loss": 41.63910675048828, "distillation_loss": 0.795266330242157, "epoch": 2.45, "learning_rate": 4.646837820914214e-05, "loss": 42.6264, "step": 2903, "task_loss": 0.8102775812149048 }, { "compression/movement_sparsity/importance_regularization_factor": 0.388434358677045, "compression/movement_sparsity/importance_threshold": -0.003969149614975721, "compression/movement_sparsity/linear_layer_sparsity": 0.30864001577042716, "compression/movement_sparsity/model_sparsity": 0.2980372817409665, "compression_loss": 41.70398712158203, "distillation_loss": 1.0853543281555176, "epoch": 2.45, "learning_rate": 4.6465247338760174e-05, "loss": 42.7023, "step": 2904, "task_loss": 1.7034764289855957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3890426798937018, "compression/movement_sparsity/importance_threshold": -0.003965201522146878, "compression/movement_sparsity/linear_layer_sparsity": 0.3096413238993135, "compression/movement_sparsity/model_sparsity": 0.29900419185524146, "compression_loss": 41.768836975097656, "distillation_loss": 0.7250006198883057, "epoch": 2.46, "learning_rate": 4.646211646837821e-05, "loss": 42.3868, "step": 2905, "task_loss": 1.0855621099472046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.38965059758054466, "compression/movement_sparsity/importance_threshold": -0.003961256048284989, "compression/movement_sparsity/linear_layer_sparsity": 0.3105017718359174, "compression/movement_sparsity/model_sparsity": 0.29983508075817555, "compression_loss": 41.83367919921875, "distillation_loss": 0.7918041944503784, "epoch": 2.46, "learning_rate": 4.6458985597996244e-05, "loss": 42.7318, "step": 2906, "task_loss": 0.3014744222164154 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39025811187145865, "compression/movement_sparsity/importance_threshold": -0.003957313192521121, "compression/movement_sparsity/linear_layer_sparsity": 0.31125503342964245, "compression/movement_sparsity/model_sparsity": 0.30056246549885246, "compression_loss": 41.89848709106445, "distillation_loss": 1.182638168334961, "epoch": 2.46, "learning_rate": 4.6455854727614276e-05, "loss": 43.364, "step": 2907, "task_loss": 1.2968885898590088 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3908652229003293, "compression/movement_sparsity/importance_threshold": -0.003953372953986338, "compression/movement_sparsity/linear_layer_sparsity": 0.31199124346364826, "compression/movement_sparsity/model_sparsity": 0.30127338445334323, "compression_loss": 41.96324157714844, "distillation_loss": 0.6459425091743469, "epoch": 2.46, "learning_rate": 4.645272385723231e-05, "loss": 42.9923, "step": 2908, "task_loss": 1.2578368186950684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3914719308010416, "compression/movement_sparsity/importance_threshold": -0.003949435331811709, "compression/movement_sparsity/linear_layer_sparsity": 0.31293485054534464, "compression/movement_sparsity/model_sparsity": 0.3021845757289084, "compression_loss": 42.02803421020508, "distillation_loss": 1.1089534759521484, "epoch": 2.46, "learning_rate": 4.6449592986850346e-05, "loss": 43.0783, "step": 2909, "task_loss": 1.5430448055267334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39207823570748057, "compression/movement_sparsity/importance_threshold": -0.003945500325128299, "compression/movement_sparsity/linear_layer_sparsity": 0.31373329281024204, "compression/movement_sparsity/model_sparsity": 0.30295558904571085, "compression_loss": 42.09272384643555, "distillation_loss": 1.2307748794555664, "epoch": 2.46, "learning_rate": 4.644646211646838e-05, "loss": 43.1943, "step": 2910, "task_loss": 1.0075352191925049 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3926841377535315, "compression/movement_sparsity/importance_threshold": -0.003941567933067172, "compression/movement_sparsity/linear_layer_sparsity": 0.31449025089593424, "compression/movement_sparsity/model_sparsity": 0.3036865432924841, "compression_loss": 42.15739440917969, "distillation_loss": 1.1939666271209717, "epoch": 2.46, "learning_rate": 4.6443331246086417e-05, "loss": 43.1837, "step": 2911, "task_loss": 1.069510579109192 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3932896370730794, "compression/movement_sparsity/importance_threshold": -0.0039376381547593965, "compression/movement_sparsity/linear_layer_sparsity": 0.31525554397480393, "compression/movement_sparsity/model_sparsity": 0.30442554619977774, "compression_loss": 42.222023010253906, "distillation_loss": 0.45554298162460327, "epoch": 2.46, "learning_rate": 4.644020037570445e-05, "loss": 43.282, "step": 2912, "task_loss": 0.21734756231307983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3938947338000094, "compression/movement_sparsity/importance_threshold": -0.003933710989336038, "compression/movement_sparsity/linear_layer_sparsity": 0.31599447271903075, "compression/movement_sparsity/model_sparsity": 0.30513909046842963, "compression_loss": 42.28664779663086, "distillation_loss": 1.2905585765838623, "epoch": 2.46, "learning_rate": 4.643706950532249e-05, "loss": 43.2745, "step": 2913, "task_loss": 0.46929341554641724 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3944994280682067, "compression/movement_sparsity/importance_threshold": -0.003929786435928163, "compression/movement_sparsity/linear_layer_sparsity": 0.31671168755399265, "compression/movement_sparsity/model_sparsity": 0.30583166676739965, "compression_loss": 42.35117721557617, "distillation_loss": 0.6128755211830139, "epoch": 2.46, "learning_rate": 4.643393863494052e-05, "loss": 43.4111, "step": 2914, "task_loss": 0.7843023538589478 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3951037200115566, "compression/movement_sparsity/importance_threshold": -0.003925864493666836, "compression/movement_sparsity/linear_layer_sparsity": 0.3175697148845664, "compression/movement_sparsity/model_sparsity": 0.30666021821956746, "compression_loss": 42.41565704345703, "distillation_loss": 0.6931375861167908, "epoch": 2.46, "learning_rate": 4.643080776455855e-05, "loss": 43.177, "step": 2915, "task_loss": 0.6785308122634888 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3957076097639439, "compression/movement_sparsity/importance_threshold": -0.003921945161683125, "compression/movement_sparsity/linear_layer_sparsity": 0.31832735264781387, "compression/movement_sparsity/model_sparsity": 0.307391828794881, "compression_loss": 42.4801139831543, "distillation_loss": 0.9215804934501648, "epoch": 2.46, "learning_rate": 4.642767689417659e-05, "loss": 43.575, "step": 2916, "task_loss": 2.18851375579834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39631109745925397, "compression/movement_sparsity/importance_threshold": -0.003918028439108094, "compression/movement_sparsity/linear_layer_sparsity": 0.3191698904846287, "compression/movement_sparsity/model_sparsity": 0.30820542286505165, "compression_loss": 42.5445442199707, "distillation_loss": 0.6391297578811646, "epoch": 2.47, "learning_rate": 4.642454602379462e-05, "loss": 43.5501, "step": 2917, "task_loss": 0.7497884035110474 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39691418323137173, "compression/movement_sparsity/importance_threshold": -0.003914114325072812, "compression/movement_sparsity/linear_layer_sparsity": 0.3198048524112384, "compression/movement_sparsity/model_sparsity": 0.30881857189611095, "compression_loss": 42.60887908935547, "distillation_loss": 1.3246917724609375, "epoch": 2.47, "learning_rate": 4.642141515341265e-05, "loss": 43.6203, "step": 2918, "task_loss": 0.418562114238739 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39751686721418267, "compression/movement_sparsity/importance_threshold": -0.003910202818708342, "compression/movement_sparsity/linear_layer_sparsity": 0.3205441865771648, "compression/movement_sparsity/model_sparsity": 0.3095325076589798, "compression_loss": 42.6732177734375, "distillation_loss": 1.0891743898391724, "epoch": 2.47, "learning_rate": 4.6418284283030684e-05, "loss": 43.8306, "step": 2919, "task_loss": 0.5415326952934265 }, { "compression/movement_sparsity/importance_regularization_factor": 0.3981191495415718, "compression/movement_sparsity/importance_threshold": -0.003906293919145751, "compression/movement_sparsity/linear_layer_sparsity": 0.32129719776336957, "compression/movement_sparsity/model_sparsity": 0.3102596505944051, "compression_loss": 42.73749542236328, "distillation_loss": 1.188812494277954, "epoch": 2.47, "learning_rate": 4.641515341264872e-05, "loss": 43.5605, "step": 2920, "task_loss": 0.28355827927589417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39872103034742423, "compression/movement_sparsity/importance_threshold": -0.0039023876255161044, "compression/movement_sparsity/linear_layer_sparsity": 0.32225288402688107, "compression/movement_sparsity/model_sparsity": 0.31118250609473014, "compression_loss": 42.80171203613281, "distillation_loss": 0.5417231917381287, "epoch": 2.47, "learning_rate": 4.6412022542266755e-05, "loss": 43.7118, "step": 2921, "task_loss": 0.6326702833175659 }, { "compression/movement_sparsity/importance_regularization_factor": 0.399322509765625, "compression/movement_sparsity/importance_threshold": -0.00389848393695047, "compression/movement_sparsity/linear_layer_sparsity": 0.3231616129182426, "compression/movement_sparsity/model_sparsity": 0.3120600173530963, "compression_loss": 42.865882873535156, "distillation_loss": 1.4310270547866821, "epoch": 2.47, "learning_rate": 4.6408891671884786e-05, "loss": 44.0238, "step": 2922, "task_loss": 1.1885594129562378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.39992358793005933, "compression/movement_sparsity/importance_threshold": -0.0038945828525799127, "compression/movement_sparsity/linear_layer_sparsity": 0.3239664226886575, "compression/movement_sparsity/model_sparsity": 0.31283717943201306, "compression_loss": 42.93000793457031, "distillation_loss": 0.8684248328208923, "epoch": 2.47, "learning_rate": 4.640576080150282e-05, "loss": 43.7433, "step": 2923, "task_loss": 0.7998301982879639 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40052426497461247, "compression/movement_sparsity/importance_threshold": -0.003890684371535498, "compression/movement_sparsity/linear_layer_sparsity": 0.32469450044033565, "compression/movement_sparsity/model_sparsity": 0.3135402454730919, "compression_loss": 42.99408721923828, "distillation_loss": 0.83378005027771, "epoch": 2.47, "learning_rate": 4.640262993112086e-05, "loss": 43.7778, "step": 2924, "task_loss": 1.6361151933670044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4011245410331694, "compression/movement_sparsity/importance_threshold": -0.003886788492948293, "compression/movement_sparsity/linear_layer_sparsity": 0.325454630354619, "compression/movement_sparsity/model_sparsity": 0.3142742625863865, "compression_loss": 43.05809783935547, "distillation_loss": 0.976805567741394, "epoch": 2.47, "learning_rate": 4.639949906073889e-05, "loss": 43.9218, "step": 2925, "task_loss": 0.4539892077445984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4017244162396153, "compression/movement_sparsity/importance_threshold": -0.0038828952159493633, "compression/movement_sparsity/linear_layer_sparsity": 0.32635516734281467, "compression/movement_sparsity/model_sparsity": 0.31514386335866185, "compression_loss": 43.12208557128906, "distillation_loss": 0.8063911199569702, "epoch": 2.47, "learning_rate": 4.639636819035692e-05, "loss": 43.9198, "step": 2926, "task_loss": 1.1421538591384888 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40232389072783514, "compression/movement_sparsity/importance_threshold": -0.0038790045396697764, "compression/movement_sparsity/linear_layer_sparsity": 0.3271673581729962, "compression/movement_sparsity/model_sparsity": 0.31592815293523535, "compression_loss": 43.18598175048828, "distillation_loss": 1.0425676107406616, "epoch": 2.47, "learning_rate": 4.639323731997495e-05, "loss": 44.0352, "step": 2927, "task_loss": 1.0535584688186646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4029229646317144, "compression/movement_sparsity/importance_threshold": -0.003875116463240596, "compression/movement_sparsity/linear_layer_sparsity": 0.3281826891230223, "compression/movement_sparsity/model_sparsity": 0.3169086041436047, "compression_loss": 43.249847412109375, "distillation_loss": 0.8331427574157715, "epoch": 2.47, "learning_rate": 4.639010644959299e-05, "loss": 44.0631, "step": 2928, "task_loss": 0.5636480450630188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4035216380851381, "compression/movement_sparsity/importance_threshold": -0.003871230985792889, "compression/movement_sparsity/linear_layer_sparsity": 0.3289308352488316, "compression/movement_sparsity/model_sparsity": 0.3176310491484258, "compression_loss": 43.313655853271484, "distillation_loss": 0.6925618648529053, "epoch": 2.48, "learning_rate": 4.638697557921102e-05, "loss": 44.3196, "step": 2929, "task_loss": 0.7738455533981323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4041199112219914, "compression/movement_sparsity/importance_threshold": -0.003867348106457721, "compression/movement_sparsity/linear_layer_sparsity": 0.3299261812939, "compression/movement_sparsity/model_sparsity": 0.31859220199480354, "compression_loss": 43.37746810913086, "distillation_loss": 0.34990906715393066, "epoch": 2.48, "learning_rate": 4.6383844708829054e-05, "loss": 44.0114, "step": 2930, "task_loss": 0.8012046813964844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40471778417615933, "compression/movement_sparsity/importance_threshold": -0.0038634678243661593, "compression/movement_sparsity/linear_layer_sparsity": 0.33078704657637115, "compression/movement_sparsity/model_sparsity": 0.3194234939064904, "compression_loss": 43.44115447998047, "distillation_loss": 0.9563947319984436, "epoch": 2.48, "learning_rate": 4.638071383844709e-05, "loss": 44.4659, "step": 2931, "task_loss": 0.6120999455451965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40531525708152716, "compression/movement_sparsity/importance_threshold": -0.0038595901386492683, "compression/movement_sparsity/linear_layer_sparsity": 0.33169105349734884, "compression/movement_sparsity/model_sparsity": 0.3202964454086819, "compression_loss": 43.50485610961914, "distillation_loss": 0.7025855779647827, "epoch": 2.48, "learning_rate": 4.6377582968065125e-05, "loss": 44.4283, "step": 2932, "task_loss": 1.3807244300842285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4059123300719799, "compression/movement_sparsity/importance_threshold": -0.0038557150484381155, "compression/movement_sparsity/linear_layer_sparsity": 0.3326149022332726, "compression/movement_sparsity/model_sparsity": 0.3211885570984355, "compression_loss": 43.568511962890625, "distillation_loss": 0.6479580402374268, "epoch": 2.48, "learning_rate": 4.6374452097683156e-05, "loss": 44.3871, "step": 2933, "task_loss": 0.6041141152381897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4065090032814027, "compression/movement_sparsity/importance_threshold": -0.003851842552863767, "compression/movement_sparsity/linear_layer_sparsity": 0.33337214649898805, "compression/movement_sparsity/model_sparsity": 0.32191978769406787, "compression_loss": 43.63206481933594, "distillation_loss": 1.1242129802703857, "epoch": 2.48, "learning_rate": 4.637132122730119e-05, "loss": 44.7485, "step": 2934, "task_loss": 0.6753441095352173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4071052768436808, "compression/movement_sparsity/importance_threshold": -0.003847972651057288, "compression/movement_sparsity/linear_layer_sparsity": 0.3342026886231612, "compression/movement_sparsity/model_sparsity": 0.3227217981412292, "compression_loss": 43.69563293457031, "distillation_loss": 0.6448209285736084, "epoch": 2.48, "learning_rate": 4.636819035691923e-05, "loss": 44.4336, "step": 2935, "task_loss": 0.06907905638217926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4077011508926993, "compression/movement_sparsity/importance_threshold": -0.0038441053421497443, "compression/movement_sparsity/linear_layer_sparsity": 0.3351139812105644, "compression/movement_sparsity/model_sparsity": 0.3236017850247912, "compression_loss": 43.7591438293457, "distillation_loss": 0.8677753210067749, "epoch": 2.48, "learning_rate": 4.636505948653726e-05, "loss": 44.5361, "step": 2936, "task_loss": 0.6613269448280334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40829662556234336, "compression/movement_sparsity/importance_threshold": -0.0038402406252722025, "compression/movement_sparsity/linear_layer_sparsity": 0.3359567456065643, "compression/movement_sparsity/model_sparsity": 0.32441559787114194, "compression_loss": 43.822532653808594, "distillation_loss": 0.9101265072822571, "epoch": 2.48, "learning_rate": 4.636192861615529e-05, "loss": 44.8325, "step": 2937, "task_loss": 0.4901188313961029 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40889170098649785, "compression/movement_sparsity/importance_threshold": -0.00383637849955573, "compression/movement_sparsity/linear_layer_sparsity": 0.33683933672246796, "compression/movement_sparsity/model_sparsity": 0.3252678692670465, "compression_loss": 43.88595962524414, "distillation_loss": 0.7091403603553772, "epoch": 2.48, "learning_rate": 4.635879774577332e-05, "loss": 44.7436, "step": 2938, "task_loss": 0.1619638353586197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.40948637729904847, "compression/movement_sparsity/importance_threshold": -0.0038325189641313898, "compression/movement_sparsity/linear_layer_sparsity": 0.33783306108073785, "compression/movement_sparsity/model_sparsity": 0.3262274561365561, "compression_loss": 43.949317932128906, "distillation_loss": 1.4223999977111816, "epoch": 2.48, "learning_rate": 4.635566687539136e-05, "loss": 45.2192, "step": 2939, "task_loss": 1.7590564489364624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4100806546338799, "compression/movement_sparsity/importance_threshold": -0.0038286620181302504, "compression/movement_sparsity/linear_layer_sparsity": 0.3386267575269162, "compression/movement_sparsity/model_sparsity": 0.32699388666811235, "compression_loss": 44.01266860961914, "distillation_loss": 1.1578500270843506, "epoch": 2.48, "learning_rate": 4.635253600500939e-05, "loss": 45.2843, "step": 2940, "task_loss": 1.1017003059387207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41067453312487745, "compression/movement_sparsity/importance_threshold": -0.0038248076606833764, "compression/movement_sparsity/linear_layer_sparsity": 0.3394116897098821, "compression/movement_sparsity/model_sparsity": 0.3277518540158596, "compression_loss": 44.07598876953125, "distillation_loss": 0.5856345295906067, "epoch": 2.49, "learning_rate": 4.6349405134627424e-05, "loss": 45.0868, "step": 2941, "task_loss": 0.9561784863471985 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4112680129059263, "compression/movement_sparsity/importance_threshold": -0.003820955890921834, "compression/movement_sparsity/linear_layer_sparsity": 0.3403250809507892, "compression/movement_sparsity/model_sparsity": 0.32863386745772144, "compression_loss": 44.13926315307617, "distillation_loss": 1.029386281967163, "epoch": 2.49, "learning_rate": 4.634627426424546e-05, "loss": 45.0443, "step": 2942, "task_loss": 1.0583741664886475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41186109411091154, "compression/movement_sparsity/importance_threshold": -0.0038171067079766897, "compression/movement_sparsity/linear_layer_sparsity": 0.3410792845537576, "compression/movement_sparsity/model_sparsity": 0.3293621618467261, "compression_loss": 44.20250701904297, "distillation_loss": 1.4305604696273804, "epoch": 2.49, "learning_rate": 4.6343143393863495e-05, "loss": 45.2288, "step": 2943, "task_loss": 0.9303393363952637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4124537768737182, "compression/movement_sparsity/importance_threshold": -0.0038132601109790104, "compression/movement_sparsity/linear_layer_sparsity": 0.3419096597395838, "compression/movement_sparsity/model_sparsity": 0.33016401109038634, "compression_loss": 44.26570129394531, "distillation_loss": 0.6474943161010742, "epoch": 2.49, "learning_rate": 4.6340012523481526e-05, "loss": 45.0706, "step": 2944, "task_loss": 0.895974338054657 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41304606132823163, "compression/movement_sparsity/importance_threshold": -0.0038094160990598605, "compression/movement_sparsity/linear_layer_sparsity": 0.34281045905946744, "compression/movement_sparsity/model_sparsity": 0.33103386518244915, "compression_loss": 44.328819274902344, "distillation_loss": 0.9988981485366821, "epoch": 2.49, "learning_rate": 4.6336881653099565e-05, "loss": 45.2618, "step": 2945, "task_loss": 1.0241179466247559 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4136379476083367, "compression/movement_sparsity/importance_threshold": -0.0038055746713503073, "compression/movement_sparsity/linear_layer_sparsity": 0.3435861023158451, "compression/movement_sparsity/model_sparsity": 0.33178286270681245, "compression_loss": 44.391929626464844, "distillation_loss": 0.5825836658477783, "epoch": 2.49, "learning_rate": 4.63337507827176e-05, "loss": 45.4396, "step": 2946, "task_loss": 0.8771296739578247 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4142294358479186, "compression/movement_sparsity/importance_threshold": -0.0038017358269814175, "compression/movement_sparsity/linear_layer_sparsity": 0.34435385177324773, "compression/movement_sparsity/model_sparsity": 0.3325242376084797, "compression_loss": 44.454986572265625, "distillation_loss": 0.6715049147605896, "epoch": 2.49, "learning_rate": 4.6330619912335635e-05, "loss": 45.2895, "step": 2947, "task_loss": 0.868720531463623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41482052618086307, "compression/movement_sparsity/importance_threshold": -0.003797899565084253, "compression/movement_sparsity/linear_layer_sparsity": 0.34518305839064567, "compression/movement_sparsity/model_sparsity": 0.3333249584276321, "compression_loss": 44.51802062988281, "distillation_loss": 0.6880049705505371, "epoch": 2.49, "learning_rate": 4.632748904195367e-05, "loss": 45.6004, "step": 2948, "task_loss": 0.5791944265365601 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41541121874105424, "compression/movement_sparsity/importance_threshold": -0.0037940658847898856, "compression/movement_sparsity/linear_layer_sparsity": 0.3459559233159641, "compression/movement_sparsity/model_sparsity": 0.33407127306515527, "compression_loss": 44.58099365234375, "distillation_loss": 1.0281972885131836, "epoch": 2.49, "learning_rate": 4.63243581715717e-05, "loss": 45.6725, "step": 2949, "task_loss": 1.3690557479858398 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41600151366237803, "compression/movement_sparsity/importance_threshold": -0.003790234785229377, "compression/movement_sparsity/linear_layer_sparsity": 0.346692228743311, "compression/movement_sparsity/model_sparsity": 0.33478228413593236, "compression_loss": 44.64393615722656, "distillation_loss": 1.6550638675689697, "epoch": 2.49, "learning_rate": 4.632122730118974e-05, "loss": 45.652, "step": 2950, "task_loss": 1.2117295265197754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41659141107871933, "compression/movement_sparsity/importance_threshold": -0.003786406265533795, "compression/movement_sparsity/linear_layer_sparsity": 0.34759622374012106, "compression/movement_sparsity/model_sparsity": 0.3356552241235881, "compression_loss": 44.706844329833984, "distillation_loss": 1.1600335836410522, "epoch": 2.49, "learning_rate": 4.631809643080777e-05, "loss": 45.832, "step": 2951, "task_loss": 1.038508653640747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41718091112396327, "compression/movement_sparsity/importance_threshold": -0.0037825803248342053, "compression/movement_sparsity/linear_layer_sparsity": 0.34844976758366375, "compression/movement_sparsity/model_sparsity": 0.3364794461102971, "compression_loss": 44.76972579956055, "distillation_loss": 1.4116768836975098, "epoch": 2.5, "learning_rate": 4.63149655604258e-05, "loss": 46.0153, "step": 2952, "task_loss": 1.126796841621399 }, { "compression/movement_sparsity/importance_regularization_factor": 0.417770013931995, "compression/movement_sparsity/importance_threshold": -0.0037787569622616743, "compression/movement_sparsity/linear_layer_sparsity": 0.34926806358767487, "compression/movement_sparsity/model_sparsity": 0.3372696311291975, "compression_loss": 44.832557678222656, "distillation_loss": 0.917719841003418, "epoch": 2.5, "learning_rate": 4.631183469004384e-05, "loss": 45.8072, "step": 2953, "task_loss": 0.3113005757331848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41835871963669957, "compression/movement_sparsity/importance_threshold": -0.003774936176947268, "compression/movement_sparsity/linear_layer_sparsity": 0.35004646132677636, "compression/movement_sparsity/model_sparsity": 0.3380212885113293, "compression_loss": 44.895347595214844, "distillation_loss": 0.6710514426231384, "epoch": 2.5, "learning_rate": 4.630870381966187e-05, "loss": 45.7014, "step": 2954, "task_loss": 0.3705901503562927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4189470283719622, "compression/movement_sparsity/importance_threshold": -0.0037711179680220523, "compression/movement_sparsity/linear_layer_sparsity": 0.35073657252870194, "compression/movement_sparsity/model_sparsity": 0.3386876922704384, "compression_loss": 44.958045959472656, "distillation_loss": 0.883277416229248, "epoch": 2.5, "learning_rate": 4.63055729492799e-05, "loss": 45.8374, "step": 2955, "task_loss": 2.5270943641662598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.41953494027166816, "compression/movement_sparsity/importance_threshold": -0.0037673023346170924, "compression/movement_sparsity/linear_layer_sparsity": 0.35152099197245956, "compression/movement_sparsity/model_sparsity": 0.33944516449314643, "compression_loss": 45.020751953125, "distillation_loss": 1.385446310043335, "epoch": 2.5, "learning_rate": 4.6302442078897935e-05, "loss": 46.2633, "step": 2956, "task_loss": 0.6488084197044373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4201224554697024, "compression/movement_sparsity/importance_threshold": -0.003763489275863456, "compression/movement_sparsity/linear_layer_sparsity": 0.3522433580478401, "compression/movement_sparsity/model_sparsity": 0.3401427150715797, "compression_loss": 45.08336639404297, "distillation_loss": 1.93520188331604, "epoch": 2.5, "learning_rate": 4.629931120851597e-05, "loss": 46.5351, "step": 2957, "task_loss": 2.136819839477539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4207095740999499, "compression/movement_sparsity/importance_threshold": -0.003759678790892209, "compression/movement_sparsity/linear_layer_sparsity": 0.35298519628897007, "compression/movement_sparsity/model_sparsity": 0.3408590688869655, "compression_loss": 45.145973205566406, "distillation_loss": 0.8209676742553711, "epoch": 2.5, "learning_rate": 4.6296180338134005e-05, "loss": 45.8936, "step": 2958, "task_loss": 0.639210045337677 }, { "compression/movement_sparsity/importance_regularization_factor": 0.42129629629629617, "compression/movement_sparsity/importance_threshold": -0.0037558708788344164, "compression/movement_sparsity/linear_layer_sparsity": 0.35368428638912547, "compression/movement_sparsity/model_sparsity": 0.3415341430915278, "compression_loss": 45.20850372314453, "distillation_loss": 0.7936142683029175, "epoch": 2.5, "learning_rate": 4.629304946775204e-05, "loss": 46.1939, "step": 2959, "task_loss": 1.103090763092041 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4218826221926263, "compression/movement_sparsity/importance_threshold": -0.0037520655388211437, "compression/movement_sparsity/linear_layer_sparsity": 0.35444783853952033, "compression/movement_sparsity/model_sparsity": 0.34227146487659543, "compression_loss": 45.270973205566406, "distillation_loss": 0.5660673379898071, "epoch": 2.5, "learning_rate": 4.628991859737007e-05, "loss": 46.2823, "step": 2960, "task_loss": 0.42161619663238525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4224685519228254, "compression/movement_sparsity/importance_threshold": -0.0037482627699834572, "compression/movement_sparsity/linear_layer_sparsity": 0.3553242052399181, "compression/movement_sparsity/model_sparsity": 0.3431177256848152, "compression_loss": 45.333404541015625, "distillation_loss": 1.181326150894165, "epoch": 2.5, "learning_rate": 4.628678772698811e-05, "loss": 46.3755, "step": 2961, "task_loss": 1.352212905883789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4230540856207784, "compression/movement_sparsity/importance_threshold": -0.0037444625714524245, "compression/movement_sparsity/linear_layer_sparsity": 0.3561853090057419, "compression/movement_sparsity/model_sparsity": 0.343949247887218, "compression_loss": 45.39580535888672, "distillation_loss": 0.8366272449493408, "epoch": 2.5, "learning_rate": 4.628365685660614e-05, "loss": 46.2074, "step": 2962, "task_loss": 0.8022566437721252 }, { "compression/movement_sparsity/importance_regularization_factor": 0.42363922342037075, "compression/movement_sparsity/importance_threshold": -0.00374066494235911, "compression/movement_sparsity/linear_layer_sparsity": 0.3570297904818814, "compression/movement_sparsity/model_sparsity": 0.34476471882672316, "compression_loss": 45.4581413269043, "distillation_loss": 0.5549689531326294, "epoch": 2.5, "learning_rate": 4.628052598622417e-05, "loss": 46.3479, "step": 2963, "task_loss": 0.5890858769416809 }, { "compression/movement_sparsity/importance_regularization_factor": 0.42422396545548735, "compression/movement_sparsity/importance_threshold": -0.0037368698818345805, "compression/movement_sparsity/linear_layer_sparsity": 0.35768646631775597, "compression/movement_sparsity/model_sparsity": 0.34539883582746433, "compression_loss": 45.520450592041016, "distillation_loss": 1.1147891283035278, "epoch": 2.51, "learning_rate": 4.62773951158422e-05, "loss": 46.5234, "step": 2964, "task_loss": 0.7623768448829651 }, { "compression/movement_sparsity/importance_regularization_factor": 0.42480831186001333, "compression/movement_sparsity/importance_threshold": -0.003733077389009903, "compression/movement_sparsity/linear_layer_sparsity": 0.3586158836399657, "compression/movement_sparsity/model_sparsity": 0.346296324805434, "compression_loss": 45.582733154296875, "distillation_loss": 0.6492785811424255, "epoch": 2.51, "learning_rate": 4.627426424546024e-05, "loss": 46.663, "step": 2965, "task_loss": 0.17176377773284912 }, { "compression/movement_sparsity/importance_regularization_factor": 0.42539226276783404, "compression/movement_sparsity/importance_threshold": -0.0037292874630161415, "compression/movement_sparsity/linear_layer_sparsity": 0.35959925782072766, "compression/movement_sparsity/model_sparsity": 0.347245917057874, "compression_loss": 45.64497375488281, "distillation_loss": 0.9244823455810547, "epoch": 2.51, "learning_rate": 4.627113337507827e-05, "loss": 46.6565, "step": 2966, "task_loss": 1.0466699600219727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4259758183128345, "compression/movement_sparsity/importance_threshold": -0.003725500102984363, "compression/movement_sparsity/linear_layer_sparsity": 0.36052754234701195, "compression/movement_sparsity/model_sparsity": 0.3481423121549432, "compression_loss": 45.70710754394531, "distillation_loss": 0.901154637336731, "epoch": 2.51, "learning_rate": 4.6268002504696305e-05, "loss": 46.8646, "step": 2967, "task_loss": 1.107824444770813 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4265589786288996, "compression/movement_sparsity/importance_threshold": -0.0037217153080456354, "compression/movement_sparsity/linear_layer_sparsity": 0.36144850543436785, "compression/movement_sparsity/model_sparsity": 0.34903163732703457, "compression_loss": 45.769264221191406, "distillation_loss": 0.9288235902786255, "epoch": 2.51, "learning_rate": 4.626487163431434e-05, "loss": 46.9507, "step": 2968, "task_loss": 1.3967339992523193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.427141743849915, "compression/movement_sparsity/importance_threshold": -0.003717933077331021, "compression/movement_sparsity/linear_layer_sparsity": 0.3622585976110455, "compression/movement_sparsity/model_sparsity": 0.34981390034530824, "compression_loss": 45.831363677978516, "distillation_loss": 0.766335666179657, "epoch": 2.51, "learning_rate": 4.6261740763932375e-05, "loss": 46.9817, "step": 2969, "task_loss": 0.6858511567115784 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4277241141097655, "compression/movement_sparsity/importance_threshold": -0.003714153409971588, "compression/movement_sparsity/linear_layer_sparsity": 0.3631852604505313, "compression/movement_sparsity/model_sparsity": 0.3507087294655094, "compression_loss": 45.893409729003906, "distillation_loss": 1.1016695499420166, "epoch": 2.51, "learning_rate": 4.625860989355041e-05, "loss": 46.9162, "step": 2970, "task_loss": 0.9459066987037659 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4283060895423363, "compression/movement_sparsity/importance_threshold": -0.0037103763050984027, "compression/movement_sparsity/linear_layer_sparsity": 0.3639856702030886, "compression/movement_sparsity/model_sparsity": 0.35148164268071797, "compression_loss": 45.955421447753906, "distillation_loss": 0.5982421636581421, "epoch": 2.51, "learning_rate": 4.625547902316844e-05, "loss": 46.7965, "step": 2971, "task_loss": 0.45809364318847656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.42888767028151265, "compression/movement_sparsity/importance_threshold": -0.00370660176184253, "compression/movement_sparsity/linear_layer_sparsity": 0.36493514397526183, "compression/movement_sparsity/model_sparsity": 0.35239849910789406, "compression_loss": 46.01740646362305, "distillation_loss": 1.3665356636047363, "epoch": 2.51, "learning_rate": 4.625234815278648e-05, "loss": 47.3068, "step": 2972, "task_loss": 0.8063247203826904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4294688564611795, "compression/movement_sparsity/importance_threshold": -0.003702829779335037, "compression/movement_sparsity/linear_layer_sparsity": 0.3656145472746504, "compression/movement_sparsity/model_sparsity": 0.3530545628138596, "compression_loss": 46.079322814941406, "distillation_loss": 0.9290417432785034, "epoch": 2.51, "learning_rate": 4.624921728240451e-05, "loss": 47.3795, "step": 2973, "task_loss": 0.7172600030899048 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4300496482152222, "compression/movement_sparsity/importance_threshold": -0.0036990603567069887, "compression/movement_sparsity/linear_layer_sparsity": 0.36640386755130633, "compression/movement_sparsity/model_sparsity": 0.35381676751077923, "compression_loss": 46.141231536865234, "distillation_loss": 1.3607105016708374, "epoch": 2.51, "learning_rate": 4.624608641202254e-05, "loss": 47.1453, "step": 2974, "task_loss": 0.9460224509239197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4306300456775256, "compression/movement_sparsity/importance_threshold": -0.0036952934930894526, "compression/movement_sparsity/linear_layer_sparsity": 0.3671693394926906, "compression/movement_sparsity/model_sparsity": 0.3545559431361098, "compression_loss": 46.203094482421875, "distillation_loss": 1.722400188446045, "epoch": 2.51, "learning_rate": 4.624295554164057e-05, "loss": 47.3999, "step": 2975, "task_loss": 1.8885767459869385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4312100489819751, "compression/movement_sparsity/importance_threshold": -0.0036915291876134936, "compression/movement_sparsity/linear_layer_sparsity": 0.367955762196611, "compression/movement_sparsity/model_sparsity": 0.35531534980083135, "compression_loss": 46.264930725097656, "distillation_loss": 1.2326338291168213, "epoch": 2.52, "learning_rate": 4.623982467125861e-05, "loss": 47.4629, "step": 2976, "task_loss": 0.939583957195282 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43178965826245586, "compression/movement_sparsity/importance_threshold": -0.0036877674394101773, "compression/movement_sparsity/linear_layer_sparsity": 0.368589472085619, "compression/movement_sparsity/model_sparsity": 0.3559272898056322, "compression_loss": 46.32670974731445, "distillation_loss": 0.8266104459762573, "epoch": 2.52, "learning_rate": 4.623669380087664e-05, "loss": 47.3876, "step": 2977, "task_loss": 0.268105685710907 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4323688736528526, "compression/movement_sparsity/importance_threshold": -0.0036840082476105726, "compression/movement_sparsity/linear_layer_sparsity": 0.36932414390199975, "compression/movement_sparsity/model_sparsity": 0.35663672338500546, "compression_loss": 46.38847351074219, "distillation_loss": 2.0449156761169434, "epoch": 2.52, "learning_rate": 4.623356293049468e-05, "loss": 47.6451, "step": 2978, "task_loss": 1.3121364116668701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.432947695287051, "compression/movement_sparsity/importance_threshold": -0.003680251611345742, "compression/movement_sparsity/linear_layer_sparsity": 0.3700689751092063, "compression/movement_sparsity/model_sparsity": 0.35735596734887565, "compression_loss": 46.450172424316406, "distillation_loss": 1.146535873413086, "epoch": 2.52, "learning_rate": 4.623043206011271e-05, "loss": 47.6191, "step": 2979, "task_loss": 0.8306382298469543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43352612329893603, "compression/movement_sparsity/importance_threshold": -0.003676497529746753, "compression/movement_sparsity/linear_layer_sparsity": 0.3708207819544798, "compression/movement_sparsity/model_sparsity": 0.35808194731618564, "compression_loss": 46.51185607910156, "distillation_loss": 1.1742773056030273, "epoch": 2.52, "learning_rate": 4.622730118973075e-05, "loss": 47.424, "step": 2980, "task_loss": 0.8094872236251831 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43410415782239276, "compression/movement_sparsity/importance_threshold": -0.003672746001944671, "compression/movement_sparsity/linear_layer_sparsity": 0.371566936744294, "compression/movement_sparsity/model_sparsity": 0.35880246939352906, "compression_loss": 46.573482513427734, "distillation_loss": 1.0738784074783325, "epoch": 2.52, "learning_rate": 4.6224170319348784e-05, "loss": 47.9315, "step": 2981, "task_loss": 0.6557512879371643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4346817989913063, "compression/movement_sparsity/importance_threshold": -0.0036689970270705635, "compression/movement_sparsity/linear_layer_sparsity": 0.37244937284601837, "compression/movement_sparsity/model_sparsity": 0.35965459110046827, "compression_loss": 46.63504409790039, "distillation_loss": 1.2947839498519897, "epoch": 2.52, "learning_rate": 4.6221039448966815e-05, "loss": 47.7694, "step": 2982, "task_loss": 0.49067604541778564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43525904693956197, "compression/movement_sparsity/importance_threshold": -0.0036652506042554947, "compression/movement_sparsity/linear_layer_sparsity": 0.3731950983657976, "compression/movement_sparsity/model_sparsity": 0.36037469865452304, "compression_loss": 46.69660186767578, "distillation_loss": 0.6079787015914917, "epoch": 2.52, "learning_rate": 4.6217908578584854e-05, "loss": 47.8819, "step": 2983, "task_loss": 1.0253175497055054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4358359018010447, "compression/movement_sparsity/importance_threshold": -0.003661506732630532, "compression/movement_sparsity/linear_layer_sparsity": 0.37398305928734304, "compression/movement_sparsity/model_sparsity": 0.3611355906943621, "compression_loss": 46.75809097290039, "distillation_loss": 0.8896089196205139, "epoch": 2.52, "learning_rate": 4.6214777708202886e-05, "loss": 47.9421, "step": 2984, "task_loss": 0.45257365703582764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43641236370963965, "compression/movement_sparsity/importance_threshold": -0.0036577654113267417, "compression/movement_sparsity/linear_layer_sparsity": 0.37482256838557837, "compression/movement_sparsity/model_sparsity": 0.36194626007244096, "compression_loss": 46.819557189941406, "distillation_loss": 0.8575799465179443, "epoch": 2.52, "learning_rate": 4.621164683782092e-05, "loss": 47.8699, "step": 2985, "task_loss": 0.37607815861701965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43698843279923205, "compression/movement_sparsity/importance_threshold": -0.003654026639475189, "compression/movement_sparsity/linear_layer_sparsity": 0.37542577625377377, "compression/movement_sparsity/model_sparsity": 0.36252874589467937, "compression_loss": 46.880977630615234, "distillation_loss": 0.9719109535217285, "epoch": 2.52, "learning_rate": 4.620851596743895e-05, "loss": 47.7689, "step": 2986, "task_loss": 1.199438214302063 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43756410920370714, "compression/movement_sparsity/importance_threshold": -0.003650290416206939, "compression/movement_sparsity/linear_layer_sparsity": 0.37615865944500915, "compression/movement_sparsity/model_sparsity": 0.36323645229368345, "compression_loss": 46.942325592041016, "distillation_loss": 1.1867564916610718, "epoch": 2.52, "learning_rate": 4.620538509705699e-05, "loss": 48.2498, "step": 2987, "task_loss": 0.6405383348464966 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4381393930569496, "compression/movement_sparsity/importance_threshold": -0.0036465567406530615, "compression/movement_sparsity/linear_layer_sparsity": 0.3768227878856562, "compression/movement_sparsity/model_sparsity": 0.3638777658792962, "compression_loss": 47.003631591796875, "distillation_loss": 1.8391717672348022, "epoch": 2.53, "learning_rate": 4.620225422667502e-05, "loss": 48.5128, "step": 2988, "task_loss": 1.7843725681304932 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4387142844928451, "compression/movement_sparsity/importance_threshold": -0.0036428256119446184, "compression/movement_sparsity/linear_layer_sparsity": 0.3775359723519572, "compression/movement_sparsity/model_sparsity": 0.36456645026516765, "compression_loss": 47.06492614746094, "distillation_loss": 1.1227415800094604, "epoch": 2.53, "learning_rate": 4.619912335629305e-05, "loss": 48.2419, "step": 2989, "task_loss": 0.4399822950363159 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4392887836452787, "compression/movement_sparsity/importance_threshold": -0.0036390970292126765, "compression/movement_sparsity/linear_layer_sparsity": 0.3782659579704571, "compression/movement_sparsity/model_sparsity": 0.3652713586319737, "compression_loss": 47.12617874145508, "distillation_loss": 1.0227115154266357, "epoch": 2.53, "learning_rate": 4.619599248591109e-05, "loss": 48.1945, "step": 2990, "task_loss": 0.7862231731414795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.43986289064813533, "compression/movement_sparsity/importance_threshold": -0.0036353709915883034, "compression/movement_sparsity/linear_layer_sparsity": 0.37904305597528626, "compression/movement_sparsity/model_sparsity": 0.3660217609297039, "compression_loss": 47.18732833862305, "distillation_loss": 1.0007266998291016, "epoch": 2.53, "learning_rate": 4.619286161552912e-05, "loss": 48.1629, "step": 2991, "task_loss": 0.8205500245094299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4404366056353003, "compression/movement_sparsity/importance_threshold": -0.003631647498202564, "compression/movement_sparsity/linear_layer_sparsity": 0.37991939882734876, "compression/movement_sparsity/model_sparsity": 0.36686799870885206, "compression_loss": 47.248443603515625, "distillation_loss": 0.7449820041656494, "epoch": 2.53, "learning_rate": 4.6189730745147153e-05, "loss": 48.4599, "step": 2992, "task_loss": 1.4738025665283203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4410099287406586, "compression/movement_sparsity/importance_threshold": -0.0036279265481865245, "compression/movement_sparsity/linear_layer_sparsity": 0.38055619707577437, "compression/movement_sparsity/model_sparsity": 0.3674829209784237, "compression_loss": 47.30953598022461, "distillation_loss": 1.039381504058838, "epoch": 2.53, "learning_rate": 4.6186599874765185e-05, "loss": 48.6303, "step": 2993, "task_loss": 0.8654632568359375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.44158286009809544, "compression/movement_sparsity/importance_threshold": -0.003624208140671252, "compression/movement_sparsity/linear_layer_sparsity": 0.3813774025766886, "compression/movement_sparsity/model_sparsity": 0.368275915544058, "compression_loss": 47.370582580566406, "distillation_loss": 1.7378082275390625, "epoch": 2.53, "learning_rate": 4.6183469004383224e-05, "loss": 48.4121, "step": 2994, "task_loss": 0.8818203210830688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.442155399841496, "compression/movement_sparsity/importance_threshold": -0.003620492274787811, "compression/movement_sparsity/linear_layer_sparsity": 0.38221160542032595, "compression/movement_sparsity/model_sparsity": 0.36908146095370825, "compression_loss": 47.43159484863281, "distillation_loss": 2.188715934753418, "epoch": 2.53, "learning_rate": 4.6180338134001256e-05, "loss": 48.6925, "step": 2995, "task_loss": 2.0982675552368164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.44272754810474546, "compression/movement_sparsity/importance_threshold": -0.0036167789496672677, "compression/movement_sparsity/linear_layer_sparsity": 0.3829935088647124, "compression/movement_sparsity/model_sparsity": 0.36983650360936365, "compression_loss": 47.492549896240234, "distillation_loss": 0.7178877592086792, "epoch": 2.53, "learning_rate": 4.617720726361929e-05, "loss": 48.7175, "step": 2996, "task_loss": 1.1483616828918457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4432993050217289, "compression/movement_sparsity/importance_threshold": -0.0036130681644406886, "compression/movement_sparsity/linear_layer_sparsity": 0.38379329856055266, "compression/movement_sparsity/model_sparsity": 0.3706088180687109, "compression_loss": 47.553489685058594, "distillation_loss": 2.2231712341308594, "epoch": 2.53, "learning_rate": 4.617407639323732e-05, "loss": 48.6895, "step": 2997, "task_loss": 1.0918434858322144 }, { "compression/movement_sparsity/importance_regularization_factor": 0.44387067072633113, "compression/movement_sparsity/importance_threshold": -0.003609359918239142, "compression/movement_sparsity/linear_layer_sparsity": 0.3845433644773513, "compression/movement_sparsity/model_sparsity": 0.3713331169137949, "compression_loss": 47.614383697509766, "distillation_loss": 1.2537977695465088, "epoch": 2.53, "learning_rate": 4.617094552285536e-05, "loss": 48.8379, "step": 2998, "task_loss": 1.1942179203033447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4444416453524378, "compression/movement_sparsity/importance_threshold": -0.0036056542101936904, "compression/movement_sparsity/linear_layer_sparsity": 0.38536836186357376, "compression/movement_sparsity/model_sparsity": 0.3721297731018118, "compression_loss": 47.67518997192383, "distillation_loss": 1.9196422100067139, "epoch": 2.53, "learning_rate": 4.616781465247339e-05, "loss": 49.0529, "step": 2999, "task_loss": 0.9124406576156616 }, { "compression/movement_sparsity/importance_regularization_factor": 0.445012229033934, "compression/movement_sparsity/importance_threshold": -0.0036019510394354, "compression/movement_sparsity/linear_layer_sparsity": 0.3861825678780858, "compression/movement_sparsity/model_sparsity": 0.3729160086349346, "compression_loss": 47.73597717285156, "distillation_loss": 1.357574224472046, "epoch": 2.54, "learning_rate": 4.616468378209142e-05, "loss": 49.0606, "step": 3000, "task_loss": 1.1198880672454834 }, { "epoch": 2.54, "eval_accuracy": 0.8525940594059406, "eval_loss": 48.499977111816406, "eval_runtime": 209.0446, "eval_samples_per_second": 120.788, "eval_steps_per_second": 0.947, "step": 3000 }, { "compression/movement_sparsity/importance_regularization_factor": 0.44558242190470465, "compression/movement_sparsity/importance_threshold": -0.003598250405095339, "compression/movement_sparsity/linear_layer_sparsity": 0.38693547174678183, "compression/movement_sparsity/model_sparsity": 0.37364304793953773, "compression_loss": 47.79674530029297, "distillation_loss": 0.7883907556533813, "epoch": 2.54, "learning_rate": 4.616155291170945e-05, "loss": 48.8798, "step": 3001, "task_loss": 0.34267064929008484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.446152224098635, "compression/movement_sparsity/importance_threshold": -0.0035945523063045724, "compression/movement_sparsity/linear_layer_sparsity": 0.3875951047762301, "compression/movement_sparsity/model_sparsity": 0.3742800205451559, "compression_loss": 47.85750198364258, "distillation_loss": 1.596421718597412, "epoch": 2.54, "learning_rate": 4.615842204132749e-05, "loss": 49.1227, "step": 3002, "task_loss": 2.00067138671875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4467216357496101, "compression/movement_sparsity/importance_threshold": -0.0035908567421941667, "compression/movement_sparsity/linear_layer_sparsity": 0.388418253916469, "compression/movement_sparsity/model_sparsity": 0.37507489198012467, "compression_loss": 47.918174743652344, "distillation_loss": 1.035849690437317, "epoch": 2.54, "learning_rate": 4.6155291170945523e-05, "loss": 49.249, "step": 3003, "task_loss": 1.3060356378555298 }, { "compression/movement_sparsity/importance_regularization_factor": 0.44729065699151516, "compression/movement_sparsity/importance_threshold": -0.003587163711895187, "compression/movement_sparsity/linear_layer_sparsity": 0.389169357235852, "compression/movement_sparsity/model_sparsity": 0.3758001925898228, "compression_loss": 47.97883605957031, "distillation_loss": 1.1889724731445312, "epoch": 2.54, "learning_rate": 4.6152160300563555e-05, "loss": 49.2524, "step": 3004, "task_loss": 1.1960442066192627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4478592879582354, "compression/movement_sparsity/importance_threshold": -0.0035834732145387, "compression/movement_sparsity/linear_layer_sparsity": 0.3898109013029967, "compression/movement_sparsity/model_sparsity": 0.37641969764464067, "compression_loss": 48.039459228515625, "distillation_loss": 0.6096097230911255, "epoch": 2.54, "learning_rate": 4.6149029430181594e-05, "loss": 49.0032, "step": 3005, "task_loss": 1.6016497611999512 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4484275287836559, "compression/movement_sparsity/importance_threshold": -0.003579785249255771, "compression/movement_sparsity/linear_layer_sparsity": 0.3906277664068915, "compression/movement_sparsity/model_sparsity": 0.3772085009192457, "compression_loss": 48.10003662109375, "distillation_loss": 1.0870728492736816, "epoch": 2.54, "learning_rate": 4.6145898559799626e-05, "loss": 49.424, "step": 3006, "task_loss": 1.130446434020996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4489953796016617, "compression/movement_sparsity/importance_threshold": -0.003576099815177467, "compression/movement_sparsity/linear_layer_sparsity": 0.3914810359945786, "compression/movement_sparsity/model_sparsity": 0.37803245807163144, "compression_loss": 48.160552978515625, "distillation_loss": 1.0132704973220825, "epoch": 2.54, "learning_rate": 4.614276768941766e-05, "loss": 49.3147, "step": 3007, "task_loss": 0.5061923861503601 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4495628405461377, "compression/movement_sparsity/importance_threshold": -0.0035724169114348557, "compression/movement_sparsity/linear_layer_sparsity": 0.3922309826697009, "compression/movement_sparsity/model_sparsity": 0.3787566417713575, "compression_loss": 48.221012115478516, "distillation_loss": 0.9048013091087341, "epoch": 2.54, "learning_rate": 4.613963681903569e-05, "loss": 49.4093, "step": 3008, "task_loss": 1.7405918836593628 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4501299117509697, "compression/movement_sparsity/importance_threshold": -0.003568736537158999, "compression/movement_sparsity/linear_layer_sparsity": 0.39305180659725086, "compression/movement_sparsity/model_sparsity": 0.3795492678718463, "compression_loss": 48.28147506713867, "distillation_loss": 1.1665695905685425, "epoch": 2.54, "learning_rate": 4.613650594865373e-05, "loss": 49.5099, "step": 3009, "task_loss": 0.8124213814735413 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45069659335004253, "compression/movement_sparsity/importance_threshold": -0.003565058691480965, "compression/movement_sparsity/linear_layer_sparsity": 0.3937940502600804, "compression/movement_sparsity/model_sparsity": 0.3802660131814491, "compression_loss": 48.34183120727539, "distillation_loss": 1.4951739311218262, "epoch": 2.54, "learning_rate": 4.613337507827176e-05, "loss": 49.3971, "step": 3010, "task_loss": 1.2046178579330444 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45126288547724125, "compression/movement_sparsity/importance_threshold": -0.00356138337353182, "compression/movement_sparsity/linear_layer_sparsity": 0.3945490050856098, "compression/movement_sparsity/model_sparsity": 0.38099503298620885, "compression_loss": 48.402183532714844, "distillation_loss": 0.9649286270141602, "epoch": 2.54, "learning_rate": 4.613024420788979e-05, "loss": 49.5819, "step": 3011, "task_loss": 0.7643091082572937 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4518287882664511, "compression/movement_sparsity/importance_threshold": -0.0035577105824426297, "compression/movement_sparsity/linear_layer_sparsity": 0.39536312763094833, "compression/movement_sparsity/model_sparsity": 0.38178118791758114, "compression_loss": 48.46248245239258, "distillation_loss": 0.853033185005188, "epoch": 2.55, "learning_rate": 4.612711333750783e-05, "loss": 49.299, "step": 3012, "task_loss": 0.7242262363433838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45239430185155705, "compression/movement_sparsity/importance_threshold": -0.003554040317344461, "compression/movement_sparsity/linear_layer_sparsity": 0.3961419546400847, "compression/movement_sparsity/model_sparsity": 0.38253325982300157, "compression_loss": 48.52271270751953, "distillation_loss": 1.8731712102890015, "epoch": 2.55, "learning_rate": 4.612398246712586e-05, "loss": 50.0661, "step": 3013, "task_loss": 0.9525712132453918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4529594263664445, "compression/movement_sparsity/importance_threshold": -0.0035503725773683784, "compression/movement_sparsity/linear_layer_sparsity": 0.3969815472074935, "compression/movement_sparsity/model_sparsity": 0.38334400980283095, "compression_loss": 48.58292770385742, "distillation_loss": 1.4801769256591797, "epoch": 2.55, "learning_rate": 4.61208515967439e-05, "loss": 49.9521, "step": 3014, "task_loss": 0.7349382638931274 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45352416194499834, "compression/movement_sparsity/importance_threshold": -0.0035467073616454493, "compression/movement_sparsity/linear_layer_sparsity": 0.397761745495908, "compression/movement_sparsity/model_sparsity": 0.3840974058798678, "compression_loss": 48.64308166503906, "distillation_loss": 1.116469383239746, "epoch": 2.55, "learning_rate": 4.611772072636193e-05, "loss": 49.7579, "step": 3015, "task_loss": 0.717325747013092 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45408850872110396, "compression/movement_sparsity/importance_threshold": -0.0035430446693067385, "compression/movement_sparsity/linear_layer_sparsity": 0.3987355684183937, "compression/movement_sparsity/model_sparsity": 0.3850377749891363, "compression_loss": 48.703208923339844, "distillation_loss": 1.763638973236084, "epoch": 2.55, "learning_rate": 4.611458985597997e-05, "loss": 50.3364, "step": 3016, "task_loss": 1.7873024940490723 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45465246682864624, "compression/movement_sparsity/importance_threshold": -0.0035393844994833132, "compression/movement_sparsity/linear_layer_sparsity": 0.3993987429256298, "compression/movement_sparsity/model_sparsity": 0.3856781674118855, "compression_loss": 48.76327896118164, "distillation_loss": 0.8483915328979492, "epoch": 2.55, "learning_rate": 4.6111458985598e-05, "loss": 49.8685, "step": 3017, "task_loss": 1.2063602209091187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4552160364015102, "compression/movement_sparsity/importance_threshold": -0.0035357268513062407, "compression/movement_sparsity/linear_layer_sparsity": 0.4001612457492727, "compression/movement_sparsity/model_sparsity": 0.3864144759178032, "compression_loss": 48.82331085205078, "distillation_loss": 0.9016829133033752, "epoch": 2.55, "learning_rate": 4.6108328115216034e-05, "loss": 49.6509, "step": 3018, "task_loss": 0.6806535124778748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45577921757358153, "compression/movement_sparsity/importance_threshold": -0.0035320717239065835, "compression/movement_sparsity/linear_layer_sparsity": 0.40080402992985154, "compression/movement_sparsity/model_sparsity": 0.3870351784843437, "compression_loss": 48.883296966552734, "distillation_loss": 1.1798243522644043, "epoch": 2.55, "learning_rate": 4.6105197244834066e-05, "loss": 49.9414, "step": 3019, "task_loss": 1.6550252437591553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.456342010478745, "compression/movement_sparsity/importance_threshold": -0.00352841911641541, "compression/movement_sparsity/linear_layer_sparsity": 0.40146988737480577, "compression/movement_sparsity/model_sparsity": 0.3876781616776467, "compression_loss": 48.943241119384766, "distillation_loss": 0.891191840171814, "epoch": 2.55, "learning_rate": 4.6102066374452104e-05, "loss": 50.035, "step": 3020, "task_loss": 0.6705802083015442 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4569044152508859, "compression/movement_sparsity/importance_threshold": -0.003524769027963785, "compression/movement_sparsity/linear_layer_sparsity": 0.4021666761106074, "compression/movement_sparsity/model_sparsity": 0.3883510135768007, "compression_loss": 49.00313186645508, "distillation_loss": 0.8527969717979431, "epoch": 2.55, "learning_rate": 4.6098935504070136e-05, "loss": 50.2651, "step": 3021, "task_loss": 0.5501022338867188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45746643202388926, "compression/movement_sparsity/importance_threshold": -0.0035211214576827752, "compression/movement_sparsity/linear_layer_sparsity": 0.40300992939748037, "compression/movement_sparsity/model_sparsity": 0.389165298519119, "compression_loss": 49.0629768371582, "distillation_loss": 0.9312759637832642, "epoch": 2.55, "learning_rate": 4.609580463368817e-05, "loss": 50.2049, "step": 3022, "task_loss": 1.307228446006775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45802806093164017, "compression/movement_sparsity/importance_threshold": -0.0035174764047034477, "compression/movement_sparsity/linear_layer_sparsity": 0.40395800804204024, "compression/movement_sparsity/model_sparsity": 0.39008080774560716, "compression_loss": 49.12279510498047, "distillation_loss": 1.3395168781280518, "epoch": 2.56, "learning_rate": 4.60926737633062e-05, "loss": 50.182, "step": 3023, "task_loss": 0.6248884797096252 }, { "compression/movement_sparsity/importance_regularization_factor": 0.45858930210802396, "compression/movement_sparsity/importance_threshold": -0.0035138338681568663, "compression/movement_sparsity/linear_layer_sparsity": 0.40470968372146976, "compression/movement_sparsity/model_sparsity": 0.3908066610530234, "compression_loss": 49.182525634765625, "distillation_loss": 0.828244686126709, "epoch": 2.56, "learning_rate": 4.608954289292424e-05, "loss": 50.1524, "step": 3024, "task_loss": 0.3207714855670929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4591501556869255, "compression/movement_sparsity/importance_threshold": -0.0035101938471740995, "compression/movement_sparsity/linear_layer_sparsity": 0.4055934553299694, "compression/movement_sparsity/model_sparsity": 0.3916600723879716, "compression_loss": 49.24222946166992, "distillation_loss": 1.1087942123413086, "epoch": 2.56, "learning_rate": 4.608641202254227e-05, "loss": 50.1614, "step": 3025, "task_loss": 1.1194664239883423 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4597106218022302, "compression/movement_sparsity/importance_threshold": -0.0035065563408862113, "compression/movement_sparsity/linear_layer_sparsity": 0.4064942069531825, "compression/movement_sparsity/model_sparsity": 0.3925298804218912, "compression_loss": 49.30186462402344, "distillation_loss": 1.2210177183151245, "epoch": 2.56, "learning_rate": 4.60832811521603e-05, "loss": 50.6592, "step": 3026, "task_loss": 1.4186598062515259 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46027070058782316, "compression/movement_sparsity/importance_threshold": -0.003502921348424268, "compression/movement_sparsity/linear_layer_sparsity": 0.4073105950903718, "compression/movement_sparsity/model_sparsity": 0.3933182231150644, "compression_loss": 49.36148452758789, "distillation_loss": 1.0785707235336304, "epoch": 2.56, "learning_rate": 4.608015028177834e-05, "loss": 50.4967, "step": 3027, "task_loss": 0.4889340102672577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46083039217758914, "compression/movement_sparsity/importance_threshold": -0.0034992888689193385, "compression/movement_sparsity/linear_layer_sparsity": 0.40812124770292835, "compression/movement_sparsity/model_sparsity": 0.3941010273165205, "compression_loss": 49.42107391357422, "distillation_loss": 1.4543458223342896, "epoch": 2.56, "learning_rate": 4.607701941139637e-05, "loss": 50.8867, "step": 3028, "task_loss": 0.844824492931366 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46138969670541374, "compression/movement_sparsity/importance_threshold": -0.0034956589015024855, "compression/movement_sparsity/linear_layer_sparsity": 0.40883104570562073, "compression/movement_sparsity/model_sparsity": 0.3947864415742263, "compression_loss": 49.48062515258789, "distillation_loss": 0.9841254949569702, "epoch": 2.56, "learning_rate": 4.6073888541014404e-05, "loss": 50.7115, "step": 3029, "task_loss": 0.6173140406608582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4619486143051821, "compression/movement_sparsity/importance_threshold": -0.003492031445304775, "compression/movement_sparsity/linear_layer_sparsity": 0.4098034377279901, "compression/movement_sparsity/model_sparsity": 0.39572542893919954, "compression_loss": 49.54011535644531, "distillation_loss": 1.3596410751342773, "epoch": 2.56, "learning_rate": 4.6070757670632436e-05, "loss": 50.9316, "step": 3030, "task_loss": 0.6872830986976624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46250714511077906, "compression/movement_sparsity/importance_threshold": -0.003488406499457275, "compression/movement_sparsity/linear_layer_sparsity": 0.4105334352706576, "compression/movement_sparsity/model_sparsity": 0.39643034882054135, "compression_loss": 49.599609375, "distillation_loss": 1.5262451171875, "epoch": 2.56, "learning_rate": 4.6067626800250474e-05, "loss": 50.5298, "step": 3031, "task_loss": 1.6148536205291748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46306528925609003, "compression/movement_sparsity/importance_threshold": -0.00348478406309105, "compression/movement_sparsity/linear_layer_sparsity": 0.41133404773406473, "compression/movement_sparsity/model_sparsity": 0.3972034577828584, "compression_loss": 49.659019470214844, "distillation_loss": 1.2843658924102783, "epoch": 2.56, "learning_rate": 4.6064495929868506e-05, "loss": 50.8546, "step": 3032, "task_loss": 1.0581368207931519 }, { "compression/movement_sparsity/importance_regularization_factor": 0.463623046875, "compression/movement_sparsity/importance_threshold": -0.003481164135337167, "compression/movement_sparsity/linear_layer_sparsity": 0.4122279429608872, "compression/movement_sparsity/model_sparsity": 0.39806664495869615, "compression_loss": 49.71843338012695, "distillation_loss": 0.647236168384552, "epoch": 2.56, "learning_rate": 4.606136505948654e-05, "loss": 50.7379, "step": 3033, "task_loss": 1.5364799499511719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4641804181013941, "compression/movement_sparsity/importance_threshold": -0.0034775467153266917, "compression/movement_sparsity/linear_layer_sparsity": 0.41298405443067726, "compression/movement_sparsity/model_sparsity": 0.39879678167342797, "compression_loss": 49.77777099609375, "distillation_loss": 1.3885314464569092, "epoch": 2.56, "learning_rate": 4.605823418910457e-05, "loss": 51.0689, "step": 3034, "task_loss": 1.4492218494415283 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46473740306915756, "compression/movement_sparsity/importance_threshold": -0.00347393180219069, "compression/movement_sparsity/linear_layer_sparsity": 0.4137628337431431, "compression/movement_sparsity/model_sparsity": 0.39954880752070526, "compression_loss": 49.83711242675781, "distillation_loss": 0.9964950680732727, "epoch": 2.57, "learning_rate": 4.605510331872261e-05, "loss": 50.9323, "step": 3035, "task_loss": 1.0271660089492798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4652940019121755, "compression/movement_sparsity/importance_threshold": -0.0034703193950602284, "compression/movement_sparsity/linear_layer_sparsity": 0.4146746629180899, "compression/movement_sparsity/model_sparsity": 0.400429312558378, "compression_loss": 49.89639663696289, "distillation_loss": 1.2218633890151978, "epoch": 2.57, "learning_rate": 4.605197244834064e-05, "loss": 50.9048, "step": 3036, "task_loss": 1.7682394981384277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46585021476433297, "compression/movement_sparsity/importance_threshold": -0.0034667094930663726, "compression/movement_sparsity/linear_layer_sparsity": 0.41537374109407765, "compression/movement_sparsity/model_sparsity": 0.40110437524840453, "compression_loss": 49.95565414428711, "distillation_loss": 0.8477005362510681, "epoch": 2.57, "learning_rate": 4.604884157795867e-05, "loss": 51.1667, "step": 3037, "task_loss": 1.0512068271636963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4664060417595153, "compression/movement_sparsity/importance_threshold": -0.003463102095340188, "compression/movement_sparsity/linear_layer_sparsity": 0.41626082762118005, "compression/movement_sparsity/model_sparsity": 0.4019609876243036, "compression_loss": 50.01482391357422, "distillation_loss": 0.6165015697479248, "epoch": 2.57, "learning_rate": 4.6045710707576704e-05, "loss": 51.074, "step": 3038, "task_loss": 0.5026558041572571 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4669614830316071, "compression/movement_sparsity/importance_threshold": -0.003459497201012744, "compression/movement_sparsity/linear_layer_sparsity": 0.4171298728826494, "compression/movement_sparsity/model_sparsity": 0.4028001785075455, "compression_loss": 50.07398986816406, "distillation_loss": 2.2888708114624023, "epoch": 2.57, "learning_rate": 4.604257983719474e-05, "loss": 51.6146, "step": 3039, "task_loss": 1.57419753074646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4675165387144943, "compression/movement_sparsity/importance_threshold": -0.0034558948092151013, "compression/movement_sparsity/linear_layer_sparsity": 0.41786951707693437, "compression/movement_sparsity/model_sparsity": 0.40351441364834506, "compression_loss": 50.13309097290039, "distillation_loss": 1.2059483528137207, "epoch": 2.57, "learning_rate": 4.6039448966812774e-05, "loss": 51.4796, "step": 3040, "task_loss": 1.3771718740463257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.46807120894206156, "compression/movement_sparsity/importance_threshold": -0.0034522949190783294, "compression/movement_sparsity/linear_layer_sparsity": 0.4186502638770601, "compression/movement_sparsity/model_sparsity": 0.40426833939402845, "compression_loss": 50.192161560058594, "distillation_loss": 2.4123082160949707, "epoch": 2.57, "learning_rate": 4.6036318096430806e-05, "loss": 51.5858, "step": 3041, "task_loss": 1.4918067455291748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4686254938481941, "compression/movement_sparsity/importance_threshold": -0.0034486975297334937, "compression/movement_sparsity/linear_layer_sparsity": 0.41942926974871103, "compression/movement_sparsity/model_sparsity": 0.4050205840174858, "compression_loss": 50.25119400024414, "distillation_loss": 2.0995240211486816, "epoch": 2.57, "learning_rate": 4.6033187226048844e-05, "loss": 51.8054, "step": 3042, "task_loss": 1.4542911052703857 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4691793935667772, "compression/movement_sparsity/importance_threshold": -0.0034451026403116594, "compression/movement_sparsity/linear_layer_sparsity": 0.42019261918825607, "compression/movement_sparsity/model_sparsity": 0.4057577100554449, "compression_loss": 50.3101806640625, "distillation_loss": 1.601941704750061, "epoch": 2.57, "learning_rate": 4.6030056355666876e-05, "loss": 51.5382, "step": 3043, "task_loss": 1.5692179203033447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4697329082316958, "compression/movement_sparsity/importance_threshold": -0.0034415102499438938, "compression/movement_sparsity/linear_layer_sparsity": 0.42101253687906565, "compression/movement_sparsity/model_sparsity": 0.4065494610512133, "compression_loss": 50.36914825439453, "distillation_loss": 1.8784395456314087, "epoch": 2.57, "learning_rate": 4.602692548528491e-05, "loss": 51.8507, "step": 3044, "task_loss": 1.3912781476974487 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4702860379768351, "compression/movement_sparsity/importance_threshold": -0.003437920357761262, "compression/movement_sparsity/linear_layer_sparsity": 0.4216344179937788, "compression/movement_sparsity/model_sparsity": 0.40714997863650604, "compression_loss": 50.42805862426758, "distillation_loss": 1.7280023097991943, "epoch": 2.57, "learning_rate": 4.6023794614902946e-05, "loss": 51.8758, "step": 3045, "task_loss": 1.6088672876358032 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4708387829360803, "compression/movement_sparsity/importance_threshold": -0.0034343329628948307, "compression/movement_sparsity/linear_layer_sparsity": 0.4223656556498805, "compression/movement_sparsity/model_sparsity": 0.40785609602957046, "compression_loss": 50.486915588378906, "distillation_loss": 1.004605770111084, "epoch": 2.57, "learning_rate": 4.602066374452098e-05, "loss": 51.4036, "step": 3046, "task_loss": 0.9863617420196533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47139114324331644, "compression/movement_sparsity/importance_threshold": -0.0034307480644756656, "compression/movement_sparsity/linear_layer_sparsity": 0.4232799650516956, "compression/movement_sparsity/model_sparsity": 0.4087389960906884, "compression_loss": 50.54570388793945, "distillation_loss": 1.3414889574050903, "epoch": 2.58, "learning_rate": 4.601753287413902e-05, "loss": 51.473, "step": 3047, "task_loss": 1.3801639080047607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4719431190324288, "compression/movement_sparsity/importance_threshold": -0.0034271656616348327, "compression/movement_sparsity/linear_layer_sparsity": 0.4240613319085384, "compression/movement_sparsity/model_sparsity": 0.4094935205922331, "compression_loss": 50.604454040527344, "distillation_loss": 1.2663168907165527, "epoch": 2.58, "learning_rate": 4.601440200375705e-05, "loss": 51.5074, "step": 3048, "task_loss": 1.840630054473877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4724947104373022, "compression/movement_sparsity/importance_threshold": -0.0034235857535033997, "compression/movement_sparsity/linear_layer_sparsity": 0.42489534396549355, "compression/movement_sparsity/model_sparsity": 0.4102988817693107, "compression_loss": 50.66312026977539, "distillation_loss": 1.275709867477417, "epoch": 2.58, "learning_rate": 4.601127113337508e-05, "loss": 51.8713, "step": 3049, "task_loss": 1.185864806175232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47304591759182224, "compression/movement_sparsity/importance_threshold": -0.0034200083392124294, "compression/movement_sparsity/linear_layer_sparsity": 0.42575223850014193, "compression/movement_sparsity/model_sparsity": 0.41112633934057796, "compression_loss": 50.7217903137207, "distillation_loss": 1.132809042930603, "epoch": 2.58, "learning_rate": 4.600814026299312e-05, "loss": 52.0783, "step": 3050, "task_loss": 1.5769896507263184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4735967406298738, "compression/movement_sparsity/importance_threshold": -0.00341643341789299, "compression/movement_sparsity/linear_layer_sparsity": 0.426576496587971, "compression/movement_sparsity/model_sparsity": 0.4119222816273756, "compression_loss": 50.78036880493164, "distillation_loss": 1.7412362098693848, "epoch": 2.58, "learning_rate": 4.600500939261115e-05, "loss": 52.0238, "step": 3051, "task_loss": 1.0740635395050049 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47414717968534215, "compression/movement_sparsity/importance_threshold": -0.0034128609886761464, "compression/movement_sparsity/linear_layer_sparsity": 0.42737924347738493, "compression/movement_sparsity/model_sparsity": 0.4126974516915999, "compression_loss": 50.838958740234375, "distillation_loss": 1.6543631553649902, "epoch": 2.58, "learning_rate": 4.600187852222918e-05, "loss": 52.0996, "step": 3052, "task_loss": 1.3523019552230835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47469723489211224, "compression/movement_sparsity/importance_threshold": -0.003409291050692966, "compression/movement_sparsity/linear_layer_sparsity": 0.4283323183481842, "compression/movement_sparsity/model_sparsity": 0.4136177855085859, "compression_loss": 50.89751434326172, "distillation_loss": 0.8148986101150513, "epoch": 2.58, "learning_rate": 4.599874765184722e-05, "loss": 51.9025, "step": 3053, "task_loss": 0.4207479953765869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4752469063840694, "compression/movement_sparsity/importance_threshold": -0.0034057236030745133, "compression/movement_sparsity/linear_layer_sparsity": 0.42932759284824673, "compression/movement_sparsity/model_sparsity": 0.4145788692677489, "compression_loss": 50.955997467041016, "distillation_loss": 1.5914688110351562, "epoch": 2.58, "learning_rate": 4.599561678146525e-05, "loss": 52.3315, "step": 3054, "task_loss": 1.6181625127792358 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4757961942950987, "compression/movement_sparsity/importance_threshold": -0.003402158644951856, "compression/movement_sparsity/linear_layer_sparsity": 0.43020112159674717, "compression/movement_sparsity/model_sparsity": 0.41542238961644956, "compression_loss": 51.01445388793945, "distillation_loss": 1.6381553411483765, "epoch": 2.58, "learning_rate": 4.5992485911083284e-05, "loss": 52.6231, "step": 3055, "task_loss": 1.3733323812484741 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47634509875908515, "compression/movement_sparsity/importance_threshold": -0.0033985961754560592, "compression/movement_sparsity/linear_layer_sparsity": 0.4309766694597837, "compression/movement_sparsity/model_sparsity": 0.4161712950245265, "compression_loss": 51.072879791259766, "distillation_loss": 1.2611162662506104, "epoch": 2.58, "learning_rate": 4.5989355040701316e-05, "loss": 52.2681, "step": 3056, "task_loss": 1.1299914121627808 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4768936199099141, "compression/movement_sparsity/importance_threshold": -0.0033950361937181888, "compression/movement_sparsity/linear_layer_sparsity": 0.4318261590863302, "compression/movement_sparsity/model_sparsity": 0.4169916020690654, "compression_loss": 51.13126754760742, "distillation_loss": 1.1304724216461182, "epoch": 2.58, "learning_rate": 4.5986224170319355e-05, "loss": 52.3906, "step": 3057, "task_loss": 0.7407874464988708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4774417578814706, "compression/movement_sparsity/importance_threshold": -0.003391478698869311, "compression/movement_sparsity/linear_layer_sparsity": 0.43262461327539525, "compression/movement_sparsity/model_sparsity": 0.4177626269004036, "compression_loss": 51.189613342285156, "distillation_loss": 1.6101806163787842, "epoch": 2.58, "learning_rate": 4.5983093299937387e-05, "loss": 52.6213, "step": 3058, "task_loss": 1.5085195302963257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47798951280763957, "compression/movement_sparsity/importance_threshold": -0.003387923690040493, "compression/movement_sparsity/linear_layer_sparsity": 0.43337295018788674, "compression/movement_sparsity/model_sparsity": 0.4184852561377974, "compression_loss": 51.24794006347656, "distillation_loss": 1.0015469789505005, "epoch": 2.59, "learning_rate": 4.597996242955542e-05, "loss": 52.5263, "step": 3059, "task_loss": 0.8378332853317261 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4785368848223065, "compression/movement_sparsity/importance_threshold": -0.0033843711663627996, "compression/movement_sparsity/linear_layer_sparsity": 0.4341944895654948, "compression/movement_sparsity/model_sparsity": 0.41927857311043387, "compression_loss": 51.306175231933594, "distillation_loss": 0.6478455662727356, "epoch": 2.59, "learning_rate": 4.597683155917345e-05, "loss": 52.6246, "step": 3060, "task_loss": 1.133970022201538 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4790838740593565, "compression/movement_sparsity/importance_threshold": -0.003380821126967296, "compression/movement_sparsity/linear_layer_sparsity": 0.434921386824577, "compression/movement_sparsity/model_sparsity": 0.4199804992124691, "compression_loss": 51.364410400390625, "distillation_loss": 1.1770353317260742, "epoch": 2.59, "learning_rate": 4.597370068879149e-05, "loss": 52.2459, "step": 3061, "task_loss": 0.5332392454147339 }, { "compression/movement_sparsity/importance_regularization_factor": 0.47963048065267455, "compression/movement_sparsity/importance_threshold": -0.0033772735709850497, "compression/movement_sparsity/linear_layer_sparsity": 0.4355445676735625, "compression/movement_sparsity/model_sparsity": 0.42058227188216335, "compression_loss": 51.42261505126953, "distillation_loss": 1.4087467193603516, "epoch": 2.59, "learning_rate": 4.597056981840952e-05, "loss": 52.8932, "step": 3062, "task_loss": 1.1240863800048828 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48017670473614593, "compression/movement_sparsity/importance_threshold": -0.0033737284975471258, "compression/movement_sparsity/linear_layer_sparsity": 0.43626558631800016, "compression/movement_sparsity/model_sparsity": 0.4212785213180519, "compression_loss": 51.48073196411133, "distillation_loss": 1.2525204420089722, "epoch": 2.59, "learning_rate": 4.596743894802755e-05, "loss": 53.2118, "step": 3063, "task_loss": 2.094228982925415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48072254644365564, "compression/movement_sparsity/importance_threshold": -0.0033701859057845914, "compression/movement_sparsity/linear_layer_sparsity": 0.43717400518100313, "compression/movement_sparsity/model_sparsity": 0.42215573319848737, "compression_loss": 51.53883743286133, "distillation_loss": 1.4521896839141846, "epoch": 2.59, "learning_rate": 4.596430807764559e-05, "loss": 52.7443, "step": 3064, "task_loss": 1.4132658243179321 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4812680059090889, "compression/movement_sparsity/importance_threshold": -0.0033666457948285113, "compression/movement_sparsity/linear_layer_sparsity": 0.438076271173506, "compression/movement_sparsity/model_sparsity": 0.4230270035784529, "compression_loss": 51.596866607666016, "distillation_loss": 2.6263747215270996, "epoch": 2.59, "learning_rate": 4.596117720726362e-05, "loss": 53.1557, "step": 3065, "task_loss": 2.2931597232818604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4818130832663309, "compression/movement_sparsity/importance_threshold": -0.0033631081638099527, "compression/movement_sparsity/linear_layer_sparsity": 0.43885212906490106, "compression/movement_sparsity/model_sparsity": 0.4237762083644605, "compression_loss": 51.65485382080078, "distillation_loss": 0.7742253541946411, "epoch": 2.59, "learning_rate": 4.5958046336881654e-05, "loss": 52.7465, "step": 3066, "task_loss": 0.24861003458499908 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48235777864926654, "compression/movement_sparsity/importance_threshold": -0.0033595730118599816, "compression/movement_sparsity/linear_layer_sparsity": 0.4395888399139476, "compression/movement_sparsity/model_sparsity": 0.4244876109294546, "compression_loss": 51.71284103393555, "distillation_loss": 1.2151256799697876, "epoch": 2.59, "learning_rate": 4.5954915466499686e-05, "loss": 53.1527, "step": 3067, "task_loss": 2.4720165729522705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48290209219178126, "compression/movement_sparsity/importance_threshold": -0.003356040338109663, "compression/movement_sparsity/linear_layer_sparsity": 0.4402817890677705, "compression/movement_sparsity/model_sparsity": 0.42515675514808277, "compression_loss": 51.77070999145508, "distillation_loss": 2.205674886703491, "epoch": 2.59, "learning_rate": 4.5951784596117725e-05, "loss": 53.3044, "step": 3068, "task_loss": 1.635906457901001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4834460240277598, "compression/movement_sparsity/importance_threshold": -0.003352510141690065, "compression/movement_sparsity/linear_layer_sparsity": 0.44098995345949676, "compression/movement_sparsity/model_sparsity": 0.4258405919143847, "compression_loss": 51.828590393066406, "distillation_loss": 1.1123504638671875, "epoch": 2.59, "learning_rate": 4.5948653725735756e-05, "loss": 53.2368, "step": 3069, "task_loss": 0.8958407044410706 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4839895742910879, "compression/movement_sparsity/importance_threshold": -0.0033489824217322506, "compression/movement_sparsity/linear_layer_sparsity": 0.4417366090643516, "compression/movement_sparsity/model_sparsity": 0.4265615976022315, "compression_loss": 51.88640594482422, "distillation_loss": 1.5869202613830566, "epoch": 2.59, "learning_rate": 4.594552285535379e-05, "loss": 53.1995, "step": 3070, "task_loss": 1.3270666599273682 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4845327431156503, "compression/movement_sparsity/importance_threshold": -0.0033454571773672876, "compression/movement_sparsity/linear_layer_sparsity": 0.4424470509720963, "compression/movement_sparsity/model_sparsity": 0.42724763364487023, "compression_loss": 51.944183349609375, "distillation_loss": 1.4062294960021973, "epoch": 2.6, "learning_rate": 4.594239198497182e-05, "loss": 53.4888, "step": 3071, "task_loss": 1.848559856414795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48507553063533226, "compression/movement_sparsity/importance_threshold": -0.003341934407726242, "compression/movement_sparsity/linear_layer_sparsity": 0.4431387242399822, "compression/movement_sparsity/model_sparsity": 0.4279155458081684, "compression_loss": 52.001895904541016, "distillation_loss": 2.3299014568328857, "epoch": 2.6, "learning_rate": 4.593926111458986e-05, "loss": 53.9, "step": 3072, "task_loss": 1.4427857398986816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4856179369840188, "compression/movement_sparsity/importance_threshold": -0.00333841411194018, "compression/movement_sparsity/linear_layer_sparsity": 0.4439005116135669, "compression/movement_sparsity/model_sparsity": 0.4286511634419384, "compression_loss": 52.059600830078125, "distillation_loss": 1.234562873840332, "epoch": 2.6, "learning_rate": 4.593613024420789e-05, "loss": 53.4152, "step": 3073, "task_loss": 0.8873119354248047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4861599622955952, "compression/movement_sparsity/importance_threshold": -0.0033348962891401666, "compression/movement_sparsity/linear_layer_sparsity": 0.4447036400763452, "compression/movement_sparsity/model_sparsity": 0.4294267019713081, "compression_loss": 52.11725997924805, "distillation_loss": 1.2186946868896484, "epoch": 2.6, "learning_rate": 4.593299937382592e-05, "loss": 53.5372, "step": 3074, "task_loss": 1.1538066864013672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48670160670394647, "compression/movement_sparsity/importance_threshold": -0.0033313809384572695, "compression/movement_sparsity/linear_layer_sparsity": 0.44545535152827764, "compression/movement_sparsity/model_sparsity": 0.43015258982233173, "compression_loss": 52.17485809326172, "distillation_loss": 1.948697805404663, "epoch": 2.6, "learning_rate": 4.5929868503443954e-05, "loss": 53.4275, "step": 3075, "task_loss": 1.1946372985839844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4872428703429579, "compression/movement_sparsity/importance_threshold": -0.003327868059022553, "compression/movement_sparsity/linear_layer_sparsity": 0.4463682181058088, "compression/movement_sparsity/model_sparsity": 0.4310340966246186, "compression_loss": 52.23242950439453, "distillation_loss": 1.536691427230835, "epoch": 2.6, "learning_rate": 4.592673763306199e-05, "loss": 53.788, "step": 3076, "task_loss": 1.3213543891906738 }, { "compression/movement_sparsity/importance_regularization_factor": 0.48778375334651447, "compression/movement_sparsity/importance_threshold": -0.0033243576499670845, "compression/movement_sparsity/linear_layer_sparsity": 0.44707974511280835, "compression/movement_sparsity/model_sparsity": 0.4317211804900147, "compression_loss": 52.28993606567383, "distillation_loss": 1.5246739387512207, "epoch": 2.6, "learning_rate": 4.5923606762680024e-05, "loss": 53.9212, "step": 3077, "task_loss": 1.3365530967712402 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4883242558485015, "compression/movement_sparsity/importance_threshold": -0.003320849710421929, "compression/movement_sparsity/linear_layer_sparsity": 0.4477852981118224, "compression/movement_sparsity/model_sparsity": 0.43240249557297766, "compression_loss": 52.34743118286133, "distillation_loss": 0.9668715596199036, "epoch": 2.6, "learning_rate": 4.5920475892298056e-05, "loss": 53.7, "step": 3078, "task_loss": 1.0256965160369873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4888643779828038, "compression/movement_sparsity/importance_threshold": -0.0033173442395181543, "compression/movement_sparsity/linear_layer_sparsity": 0.4485775517337167, "compression/movement_sparsity/model_sparsity": 0.4331675328457027, "compression_loss": 52.404876708984375, "distillation_loss": 0.8775773048400879, "epoch": 2.6, "learning_rate": 4.5917345021916095e-05, "loss": 53.7763, "step": 3079, "task_loss": 0.8935506343841553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4894041198833069, "compression/movement_sparsity/importance_threshold": -0.0033138412363868243, "compression/movement_sparsity/linear_layer_sparsity": 0.44931814986141255, "compression/movement_sparsity/model_sparsity": 0.43388268914936584, "compression_loss": 52.46227264404297, "distillation_loss": 1.9911421537399292, "epoch": 2.6, "learning_rate": 4.5914214151534126e-05, "loss": 53.981, "step": 3080, "task_loss": 1.5246981382369995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4899434816838958, "compression/movement_sparsity/importance_threshold": -0.003310340700159006, "compression/movement_sparsity/linear_layer_sparsity": 0.4500237505570971, "compression/movement_sparsity/model_sparsity": 0.434564050290472, "compression_loss": 52.5196647644043, "distillation_loss": 1.6352965831756592, "epoch": 2.6, "learning_rate": 4.5911083281152165e-05, "loss": 53.6588, "step": 3081, "task_loss": 1.930827260017395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4904824635184556, "compression/movement_sparsity/importance_threshold": -0.0033068426299657653, "compression/movement_sparsity/linear_layer_sparsity": 0.4506157854802177, "compression/movement_sparsity/model_sparsity": 0.435135746992671, "compression_loss": 52.577003479003906, "distillation_loss": 2.798123836517334, "epoch": 2.6, "learning_rate": 4.59079524107702e-05, "loss": 54.1722, "step": 3082, "task_loss": 1.606149673461914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49102106552087144, "compression/movement_sparsity/importance_threshold": -0.003303347024938168, "compression/movement_sparsity/linear_layer_sparsity": 0.45135180472754133, "compression/movement_sparsity/model_sparsity": 0.435846481714589, "compression_loss": 52.63432693481445, "distillation_loss": 1.4533947706222534, "epoch": 2.61, "learning_rate": 4.5904821540388235e-05, "loss": 54.0999, "step": 3083, "task_loss": 0.660954475402832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49155928782502856, "compression/movement_sparsity/importance_threshold": -0.003299853884207281, "compression/movement_sparsity/linear_layer_sparsity": 0.45215693645048244, "compression/movement_sparsity/model_sparsity": 0.4366239546859722, "compression_loss": 52.69159698486328, "distillation_loss": 1.691418170928955, "epoch": 2.61, "learning_rate": 4.590169067000627e-05, "loss": 54.3145, "step": 3084, "task_loss": 1.214664101600647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49209713056481197, "compression/movement_sparsity/importance_threshold": -0.003296363206904169, "compression/movement_sparsity/linear_layer_sparsity": 0.45279957754104966, "compression/movement_sparsity/model_sparsity": 0.43724451907808315, "compression_loss": 52.74886703491211, "distillation_loss": 1.3082987070083618, "epoch": 2.61, "learning_rate": 4.58985597996243e-05, "loss": 54.1067, "step": 3085, "task_loss": 2.009155750274658 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4926345938741069, "compression/movement_sparsity/importance_threshold": -0.0032928749921598997, "compression/movement_sparsity/linear_layer_sparsity": 0.45338178695003833, "compression/movement_sparsity/model_sparsity": 0.4378067278027874, "compression_loss": 52.80603790283203, "distillation_loss": 1.4681005477905273, "epoch": 2.61, "learning_rate": 4.589542892924233e-05, "loss": 54.0182, "step": 3086, "task_loss": 1.066341757774353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4931716778867985, "compression/movement_sparsity/importance_threshold": -0.0032893892391055372, "compression/movement_sparsity/linear_layer_sparsity": 0.45412063222509164, "compression/movement_sparsity/model_sparsity": 0.43852019146968874, "compression_loss": 52.863197326660156, "distillation_loss": 0.7996758818626404, "epoch": 2.61, "learning_rate": 4.589229805886037e-05, "loss": 54.383, "step": 3087, "task_loss": 0.15307585895061493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49370838273677176, "compression/movement_sparsity/importance_threshold": -0.0032859059468721497, "compression/movement_sparsity/linear_layer_sparsity": 0.454865666143148, "compression/movement_sparsity/model_sparsity": 0.43923963118066744, "compression_loss": 52.92028045654297, "distillation_loss": 0.8706608414649963, "epoch": 2.61, "learning_rate": 4.58891671884784e-05, "loss": 54.3212, "step": 3088, "task_loss": 1.713845133781433 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4942447085579118, "compression/movement_sparsity/importance_threshold": -0.003282425114590803, "compression/movement_sparsity/linear_layer_sparsity": 0.4556747924622471, "compression/movement_sparsity/model_sparsity": 0.4400209615215418, "compression_loss": 52.977378845214844, "distillation_loss": 2.0983335971832275, "epoch": 2.61, "learning_rate": 4.588603631809643e-05, "loss": 54.5307, "step": 3089, "task_loss": 1.755318284034729 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4947806554841041, "compression/movement_sparsity/importance_threshold": -0.0032789467413925607, "compression/movement_sparsity/linear_layer_sparsity": 0.4565068251073747, "compression/movement_sparsity/model_sparsity": 0.4408244112856775, "compression_loss": 53.03441619873047, "distillation_loss": 1.5715292692184448, "epoch": 2.61, "learning_rate": 4.588290544771447e-05, "loss": 54.4939, "step": 3090, "task_loss": 1.1951282024383545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49531622364923356, "compression/movement_sparsity/importance_threshold": -0.0032754708264084906, "compression/movement_sparsity/linear_layer_sparsity": 0.4572940586546944, "compression/movement_sparsity/model_sparsity": 0.44158460093883306, "compression_loss": 53.09141159057617, "distillation_loss": 1.8836911916732788, "epoch": 2.61, "learning_rate": 4.58797745773325e-05, "loss": 54.3144, "step": 3091, "task_loss": 0.9058290719985962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4958514131871853, "compression/movement_sparsity/importance_threshold": -0.003271997368769659, "compression/movement_sparsity/linear_layer_sparsity": 0.4580343825265346, "compression/movement_sparsity/model_sparsity": 0.44229949240817296, "compression_loss": 53.1483268737793, "distillation_loss": 1.3034024238586426, "epoch": 2.61, "learning_rate": 4.5876643706950535e-05, "loss": 54.3541, "step": 3092, "task_loss": 0.6587711572647095 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49638622423184464, "compression/movement_sparsity/importance_threshold": -0.003268526367607131, "compression/movement_sparsity/linear_layer_sparsity": 0.4587593838429627, "compression/movement_sparsity/model_sparsity": 0.4429995876990168, "compression_loss": 53.20524215698242, "distillation_loss": 1.620992660522461, "epoch": 2.61, "learning_rate": 4.587351283656857e-05, "loss": 54.5744, "step": 3093, "task_loss": 0.6328001022338867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4969206569170964, "compression/movement_sparsity/importance_threshold": -0.0032650578220519734, "compression/movement_sparsity/linear_layer_sparsity": 0.459540571837291, "compression/movement_sparsity/model_sparsity": 0.44375393948252456, "compression_loss": 53.262142181396484, "distillation_loss": 1.991057276725769, "epoch": 2.61, "learning_rate": 4.5870381966186605e-05, "loss": 54.596, "step": 3094, "task_loss": 2.099888563156128 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49745471137682606, "compression/movement_sparsity/importance_threshold": -0.003261591731235252, "compression/movement_sparsity/linear_layer_sparsity": 0.4602721433700865, "compression/movement_sparsity/model_sparsity": 0.44446037928259124, "compression_loss": 53.318946838378906, "distillation_loss": 1.033545732498169, "epoch": 2.62, "learning_rate": 4.586725109580464e-05, "loss": 54.7118, "step": 3095, "task_loss": 1.25397789478302 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4979883877449186, "compression/movement_sparsity/importance_threshold": -0.0032581280942880326, "compression/movement_sparsity/linear_layer_sparsity": 0.46099015712428, "compression/movement_sparsity/model_sparsity": 0.44515372705545947, "compression_loss": 53.37570571899414, "distillation_loss": 1.2468268871307373, "epoch": 2.62, "learning_rate": 4.586412022542267e-05, "loss": 54.8567, "step": 3096, "task_loss": 1.2881566286087036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49852168615525916, "compression/movement_sparsity/importance_threshold": -0.0032546669103413815, "compression/movement_sparsity/linear_layer_sparsity": 0.4618470278105931, "compression/movement_sparsity/model_sparsity": 0.4459811615976552, "compression_loss": 53.432464599609375, "distillation_loss": 1.7280619144439697, "epoch": 2.62, "learning_rate": 4.58609893550407e-05, "loss": 54.8545, "step": 3097, "task_loss": 1.6398721933364868 }, { "compression/movement_sparsity/importance_regularization_factor": 0.49905460674173285, "compression/movement_sparsity/importance_threshold": -0.003251208178526365, "compression/movement_sparsity/linear_layer_sparsity": 0.46261972579756466, "compression/movement_sparsity/model_sparsity": 0.44672731503167723, "compression_loss": 53.48917770385742, "distillation_loss": 1.3690271377563477, "epoch": 2.62, "learning_rate": 4.585785848465874e-05, "loss": 54.9397, "step": 3098, "task_loss": 1.237884759902954 }, { "compression/movement_sparsity/importance_regularization_factor": 0.4995871496382247, "compression/movement_sparsity/importance_threshold": -0.00324775189797405, "compression/movement_sparsity/linear_layer_sparsity": 0.4635356688103458, "compression/movement_sparsity/model_sparsity": 0.4476117925841991, "compression_loss": 53.545814514160156, "distillation_loss": 2.4796550273895264, "epoch": 2.62, "learning_rate": 4.585472761427677e-05, "loss": 55.3061, "step": 3099, "task_loss": 1.8659636974334717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5001193149786202, "compression/movement_sparsity/importance_threshold": -0.0032442980678155, "compression/movement_sparsity/linear_layer_sparsity": 0.46444886926457074, "compression/movement_sparsity/model_sparsity": 0.44849362179348823, "compression_loss": 53.602413177490234, "distillation_loss": 1.762245535850525, "epoch": 2.62, "learning_rate": 4.58515967438948e-05, "loss": 55.3457, "step": 3100, "task_loss": 0.7994271516799927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5006511028968043, "compression/movement_sparsity/importance_threshold": -0.0032408466871817823, "compression/movement_sparsity/linear_layer_sparsity": 0.4652108951215082, "compression/movement_sparsity/model_sparsity": 0.4492294697179741, "compression_loss": 53.65900802612305, "distillation_loss": 2.1681902408599854, "epoch": 2.62, "learning_rate": 4.584846587351284e-05, "loss": 55.2365, "step": 3101, "task_loss": 1.702002763748169 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5011825135266621, "compression/movement_sparsity/importance_threshold": -0.003237397755203963, "compression/movement_sparsity/linear_layer_sparsity": 0.4659406422566554, "compression/movement_sparsity/model_sparsity": 0.4499341477940642, "compression_loss": 53.71554183959961, "distillation_loss": 1.4499709606170654, "epoch": 2.62, "learning_rate": 4.584533500313087e-05, "loss": 54.8698, "step": 3102, "task_loss": 1.130844235420227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5017135470020788, "compression/movement_sparsity/importance_threshold": -0.0032339512710131083, "compression/movement_sparsity/linear_layer_sparsity": 0.4667729491576386, "compression/movement_sparsity/model_sparsity": 0.45073786239252317, "compression_loss": 53.77202606201172, "distillation_loss": 1.6423511505126953, "epoch": 2.62, "learning_rate": 4.5842204132748905e-05, "loss": 55.2505, "step": 3103, "task_loss": 0.7270593047142029 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5022442034569394, "compression/movement_sparsity/importance_threshold": -0.003230507233740285, "compression/movement_sparsity/linear_layer_sparsity": 0.4674846431029851, "compression/movement_sparsity/model_sparsity": 0.45142510746142034, "compression_loss": 53.82844924926758, "distillation_loss": 2.0480289459228516, "epoch": 2.62, "learning_rate": 4.5839073262366937e-05, "loss": 55.4794, "step": 3104, "task_loss": 1.0429370403289795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5027744830251293, "compression/movement_sparsity/importance_threshold": -0.0032270656425165576, "compression/movement_sparsity/linear_layer_sparsity": 0.4680614627882024, "compression/movement_sparsity/model_sparsity": 0.4519821116159455, "compression_loss": 53.8848876953125, "distillation_loss": 1.413723349571228, "epoch": 2.62, "learning_rate": 4.5835942391984975e-05, "loss": 55.804, "step": 3105, "task_loss": 0.6585744619369507 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5033043858405335, "compression/movement_sparsity/importance_threshold": -0.0032236264964729928, "compression/movement_sparsity/linear_layer_sparsity": 0.468861502891563, "compression/movement_sparsity/model_sparsity": 0.45275466788054436, "compression_loss": 53.9412841796875, "distillation_loss": 1.5433050394058228, "epoch": 2.63, "learning_rate": 4.583281152160301e-05, "loss": 55.3201, "step": 3106, "task_loss": 1.163108229637146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5038339120370371, "compression/movement_sparsity/importance_threshold": -0.0032201897947406566, "compression/movement_sparsity/linear_layer_sparsity": 0.4695863968904823, "compression/movement_sparsity/model_sparsity": 0.45345465954056613, "compression_loss": 53.99762725830078, "distillation_loss": 0.9867916107177734, "epoch": 2.63, "learning_rate": 4.582968065122104e-05, "loss": 55.3756, "step": 3107, "task_loss": 0.8338168859481812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5043630617485253, "compression/movement_sparsity/importance_threshold": -0.003216755536450615, "compression/movement_sparsity/linear_layer_sparsity": 0.47029838894001974, "compression/movement_sparsity/model_sparsity": 0.45414219247285814, "compression_loss": 54.05391311645508, "distillation_loss": 1.3561310768127441, "epoch": 2.63, "learning_rate": 4.582654978083907e-05, "loss": 55.5452, "step": 3108, "task_loss": 0.6082383990287781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5048918351088829, "compression/movement_sparsity/importance_threshold": -0.003213323720733936, "compression/movement_sparsity/linear_layer_sparsity": 0.4709764686568007, "compression/movement_sparsity/model_sparsity": 0.4547969780653505, "compression_loss": 54.11016845703125, "distillation_loss": 1.1709469556808472, "epoch": 2.63, "learning_rate": 4.582341891045711e-05, "loss": 55.6201, "step": 3109, "task_loss": 0.863995373249054 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5054202322519957, "compression/movement_sparsity/importance_threshold": -0.003209894346721682, "compression/movement_sparsity/linear_layer_sparsity": 0.4717967440726394, "compression/movement_sparsity/model_sparsity": 0.45558907449719277, "compression_loss": 54.16639709472656, "distillation_loss": 0.8826963901519775, "epoch": 2.63, "learning_rate": 4.582028804007514e-05, "loss": 55.5378, "step": 3110, "task_loss": 0.9433386921882629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5059482533117485, "compression/movement_sparsity/importance_threshold": -0.0032064674135449207, "compression/movement_sparsity/linear_layer_sparsity": 0.4725345161726055, "compression/movement_sparsity/model_sparsity": 0.4563015018558726, "compression_loss": 54.22260284423828, "distillation_loss": 1.4952242374420166, "epoch": 2.63, "learning_rate": 4.581715716969317e-05, "loss": 55.803, "step": 3111, "task_loss": 0.8440118432044983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5064758984220263, "compression/movement_sparsity/importance_threshold": -0.003203042920334719, "compression/movement_sparsity/linear_layer_sparsity": 0.47324634128379595, "compression/movement_sparsity/model_sparsity": 0.4569888735846635, "compression_loss": 54.278690338134766, "distillation_loss": 0.9158492088317871, "epoch": 2.63, "learning_rate": 4.581402629931121e-05, "loss": 55.8268, "step": 3112, "task_loss": 1.6044416427612305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5070031677167146, "compression/movement_sparsity/importance_threshold": -0.0031996208662221416, "compression/movement_sparsity/linear_layer_sparsity": 0.47399085053847634, "compression/movement_sparsity/model_sparsity": 0.45770780665606725, "compression_loss": 54.334808349609375, "distillation_loss": 1.6467803716659546, "epoch": 2.63, "learning_rate": 4.581089542892924e-05, "loss": 55.9382, "step": 3113, "task_loss": 1.5794326066970825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5075300613296982, "compression/movement_sparsity/importance_threshold": -0.003196201250338255, "compression/movement_sparsity/linear_layer_sparsity": 0.4746537269415216, "compression/movement_sparsity/model_sparsity": 0.45834791121542157, "compression_loss": 54.390872955322266, "distillation_loss": 1.3191208839416504, "epoch": 2.63, "learning_rate": 4.5807764558547275e-05, "loss": 55.7643, "step": 3114, "task_loss": 1.2441669702529907 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5080565793948624, "compression/movement_sparsity/importance_threshold": -0.003192784071814126, "compression/movement_sparsity/linear_layer_sparsity": 0.4753824247499168, "compression/movement_sparsity/model_sparsity": 0.45905157601236174, "compression_loss": 54.4468994140625, "distillation_loss": 1.8815233707427979, "epoch": 2.63, "learning_rate": 4.580463368816531e-05, "loss": 55.9003, "step": 3115, "task_loss": 1.942732810974121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5085827220460923, "compression/movement_sparsity/importance_threshold": -0.0031893693297808195, "compression/movement_sparsity/linear_layer_sparsity": 0.47602873848411587, "compression/movement_sparsity/model_sparsity": 0.4596756868814974, "compression_loss": 54.502906799316406, "distillation_loss": 1.5822970867156982, "epoch": 2.63, "learning_rate": 4.5801502817783345e-05, "loss": 55.9818, "step": 3116, "task_loss": 1.4449480772018433 }, { "compression/movement_sparsity/importance_regularization_factor": 0.509108489417273, "compression/movement_sparsity/importance_threshold": -0.0031859570233694024, "compression/movement_sparsity/linear_layer_sparsity": 0.47694359639764217, "compression/movement_sparsity/model_sparsity": 0.460559116611262, "compression_loss": 54.55882263183594, "distillation_loss": 1.760533094406128, "epoch": 2.63, "learning_rate": 4.5798371947401384e-05, "loss": 56.0086, "step": 3117, "task_loss": 1.8802335262298584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5096338816422896, "compression/movement_sparsity/importance_threshold": -0.003182547151710941, "compression/movement_sparsity/linear_layer_sparsity": 0.4776476588757017, "compression/movement_sparsity/model_sparsity": 0.46123899237725063, "compression_loss": 54.61473083496094, "distillation_loss": 1.7506966590881348, "epoch": 2.64, "learning_rate": 4.5795241077019415e-05, "loss": 56.8412, "step": 3118, "task_loss": 1.9673244953155518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5101588988550273, "compression/movement_sparsity/importance_threshold": -0.003179139713936501, "compression/movement_sparsity/linear_layer_sparsity": 0.4783226263847297, "compression/movement_sparsity/model_sparsity": 0.4618907726759006, "compression_loss": 54.670562744140625, "distillation_loss": 1.2857697010040283, "epoch": 2.64, "learning_rate": 4.579211020663745e-05, "loss": 55.8924, "step": 3119, "task_loss": 0.3296647071838379 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5106835411893714, "compression/movement_sparsity/importance_threshold": -0.0031757347091771475, "compression/movement_sparsity/linear_layer_sparsity": 0.4790101500422783, "compression/movement_sparsity/model_sparsity": 0.4625546777807423, "compression_loss": 54.72636032104492, "distillation_loss": 1.0580989122390747, "epoch": 2.64, "learning_rate": 4.5788979336255486e-05, "loss": 55.8368, "step": 3120, "task_loss": 0.3070012331008911 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5112078087792069, "compression/movement_sparsity/importance_threshold": -0.003172332136563947, "compression/movement_sparsity/linear_layer_sparsity": 0.479781631764151, "compression/movement_sparsity/model_sparsity": 0.46329965673211326, "compression_loss": 54.782135009765625, "distillation_loss": 1.915335774421692, "epoch": 2.64, "learning_rate": 4.578584846587352e-05, "loss": 56.0492, "step": 3121, "task_loss": 0.7394202351570129 }, { "compression/movement_sparsity/importance_regularization_factor": 0.511731701758419, "compression/movement_sparsity/importance_threshold": -0.003168931995227966, "compression/movement_sparsity/linear_layer_sparsity": 0.48052256376854063, "compression/movement_sparsity/model_sparsity": 0.4640151354427786, "compression_loss": 54.837825775146484, "distillation_loss": 2.6138763427734375, "epoch": 2.64, "learning_rate": 4.578271759549155e-05, "loss": 56.9156, "step": 3122, "task_loss": 0.6614902019500732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5122552202608928, "compression/movement_sparsity/importance_threshold": -0.0031655342843002698, "compression/movement_sparsity/linear_layer_sparsity": 0.4812702925484827, "compression/movement_sparsity/model_sparsity": 0.4647371774388469, "compression_loss": 54.8934440612793, "distillation_loss": 2.4833853244781494, "epoch": 2.64, "learning_rate": 4.577958672510958e-05, "loss": 56.6606, "step": 3123, "task_loss": 1.6782410144805908 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5127783644205133, "compression/movement_sparsity/importance_threshold": -0.0031621390029119255, "compression/movement_sparsity/linear_layer_sparsity": 0.48183281515670456, "compression/movement_sparsity/model_sparsity": 0.4652803756649544, "compression_loss": 54.949092864990234, "distillation_loss": 1.297171711921692, "epoch": 2.64, "learning_rate": 4.577645585472762e-05, "loss": 56.3369, "step": 3124, "task_loss": 0.34102511405944824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5133011343711658, "compression/movement_sparsity/importance_threshold": -0.0031587461501939988, "compression/movement_sparsity/linear_layer_sparsity": 0.48262937340311546, "compression/movement_sparsity/model_sparsity": 0.4660495696851013, "compression_loss": 55.00471496582031, "distillation_loss": 1.0936195850372314, "epoch": 2.64, "learning_rate": 4.577332498434565e-05, "loss": 56.5949, "step": 3125, "task_loss": 0.5886197090148926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5138235302467355, "compression/movement_sparsity/importance_threshold": -0.0031553557252775545, "compression/movement_sparsity/linear_layer_sparsity": 0.4833744550178424, "compression/movement_sparsity/model_sparsity": 0.46676905545422315, "compression_loss": 55.060264587402344, "distillation_loss": 1.2534427642822266, "epoch": 2.64, "learning_rate": 4.577019411396368e-05, "loss": 56.6932, "step": 3126, "task_loss": 0.5337661504745483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5143455521811073, "compression/movement_sparsity/importance_threshold": -0.003151967727293661, "compression/movement_sparsity/linear_layer_sparsity": 0.484256294911185, "compression/movement_sparsity/model_sparsity": 0.4676206014343726, "compression_loss": 55.11577224731445, "distillation_loss": 1.262599229812622, "epoch": 2.64, "learning_rate": 4.576706324358172e-05, "loss": 56.2964, "step": 3127, "task_loss": 0.7880988121032715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5148672003081667, "compression/movement_sparsity/importance_threshold": -0.0031485821553733817, "compression/movement_sparsity/linear_layer_sparsity": 0.4850134437835594, "compression/movement_sparsity/model_sparsity": 0.4683517399137186, "compression_loss": 55.171207427978516, "distillation_loss": 1.3007023334503174, "epoch": 2.64, "learning_rate": 4.5763932373199754e-05, "loss": 56.9342, "step": 3128, "task_loss": 1.322317123413086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5153884747617983, "compression/movement_sparsity/importance_threshold": -0.0031451990086477865, "compression/movement_sparsity/linear_layer_sparsity": 0.48576446363376896, "compression/movement_sparsity/model_sparsity": 0.46907695992166615, "compression_loss": 55.226661682128906, "distillation_loss": 0.6841090321540833, "epoch": 2.64, "learning_rate": 4.5760801502817785e-05, "loss": 56.527, "step": 3129, "task_loss": 0.5773928165435791 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5159093756758878, "compression/movement_sparsity/importance_threshold": -0.0031418182862479366, "compression/movement_sparsity/linear_layer_sparsity": 0.48648816521592475, "compression/movement_sparsity/model_sparsity": 0.46977580012810843, "compression_loss": 55.28203582763672, "distillation_loss": 1.7404959201812744, "epoch": 2.65, "learning_rate": 4.575767063243582e-05, "loss": 57.2405, "step": 3130, "task_loss": 1.5675458908081055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5164299031843201, "compression/movement_sparsity/importance_threshold": -0.003138439987304901, "compression/movement_sparsity/linear_layer_sparsity": 0.48725479380156966, "compression/movement_sparsity/model_sparsity": 0.47051609266341105, "compression_loss": 55.33741760253906, "distillation_loss": 1.6505706310272217, "epoch": 2.65, "learning_rate": 4.5754539762053856e-05, "loss": 57.2038, "step": 3131, "task_loss": 0.7763373255729675 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5169500574209804, "compression/movement_sparsity/importance_threshold": -0.003135064110949744, "compression/movement_sparsity/linear_layer_sparsity": 0.4879686460212582, "compression/movement_sparsity/model_sparsity": 0.471205421863287, "compression_loss": 55.39278793334961, "distillation_loss": 1.2994023561477661, "epoch": 2.65, "learning_rate": 4.575140889167189e-05, "loss": 57.1774, "step": 3132, "task_loss": 0.8017194867134094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5174698385197538, "compression/movement_sparsity/importance_threshold": -0.003131690656313534, "compression/movement_sparsity/linear_layer_sparsity": 0.48863836689652645, "compression/movement_sparsity/model_sparsity": 0.4718521357661874, "compression_loss": 55.44806671142578, "distillation_loss": 2.9117164611816406, "epoch": 2.65, "learning_rate": 4.574827802128992e-05, "loss": 57.4458, "step": 3133, "task_loss": 2.2286953926086426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5179892466145253, "compression/movement_sparsity/importance_threshold": -0.0031283196225273357, "compression/movement_sparsity/linear_layer_sparsity": 0.4893332239171711, "compression/movement_sparsity/model_sparsity": 0.47252312231054266, "compression_loss": 55.50334167480469, "distillation_loss": 1.5091426372528076, "epoch": 2.65, "learning_rate": 4.574514715090795e-05, "loss": 57.4859, "step": 3134, "task_loss": 1.2677737474441528 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5185082818391804, "compression/movement_sparsity/importance_threshold": -0.003124951008722214, "compression/movement_sparsity/linear_layer_sparsity": 0.4900904681828866, "compression/movement_sparsity/model_sparsity": 0.473254352906175, "compression_loss": 55.558597564697266, "distillation_loss": 1.03953218460083, "epoch": 2.65, "learning_rate": 4.574201628052599e-05, "loss": 57.4183, "step": 3135, "task_loss": 1.2632111310958862 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5190269443276039, "compression/movement_sparsity/importance_threshold": -0.003121584814029237, "compression/movement_sparsity/linear_layer_sparsity": 0.49067085319422693, "compression/movement_sparsity/model_sparsity": 0.4738147999069027, "compression_loss": 55.613792419433594, "distillation_loss": 0.9954079985618591, "epoch": 2.65, "learning_rate": 4.573888541014402e-05, "loss": 57.2082, "step": 3136, "task_loss": 1.7546303272247314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.519545234213681, "compression/movement_sparsity/importance_threshold": -0.00311822103757947, "compression/movement_sparsity/linear_layer_sparsity": 0.49143491808383016, "compression/movement_sparsity/model_sparsity": 0.47455261681700944, "compression_loss": 55.6689567565918, "distillation_loss": 1.6619186401367188, "epoch": 2.65, "learning_rate": 4.573575453976205e-05, "loss": 57.3074, "step": 3137, "task_loss": 1.6316941976547241 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5200631516312969, "compression/movement_sparsity/importance_threshold": -0.003114859678503979, "compression/movement_sparsity/linear_layer_sparsity": 0.49200789818706864, "compression/movement_sparsity/model_sparsity": 0.47510591329100876, "compression_loss": 55.72410583496094, "distillation_loss": 1.3614782094955444, "epoch": 2.65, "learning_rate": 4.573262366938009e-05, "loss": 57.3677, "step": 3138, "task_loss": 1.9098525047302246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5205806967143366, "compression/movement_sparsity/importance_threshold": -0.003111500735933831, "compression/movement_sparsity/linear_layer_sparsity": 0.4927497006556957, "compression/movement_sparsity/model_sparsity": 0.4758222325627871, "compression_loss": 55.779144287109375, "distillation_loss": 1.563291072845459, "epoch": 2.65, "learning_rate": 4.5729492798998123e-05, "loss": 57.8218, "step": 3139, "task_loss": 1.282370924949646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5210978695966855, "compression/movement_sparsity/importance_threshold": -0.0031081442090000904, "compression/movement_sparsity/linear_layer_sparsity": 0.4932505753172401, "compression/movement_sparsity/model_sparsity": 0.4763059006388368, "compression_loss": 55.83419418334961, "distillation_loss": 1.2283542156219482, "epoch": 2.65, "learning_rate": 4.5726361928616155e-05, "loss": 57.2167, "step": 3140, "task_loss": 2.0165770053863525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5216146704122288, "compression/movement_sparsity/importance_threshold": -0.0031047900968338223, "compression/movement_sparsity/linear_layer_sparsity": 0.49390983869749167, "compression/movement_sparsity/model_sparsity": 0.4769425162938454, "compression_loss": 55.88915252685547, "distillation_loss": 1.1150891780853271, "epoch": 2.65, "learning_rate": 4.572323105823419e-05, "loss": 57.3949, "step": 3141, "task_loss": 0.867617130279541 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5221310992948514, "compression/movement_sparsity/importance_threshold": -0.003101438398566095, "compression/movement_sparsity/linear_layer_sparsity": 0.4946145212322683, "compression/movement_sparsity/model_sparsity": 0.47762299081569537, "compression_loss": 55.944068908691406, "distillation_loss": 1.994683027267456, "epoch": 2.66, "learning_rate": 4.5720100187852226e-05, "loss": 57.5402, "step": 3142, "task_loss": 0.8451590538024902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5226471563784383, "compression/movement_sparsity/importance_threshold": -0.0030980891133279747, "compression/movement_sparsity/linear_layer_sparsity": 0.49526658241526783, "compression/movement_sparsity/model_sparsity": 0.47825265169108405, "compression_loss": 55.99893569946289, "distillation_loss": 1.0859869718551636, "epoch": 2.66, "learning_rate": 4.571696931747026e-05, "loss": 57.5646, "step": 3143, "task_loss": 1.8218578100204468 }, { "compression/movement_sparsity/importance_regularization_factor": 0.523162841796875, "compression/movement_sparsity/importance_threshold": -0.0030947422402505254, "compression/movement_sparsity/linear_layer_sparsity": 0.49598922274650403, "compression/movement_sparsity/model_sparsity": 0.4789504671038406, "compression_loss": 56.053775787353516, "distillation_loss": 1.448288917541504, "epoch": 2.66, "learning_rate": 4.571383844708829e-05, "loss": 57.8134, "step": 3144, "task_loss": 0.7190232276916504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5236781556840464, "compression/movement_sparsity/importance_threshold": -0.0030913977784648152, "compression/movement_sparsity/linear_layer_sparsity": 0.4966838293596283, "compression/movement_sparsity/model_sparsity": 0.47962121184294415, "compression_loss": 56.108577728271484, "distillation_loss": 1.4276745319366455, "epoch": 2.66, "learning_rate": 4.571070757670632e-05, "loss": 58.0955, "step": 3145, "task_loss": 2.399954319000244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5241930981738379, "compression/movement_sparsity/importance_threshold": -0.0030880557271019083, "compression/movement_sparsity/linear_layer_sparsity": 0.49734460710916967, "compression/movement_sparsity/model_sparsity": 0.4802592898439986, "compression_loss": 56.16329574584961, "distillation_loss": 1.5720491409301758, "epoch": 2.66, "learning_rate": 4.570757670632436e-05, "loss": 58.1548, "step": 3146, "task_loss": 0.8407875299453735 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5247076694001342, "compression/movement_sparsity/importance_threshold": -0.0030847160852928725, "compression/movement_sparsity/linear_layer_sparsity": 0.49795421825538555, "compression/movement_sparsity/model_sparsity": 0.48084795897195876, "compression_loss": 56.218040466308594, "distillation_loss": 2.1743671894073486, "epoch": 2.66, "learning_rate": 4.570444583594239e-05, "loss": 57.674, "step": 3147, "task_loss": 1.8997976779937744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5252218694968209, "compression/movement_sparsity/importance_threshold": -0.0030813788521687727, "compression/movement_sparsity/linear_layer_sparsity": 0.4987059177831503, "compression/movement_sparsity/model_sparsity": 0.4815738353084466, "compression_loss": 56.27268981933594, "distillation_loss": 1.742993950843811, "epoch": 2.66, "learning_rate": 4.570131496556043e-05, "loss": 58.4274, "step": 3148, "task_loss": 1.5922030210494995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.525735698597783, "compression/movement_sparsity/importance_threshold": -0.0030780440268606745, "compression/movement_sparsity/linear_layer_sparsity": 0.4994278307401607, "compression/movement_sparsity/model_sparsity": 0.4822709483345197, "compression_loss": 56.32732009887695, "distillation_loss": 2.032843828201294, "epoch": 2.66, "learning_rate": 4.569818409517846e-05, "loss": 58.034, "step": 3149, "task_loss": 1.0167104005813599 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5262491568369052, "compression/movement_sparsity/importance_threshold": -0.0030747116084996473, "compression/movement_sparsity/linear_layer_sparsity": 0.5001428515282776, "compression/movement_sparsity/model_sparsity": 0.4829614059589035, "compression_loss": 56.38195037841797, "distillation_loss": 1.2965011596679688, "epoch": 2.66, "learning_rate": 4.56950532247965e-05, "loss": 57.8847, "step": 3150, "task_loss": 1.049439549446106 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5267622443480733, "compression/movement_sparsity/importance_threshold": -0.003071381596216753, "compression/movement_sparsity/linear_layer_sparsity": 0.5008958507903147, "compression/movement_sparsity/model_sparsity": 0.483688537379793, "compression_loss": 56.43653106689453, "distillation_loss": 1.8541642427444458, "epoch": 2.66, "learning_rate": 4.569192235441453e-05, "loss": 57.9405, "step": 3151, "task_loss": 1.509620189666748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5272749612651723, "compression/movement_sparsity/importance_threshold": -0.003068053989143058, "compression/movement_sparsity/linear_layer_sparsity": 0.5014942055222824, "compression/movement_sparsity/model_sparsity": 0.48426633678596304, "compression_loss": 56.49104309082031, "distillation_loss": 2.761598825454712, "epoch": 2.66, "learning_rate": 4.5688791484032564e-05, "loss": 58.8003, "step": 3152, "task_loss": 1.6670842170715332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.527787307722087, "compression/movement_sparsity/importance_threshold": -0.0030647287864096306, "compression/movement_sparsity/linear_layer_sparsity": 0.5021187576505459, "compression/movement_sparsity/model_sparsity": 0.4848694336272737, "compression_loss": 56.5455207824707, "distillation_loss": 1.3612945079803467, "epoch": 2.66, "learning_rate": 4.56856606136506e-05, "loss": 57.8587, "step": 3153, "task_loss": 0.725012481212616 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5282992838527029, "compression/movement_sparsity/importance_threshold": -0.0030614059871475357, "compression/movement_sparsity/linear_layer_sparsity": 0.5026414296905287, "compression/movement_sparsity/model_sparsity": 0.48537415027475583, "compression_loss": 56.59991455078125, "distillation_loss": 1.7281944751739502, "epoch": 2.67, "learning_rate": 4.5682529743268634e-05, "loss": 58.2771, "step": 3154, "task_loss": 1.4448251724243164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5288108897909048, "compression/movement_sparsity/importance_threshold": -0.0030580855904878395, "compression/movement_sparsity/linear_layer_sparsity": 0.5032653856104106, "compression/movement_sparsity/model_sparsity": 0.48597667138927675, "compression_loss": 56.65431213378906, "distillation_loss": 1.4960108995437622, "epoch": 2.67, "learning_rate": 4.5679398872886666e-05, "loss": 58.559, "step": 3155, "task_loss": 1.732715368270874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5293221256705782, "compression/movement_sparsity/importance_threshold": -0.0030547675955616073, "compression/movement_sparsity/linear_layer_sparsity": 0.5039705451118925, "compression/movement_sparsity/model_sparsity": 0.4866576064925585, "compression_loss": 56.70866775512695, "distillation_loss": 2.222066640853882, "epoch": 2.67, "learning_rate": 4.56762680025047e-05, "loss": 58.9815, "step": 3156, "task_loss": 2.2546300888061523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.529832991625608, "compression/movement_sparsity/importance_threshold": -0.003051452001499906, "compression/movement_sparsity/linear_layer_sparsity": 0.5046213900297932, "compression/movement_sparsity/model_sparsity": 0.4872860928852961, "compression_loss": 56.762977600097656, "distillation_loss": 2.0816359519958496, "epoch": 2.67, "learning_rate": 4.5673137132122736e-05, "loss": 58.1723, "step": 3157, "task_loss": 0.9676244854927063 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5303434877898795, "compression/movement_sparsity/importance_threshold": -0.003048138807433802, "compression/movement_sparsity/linear_layer_sparsity": 0.5052491616833186, "compression/movement_sparsity/model_sparsity": 0.4878922986512713, "compression_loss": 56.8172492980957, "distillation_loss": 1.403087854385376, "epoch": 2.67, "learning_rate": 4.567000626174077e-05, "loss": 58.579, "step": 3158, "task_loss": 1.395765781402588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5308536142972777, "compression/movement_sparsity/importance_threshold": -0.00304482801249436, "compression/movement_sparsity/linear_layer_sparsity": 0.5058653072733988, "compression/movement_sparsity/model_sparsity": 0.48848727774484685, "compression_loss": 56.871490478515625, "distillation_loss": 1.5918793678283691, "epoch": 2.67, "learning_rate": 4.56668753913588e-05, "loss": 58.6714, "step": 3159, "task_loss": 1.3594141006469727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5313633712816875, "compression/movement_sparsity/importance_threshold": -0.003041519615812649, "compression/movement_sparsity/linear_layer_sparsity": 0.5065123006851532, "compression/movement_sparsity/model_sparsity": 0.48911204494252286, "compression_loss": 56.92570114135742, "distillation_loss": 1.9328117370605469, "epoch": 2.67, "learning_rate": 4.566374452097683e-05, "loss": 58.8175, "step": 3160, "task_loss": 2.387741804122925 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5318727588769947, "compression/movement_sparsity/importance_threshold": -0.0030382136165197303, "compression/movement_sparsity/linear_layer_sparsity": 0.5071339671648488, "compression/movement_sparsity/model_sparsity": 0.48971235526617124, "compression_loss": 56.97985076904297, "distillation_loss": 2.7363977432250977, "epoch": 2.67, "learning_rate": 4.566061365059487e-05, "loss": 58.8928, "step": 3161, "task_loss": 1.8601887226104736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5323817772170839, "compression/movement_sparsity/importance_threshold": -0.003034910013746674, "compression/movement_sparsity/linear_layer_sparsity": 0.507881791338132, "compression/movement_sparsity/model_sparsity": 0.49043448937852585, "compression_loss": 57.0339469909668, "distillation_loss": 1.8591666221618652, "epoch": 2.67, "learning_rate": 4.56574827802129e-05, "loss": 58.5528, "step": 3162, "task_loss": 0.6923894286155701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5328904264358405, "compression/movement_sparsity/importance_threshold": -0.0030316088066245435, "compression/movement_sparsity/linear_layer_sparsity": 0.5086007947982393, "compression/movement_sparsity/model_sparsity": 0.4911287928578651, "compression_loss": 57.088016510009766, "distillation_loss": 1.2232180833816528, "epoch": 2.67, "learning_rate": 4.5654351909830934e-05, "loss": 58.5999, "step": 3163, "task_loss": 1.5347849130630493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5333987066671495, "compression/movement_sparsity/importance_threshold": -0.003028309994284407, "compression/movement_sparsity/linear_layer_sparsity": 0.509269156318397, "compression/movement_sparsity/model_sparsity": 0.4917741941036849, "compression_loss": 57.142051696777344, "distillation_loss": 2.0505456924438477, "epoch": 2.67, "learning_rate": 4.565122103944897e-05, "loss": 58.6645, "step": 3164, "task_loss": 2.175471305847168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5339066180448961, "compression/movement_sparsity/importance_threshold": -0.003025013575857329, "compression/movement_sparsity/linear_layer_sparsity": 0.5100045674331712, "compression/movement_sparsity/model_sparsity": 0.4924843415842774, "compression_loss": 57.19607925415039, "distillation_loss": 1.7334367036819458, "epoch": 2.67, "learning_rate": 4.5648090169067004e-05, "loss": 58.8548, "step": 3165, "task_loss": 1.6553179025650024 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5344141607029655, "compression/movement_sparsity/importance_threshold": -0.0030217195504743753, "compression/movement_sparsity/linear_layer_sparsity": 0.5105887920264903, "compression/movement_sparsity/model_sparsity": 0.49304849626553093, "compression_loss": 57.25006103515625, "distillation_loss": 1.932754397392273, "epoch": 2.68, "learning_rate": 4.5644959298685036e-05, "loss": 58.632, "step": 3166, "task_loss": 1.9111816883087158 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5349213347752428, "compression/movement_sparsity/importance_threshold": -0.003018427917266613, "compression/movement_sparsity/linear_layer_sparsity": 0.5112894918892765, "compression/movement_sparsity/model_sparsity": 0.4937251249324255, "compression_loss": 57.303951263427734, "distillation_loss": 2.7239201068878174, "epoch": 2.68, "learning_rate": 4.564182842830307e-05, "loss": 59.1513, "step": 3167, "task_loss": 2.0524888038635254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.535428140395613, "compression/movement_sparsity/importance_threshold": -0.0030151386753651086, "compression/movement_sparsity/linear_layer_sparsity": 0.511959272385383, "compression/movement_sparsity/model_sparsity": 0.4943718964080049, "compression_loss": 57.357818603515625, "distillation_loss": 1.4924352169036865, "epoch": 2.68, "learning_rate": 4.5638697557921106e-05, "loss": 59.0783, "step": 3168, "task_loss": 0.639301598072052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5359345776979615, "compression/movement_sparsity/importance_threshold": -0.003011851823900927, "compression/movement_sparsity/linear_layer_sparsity": 0.5126957685994121, "compression/movement_sparsity/model_sparsity": 0.4950830917113547, "compression_loss": 57.41164779663086, "distillation_loss": 2.033994674682617, "epoch": 2.68, "learning_rate": 4.563556668753914e-05, "loss": 59.1522, "step": 3169, "task_loss": 1.735673189163208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.536440646816173, "compression/movement_sparsity/importance_threshold": -0.0030085673620051357, "compression/movement_sparsity/linear_layer_sparsity": 0.5134534302109948, "compression/movement_sparsity/model_sparsity": 0.4958147253157398, "compression_loss": 57.46541976928711, "distillation_loss": 1.0567972660064697, "epoch": 2.68, "learning_rate": 4.563243581715717e-05, "loss": 58.7426, "step": 3170, "task_loss": 0.803926944732666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5369463478841332, "compression/movement_sparsity/importance_threshold": -0.003005285288808798, "compression/movement_sparsity/linear_layer_sparsity": 0.5142246615253472, "compression/movement_sparsity/model_sparsity": 0.4965594624618591, "compression_loss": 57.51912307739258, "distillation_loss": 1.3474661111831665, "epoch": 2.68, "learning_rate": 4.56293049467752e-05, "loss": 58.8155, "step": 3171, "task_loss": 1.5947743654251099 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5374516810357268, "compression/movement_sparsity/importance_threshold": -0.003002005603442983, "compression/movement_sparsity/linear_layer_sparsity": 0.5149143911539084, "compression/movement_sparsity/model_sparsity": 0.49722549775582275, "compression_loss": 57.57280349731445, "distillation_loss": 0.8849033713340759, "epoch": 2.68, "learning_rate": 4.562617407639324e-05, "loss": 59.1709, "step": 3172, "task_loss": 0.4439680874347687 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5379566464048393, "compression/movement_sparsity/importance_threshold": -0.002998728305038754, "compression/movement_sparsity/linear_layer_sparsity": 0.5156072687627254, "compression/movement_sparsity/model_sparsity": 0.4978945728872361, "compression_loss": 57.626434326171875, "distillation_loss": 1.181388258934021, "epoch": 2.68, "learning_rate": 4.562304320601127e-05, "loss": 59.5859, "step": 3173, "task_loss": 0.4399968087673187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5384612441253557, "compression/movement_sparsity/importance_threshold": -0.0029954533927271784, "compression/movement_sparsity/linear_layer_sparsity": 0.5163801098397086, "compression/movement_sparsity/model_sparsity": 0.4986408644956877, "compression_loss": 57.68001174926758, "distillation_loss": 1.5711721181869507, "epoch": 2.68, "learning_rate": 4.5619912335629304e-05, "loss": 59.2744, "step": 3174, "task_loss": 1.3597255945205688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5389654743311609, "compression/movement_sparsity/importance_threshold": -0.002992180865639323, "compression/movement_sparsity/linear_layer_sparsity": 0.5169704753793603, "compression/movement_sparsity/model_sparsity": 0.4992109491628754, "compression_loss": 57.733585357666016, "distillation_loss": 1.802870273590088, "epoch": 2.68, "learning_rate": 4.561678146524734e-05, "loss": 59.3992, "step": 3175, "task_loss": 1.3098247051239014 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5394693371561404, "compression/movement_sparsity/importance_threshold": -0.0029889107229062522, "compression/movement_sparsity/linear_layer_sparsity": 0.517743590712199, "compression/movement_sparsity/model_sparsity": 0.4999575056056502, "compression_loss": 57.78712463378906, "distillation_loss": 1.9867966175079346, "epoch": 2.68, "learning_rate": 4.5613650594865374e-05, "loss": 59.5746, "step": 3176, "task_loss": 1.59889817237854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5399728327341792, "compression/movement_sparsity/importance_threshold": -0.0029856429636590334, "compression/movement_sparsity/linear_layer_sparsity": 0.518535892030764, "compression/movement_sparsity/model_sparsity": 0.5007225889365186, "compression_loss": 57.84061813354492, "distillation_loss": 2.5011610984802246, "epoch": 2.69, "learning_rate": 4.5610519724483406e-05, "loss": 60.542, "step": 3177, "task_loss": 1.4533343315124512 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5404759611991623, "compression/movement_sparsity/importance_threshold": -0.002982377587028732, "compression/movement_sparsity/linear_layer_sparsity": 0.5193067775442549, "compression/movement_sparsity/model_sparsity": 0.5014669921610998, "compression_loss": 57.89409255981445, "distillation_loss": 2.636124610900879, "epoch": 2.69, "learning_rate": 4.560738885410144e-05, "loss": 59.7093, "step": 3178, "task_loss": 2.1889402866363525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.540978722684975, "compression/movement_sparsity/importance_threshold": -0.0029791145921464145, "compression/movement_sparsity/linear_layer_sparsity": 0.5200200693280645, "compression/movement_sparsity/model_sparsity": 0.5021557801777934, "compression_loss": 57.947547912597656, "distillation_loss": 1.737799882888794, "epoch": 2.69, "learning_rate": 4.5604257983719476e-05, "loss": 59.3658, "step": 3179, "task_loss": 1.4039278030395508 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5414811173255023, "compression/movement_sparsity/importance_threshold": -0.0029758539781431473, "compression/movement_sparsity/linear_layer_sparsity": 0.5207139247186278, "compression/movement_sparsity/model_sparsity": 0.5028257995011419, "compression_loss": 58.00093078613281, "distillation_loss": 2.938216209411621, "epoch": 2.69, "learning_rate": 4.560112711333751e-05, "loss": 59.7839, "step": 3180, "task_loss": 2.086880922317505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5419831452546295, "compression/movement_sparsity/importance_threshold": -0.002972595744149995, "compression/movement_sparsity/linear_layer_sparsity": 0.521341994476344, "compression/movement_sparsity/model_sparsity": 0.503432293130512, "compression_loss": 58.0543212890625, "distillation_loss": 1.638197898864746, "epoch": 2.69, "learning_rate": 4.559799624295554e-05, "loss": 59.945, "step": 3181, "task_loss": 0.5111116766929626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5424848066062418, "compression/movement_sparsity/importance_threshold": -0.002969339889298024, "compression/movement_sparsity/linear_layer_sparsity": 0.5221111390613601, "compression/movement_sparsity/model_sparsity": 0.5041750152328672, "compression_loss": 58.107669830322266, "distillation_loss": 1.663358449935913, "epoch": 2.69, "learning_rate": 4.559486537257358e-05, "loss": 59.861, "step": 3182, "task_loss": 1.9720232486724854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5429861015142242, "compression/movement_sparsity/importance_threshold": -0.0029660864127183007, "compression/movement_sparsity/linear_layer_sparsity": 0.5227069539456212, "compression/movement_sparsity/model_sparsity": 0.5047503620429131, "compression_loss": 58.160919189453125, "distillation_loss": 2.11517333984375, "epoch": 2.69, "learning_rate": 4.559173450219161e-05, "loss": 59.6476, "step": 3183, "task_loss": 1.4201602935791016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5434870301124619, "compression/movement_sparsity/importance_threshold": -0.0029628353135418913, "compression/movement_sparsity/linear_layer_sparsity": 0.5233342367082734, "compression/movement_sparsity/model_sparsity": 0.5053560957129207, "compression_loss": 58.21417999267578, "distillation_loss": 2.7555124759674072, "epoch": 2.69, "learning_rate": 4.558860363180965e-05, "loss": 59.9418, "step": 3184, "task_loss": 2.3349759578704834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5439875925348399, "compression/movement_sparsity/importance_threshold": -0.0029595865908998626, "compression/movement_sparsity/linear_layer_sparsity": 0.5240343284385102, "compression/movement_sparsity/model_sparsity": 0.5060321371384898, "compression_loss": 58.26737976074219, "distillation_loss": 1.720668077468872, "epoch": 2.69, "learning_rate": 4.558547276142768e-05, "loss": 60.0101, "step": 3185, "task_loss": 1.656842589378357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5444877889152435, "compression/movement_sparsity/importance_threshold": -0.0029563402439232783, "compression/movement_sparsity/linear_layer_sparsity": 0.5247672474022486, "compression/movement_sparsity/model_sparsity": 0.5067398780811012, "compression_loss": 58.32054138183594, "distillation_loss": 1.6099553108215332, "epoch": 2.69, "learning_rate": 4.558234189104572e-05, "loss": 59.9403, "step": 3186, "task_loss": 1.1747695207595825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5449876193875578, "compression/movement_sparsity/importance_threshold": -0.0029530962717432066, "compression/movement_sparsity/linear_layer_sparsity": 0.5255157870255899, "compression/movement_sparsity/model_sparsity": 0.5074627030656035, "compression_loss": 58.37366485595703, "distillation_loss": 2.378873348236084, "epoch": 2.69, "learning_rate": 4.557921102066375e-05, "loss": 60.2851, "step": 3187, "task_loss": 1.1357897520065308 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5454870840856679, "compression/movement_sparsity/importance_threshold": -0.002949854673490713, "compression/movement_sparsity/linear_layer_sparsity": 0.5261816683188794, "compression/movement_sparsity/model_sparsity": 0.508105709287978, "compression_loss": 58.426719665527344, "distillation_loss": 1.7443965673446655, "epoch": 2.69, "learning_rate": 4.557608015028178e-05, "loss": 60.2827, "step": 3188, "task_loss": 1.4673521518707275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.545986183143459, "compression/movement_sparsity/importance_threshold": -0.002946615448296863, "compression/movement_sparsity/linear_layer_sparsity": 0.5269972098401665, "compression/movement_sparsity/model_sparsity": 0.5088932344491099, "compression_loss": 58.47976303100586, "distillation_loss": 1.3626937866210938, "epoch": 2.7, "learning_rate": 4.5572949279899814e-05, "loss": 60.6802, "step": 3189, "task_loss": 1.0774041414260864 }, { "compression/movement_sparsity/importance_regularization_factor": 0.546484916694816, "compression/movement_sparsity/importance_threshold": -0.0029433785952927244, "compression/movement_sparsity/linear_layer_sparsity": 0.5277752140817361, "compression/movement_sparsity/model_sparsity": 0.5096445118515605, "compression_loss": 58.53273391723633, "distillation_loss": 1.6590819358825684, "epoch": 2.7, "learning_rate": 4.556981840951785e-05, "loss": 60.1896, "step": 3190, "task_loss": 0.9673817157745361 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5469832848736245, "compression/movement_sparsity/importance_threshold": -0.0029401441136093606, "compression/movement_sparsity/linear_layer_sparsity": 0.5284550466511595, "compression/movement_sparsity/model_sparsity": 0.5103009900808146, "compression_loss": 58.585670471191406, "distillation_loss": 1.1273374557495117, "epoch": 2.7, "learning_rate": 4.5566687539135885e-05, "loss": 60.2601, "step": 3191, "task_loss": 0.6737694144248962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5474812878137694, "compression/movement_sparsity/importance_threshold": -0.0029369120023778387, "compression/movement_sparsity/linear_layer_sparsity": 0.529019775230394, "compression/movement_sparsity/model_sparsity": 0.5108463184960441, "compression_loss": 58.6385498046875, "distillation_loss": 1.1908236742019653, "epoch": 2.7, "learning_rate": 4.5563556668753916e-05, "loss": 60.1372, "step": 3192, "task_loss": 0.5164015293121338 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5479789256491356, "compression/movement_sparsity/importance_threshold": -0.002933682260729226, "compression/movement_sparsity/linear_layer_sparsity": 0.5296407024116963, "compression/movement_sparsity/model_sparsity": 0.5114459149184732, "compression_loss": 58.69143295288086, "distillation_loss": 1.3725593090057373, "epoch": 2.7, "learning_rate": 4.556042579837195e-05, "loss": 60.5575, "step": 3193, "task_loss": 1.62465500831604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5484761985136087, "compression/movement_sparsity/importance_threshold": -0.002930454887794586, "compression/movement_sparsity/linear_layer_sparsity": 0.5303601112935031, "compression/movement_sparsity/model_sparsity": 0.5121406098920295, "compression_loss": 58.74427032470703, "distillation_loss": 1.028214931488037, "epoch": 2.7, "learning_rate": 4.555729492798999e-05, "loss": 60.8568, "step": 3194, "task_loss": 1.0253548622131348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5489731065410736, "compression/movement_sparsity/importance_threshold": -0.002927229882704987, "compression/movement_sparsity/linear_layer_sparsity": 0.5311601037001932, "compression/movement_sparsity/model_sparsity": 0.5129131200984852, "compression_loss": 58.797061920166016, "distillation_loss": 1.8850799798965454, "epoch": 2.7, "learning_rate": 4.555416405760802e-05, "loss": 60.6784, "step": 3195, "task_loss": 1.3054112195968628 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5494696498654155, "compression/movement_sparsity/importance_threshold": -0.002924007244591494, "compression/movement_sparsity/linear_layer_sparsity": 0.5318985197052117, "compression/movement_sparsity/model_sparsity": 0.5136261692420979, "compression_loss": 58.84984588623047, "distillation_loss": 1.9098048210144043, "epoch": 2.7, "learning_rate": 4.555103318722605e-05, "loss": 60.4531, "step": 3196, "task_loss": 0.7044451832771301 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5499658286205193, "compression/movement_sparsity/importance_threshold": -0.0029207869725851733, "compression/movement_sparsity/linear_layer_sparsity": 0.5325825615058106, "compression/movement_sparsity/model_sparsity": 0.5142867121024876, "compression_loss": 58.902591705322266, "distillation_loss": 1.6937456130981445, "epoch": 2.7, "learning_rate": 4.554790231684408e-05, "loss": 60.791, "step": 3197, "task_loss": 0.7342414855957031 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5504616429402704, "compression/movement_sparsity/importance_threshold": -0.0029175690658170914, "compression/movement_sparsity/linear_layer_sparsity": 0.5332653154963048, "compression/movement_sparsity/model_sparsity": 0.5149460113930114, "compression_loss": 58.95527267456055, "distillation_loss": 3.5602824687957764, "epoch": 2.7, "learning_rate": 4.554477144646212e-05, "loss": 61.2161, "step": 3198, "task_loss": 1.8277860879898071 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5509570929585539, "compression/movement_sparsity/importance_threshold": -0.002914353523418314, "compression/movement_sparsity/linear_layer_sparsity": 0.5338885559661285, "compression/movement_sparsity/model_sparsity": 0.5155478416353847, "compression_loss": 59.007930755615234, "distillation_loss": 2.631556987762451, "epoch": 2.7, "learning_rate": 4.554164057608015e-05, "loss": 61.2554, "step": 3199, "task_loss": 1.934686303138733 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5514521788092548, "compression/movement_sparsity/importance_threshold": -0.0029111403445199074, "compression/movement_sparsity/linear_layer_sparsity": 0.5345891127389031, "compression/movement_sparsity/model_sparsity": 0.5162243321278497, "compression_loss": 59.060569763183594, "distillation_loss": 2.0677623748779297, "epoch": 2.7, "learning_rate": 4.5538509705698184e-05, "loss": 60.8322, "step": 3200, "task_loss": 2.3284976482391357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5519469006262585, "compression/movement_sparsity/importance_threshold": -0.0029079295282529363, "compression/movement_sparsity/linear_layer_sparsity": 0.5354168049870113, "compression/movement_sparsity/model_sparsity": 0.5170235906009562, "compression_loss": 59.11317443847656, "distillation_loss": 1.7525142431259155, "epoch": 2.71, "learning_rate": 4.553537883531622e-05, "loss": 60.7834, "step": 3201, "task_loss": 1.2154886722564697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5524412585434499, "compression/movement_sparsity/importance_threshold": -0.002904721073748468, "compression/movement_sparsity/linear_layer_sparsity": 0.5360712271552027, "compression/movement_sparsity/model_sparsity": 0.5176555313544322, "compression_loss": 59.16571044921875, "distillation_loss": 2.2072129249572754, "epoch": 2.71, "learning_rate": 4.5532247964934254e-05, "loss": 61.3189, "step": 3202, "task_loss": 2.370945692062378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5529352526947144, "compression/movement_sparsity/importance_threshold": -0.0029015149801375675, "compression/movement_sparsity/linear_layer_sparsity": 0.5366043328418668, "compression/movement_sparsity/model_sparsity": 0.5181703232207345, "compression_loss": 59.21821594238281, "distillation_loss": 1.5116181373596191, "epoch": 2.71, "learning_rate": 4.5529117094552286e-05, "loss": 61.3715, "step": 3203, "task_loss": 1.3669803142547607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5534288832139369, "compression/movement_sparsity/importance_threshold": -0.0028983112465513026, "compression/movement_sparsity/linear_layer_sparsity": 0.5371084389528406, "compression/movement_sparsity/model_sparsity": 0.5186571117359845, "compression_loss": 59.27071762084961, "distillation_loss": 1.845075249671936, "epoch": 2.71, "learning_rate": 4.552598622417032e-05, "loss": 61.1098, "step": 3204, "task_loss": 1.1053732633590698 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5539221502350027, "compression/movement_sparsity/importance_threshold": -0.0028951098721207375, "compression/movement_sparsity/linear_layer_sparsity": 0.5376104225619751, "compression/movement_sparsity/model_sparsity": 0.5191418506638631, "compression_loss": 59.323150634765625, "distillation_loss": 1.8696494102478027, "epoch": 2.71, "learning_rate": 4.5522855353788357e-05, "loss": 60.802, "step": 3205, "task_loss": 1.0870414972305298 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5544150538917967, "compression/movement_sparsity/importance_threshold": -0.00289191085597694, "compression/movement_sparsity/linear_layer_sparsity": 0.5381150414121573, "compression/movement_sparsity/model_sparsity": 0.5196291343041523, "compression_loss": 59.37552261352539, "distillation_loss": 1.8880031108856201, "epoch": 2.71, "learning_rate": 4.551972448340639e-05, "loss": 61.1355, "step": 3206, "task_loss": 1.2625141143798828 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5549075943182042, "compression/movement_sparsity/importance_threshold": -0.002888714197250975, "compression/movement_sparsity/linear_layer_sparsity": 0.5386884746337659, "compression/movement_sparsity/model_sparsity": 0.5201828683305117, "compression_loss": 59.42783737182617, "distillation_loss": 2.0919203758239746, "epoch": 2.71, "learning_rate": 4.551659361302442e-05, "loss": 61.3216, "step": 3207, "task_loss": 0.8890621662139893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5553997716481105, "compression/movement_sparsity/importance_threshold": -0.002885519895073908, "compression/movement_sparsity/linear_layer_sparsity": 0.5394429167200869, "compression/movement_sparsity/model_sparsity": 0.5209113930102323, "compression_loss": 59.48012924194336, "distillation_loss": 1.3766056299209595, "epoch": 2.71, "learning_rate": 4.551346274264245e-05, "loss": 61.3789, "step": 3208, "task_loss": 1.8640154600143433 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5558915860154006, "compression/movement_sparsity/importance_threshold": -0.0028823279485768057, "compression/movement_sparsity/linear_layer_sparsity": 0.5401862812546743, "compression/movement_sparsity/model_sparsity": 0.5216292206861998, "compression_loss": 59.53236389160156, "distillation_loss": 2.090916633605957, "epoch": 2.71, "learning_rate": 4.551033187226049e-05, "loss": 61.7791, "step": 3209, "task_loss": 2.8921401500701904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5563830375539592, "compression/movement_sparsity/importance_threshold": -0.0028791383568907364, "compression/movement_sparsity/linear_layer_sparsity": 0.540781225674698, "compression/movement_sparsity/model_sparsity": 0.5222037269351326, "compression_loss": 59.58456039428711, "distillation_loss": 1.3313020467758179, "epoch": 2.71, "learning_rate": 4.550720100187852e-05, "loss": 61.1437, "step": 3210, "task_loss": 1.3936184644699097 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5568741263976722, "compression/movement_sparsity/importance_threshold": -0.002875951119146762, "compression/movement_sparsity/linear_layer_sparsity": 0.54144447172694, "compression/movement_sparsity/model_sparsity": 0.5228441884450965, "compression_loss": 59.63673400878906, "distillation_loss": 2.1473522186279297, "epoch": 2.71, "learning_rate": 4.5504070131496554e-05, "loss": 61.8289, "step": 3211, "task_loss": 1.5665907859802246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5573648526804245, "compression/movement_sparsity/importance_threshold": -0.0028727662344759505, "compression/movement_sparsity/linear_layer_sparsity": 0.5421484745841614, "compression/movement_sparsity/model_sparsity": 0.5235240066384063, "compression_loss": 59.688873291015625, "distillation_loss": 2.785092353820801, "epoch": 2.71, "learning_rate": 4.550093926111459e-05, "loss": 62.0144, "step": 3212, "task_loss": 2.286284923553467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.557855216536101, "compression/movement_sparsity/importance_threshold": -0.002869583702009368, "compression/movement_sparsity/linear_layer_sparsity": 0.5428555658008004, "compression/movement_sparsity/model_sparsity": 0.5242068070964867, "compression_loss": 59.7409553527832, "distillation_loss": 3.1469435691833496, "epoch": 2.72, "learning_rate": 4.5497808390732624e-05, "loss": 61.4725, "step": 3213, "task_loss": 1.5217348337173462 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5583452180985871, "compression/movement_sparsity/importance_threshold": -0.0028664035208780807, "compression/movement_sparsity/linear_layer_sparsity": 0.5435181798721577, "compression/movement_sparsity/model_sparsity": 0.5248466583360536, "compression_loss": 59.793006896972656, "distillation_loss": 1.5433502197265625, "epoch": 2.72, "learning_rate": 4.5494677520350656e-05, "loss": 62.0626, "step": 3214, "task_loss": 0.7013033628463745 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5588348575017676, "compression/movement_sparsity/importance_threshold": -0.002863225690213155, "compression/movement_sparsity/linear_layer_sparsity": 0.5442654197612267, "compression/movement_sparsity/model_sparsity": 0.5255682282361542, "compression_loss": 59.84503173828125, "distillation_loss": 3.040944814682007, "epoch": 2.72, "learning_rate": 4.5491546649968695e-05, "loss": 61.6214, "step": 3215, "task_loss": 1.7532404661178589 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5593241348795281, "compression/movement_sparsity/importance_threshold": -0.002860050209145655, "compression/movement_sparsity/linear_layer_sparsity": 0.5449777337632902, "compression/movement_sparsity/model_sparsity": 0.5262560720609127, "compression_loss": 59.897037506103516, "distillation_loss": 2.32039213180542, "epoch": 2.72, "learning_rate": 4.5488415779586726e-05, "loss": 62.1762, "step": 3216, "task_loss": 0.9362318515777588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5598130503657535, "compression/movement_sparsity/importance_threshold": -0.002856877076806649, "compression/movement_sparsity/linear_layer_sparsity": 0.5456036690949996, "compression/movement_sparsity/model_sparsity": 0.5268605045883755, "compression_loss": 59.94902038574219, "distillation_loss": 1.6348111629486084, "epoch": 2.72, "learning_rate": 4.5485284909204765e-05, "loss": 61.5233, "step": 3217, "task_loss": 1.2171554565429688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5603016040943287, "compression/movement_sparsity/importance_threshold": -0.0028537062923272025, "compression/movement_sparsity/linear_layer_sparsity": 0.5463294812548269, "compression/movement_sparsity/model_sparsity": 0.5275613828676534, "compression_loss": 60.00094985961914, "distillation_loss": 2.0153145790100098, "epoch": 2.72, "learning_rate": 4.54821540388228e-05, "loss": 61.8111, "step": 3218, "task_loss": 1.0507100820541382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5607897961991393, "compression/movement_sparsity/importance_threshold": -0.002850537854838381, "compression/movement_sparsity/linear_layer_sparsity": 0.5470102558334936, "compression/movement_sparsity/model_sparsity": 0.5282187707452354, "compression_loss": 60.052860260009766, "distillation_loss": 2.3539702892303467, "epoch": 2.72, "learning_rate": 4.547902316844083e-05, "loss": 62.3723, "step": 3219, "task_loss": 1.7692409753799438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.56127762681407, "compression/movement_sparsity/importance_threshold": -0.002847371763471251, "compression/movement_sparsity/linear_layer_sparsity": 0.5476944884207747, "compression/movement_sparsity/model_sparsity": 0.5288794978381978, "compression_loss": 60.10472106933594, "distillation_loss": 1.8585728406906128, "epoch": 2.72, "learning_rate": 4.547589229805887e-05, "loss": 61.7738, "step": 3220, "task_loss": 1.469277262687683 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5617650960730064, "compression/movement_sparsity/importance_threshold": -0.0028442080173568776, "compression/movement_sparsity/linear_layer_sparsity": 0.5483274232388863, "compression/movement_sparsity/model_sparsity": 0.5294906893981719, "compression_loss": 60.15651321411133, "distillation_loss": 1.9224746227264404, "epoch": 2.72, "learning_rate": 4.54727614276769e-05, "loss": 62.1822, "step": 3221, "task_loss": 1.116389274597168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5622522041098333, "compression/movement_sparsity/importance_threshold": -0.0028410466156263275, "compression/movement_sparsity/linear_layer_sparsity": 0.5489588198393729, "compression/movement_sparsity/model_sparsity": 0.5301003955830287, "compression_loss": 60.208290100097656, "distillation_loss": 1.7739439010620117, "epoch": 2.72, "learning_rate": 4.546963055729493e-05, "loss": 62.1098, "step": 3222, "task_loss": 1.5148887634277344 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5627389510584361, "compression/movement_sparsity/importance_threshold": -0.002837887557410667, "compression/movement_sparsity/linear_layer_sparsity": 0.5496650405917745, "compression/movement_sparsity/model_sparsity": 0.5307823554799962, "compression_loss": 60.26002502441406, "distillation_loss": 2.1653988361358643, "epoch": 2.72, "learning_rate": 4.546649968691297e-05, "loss": 62.3949, "step": 3223, "task_loss": 1.7051787376403809 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5632253370526996, "compression/movement_sparsity/importance_threshold": -0.002834730841840962, "compression/movement_sparsity/linear_layer_sparsity": 0.5502848349771514, "compression/movement_sparsity/model_sparsity": 0.5313808580215248, "compression_loss": 60.31169891357422, "distillation_loss": 2.742476463317871, "epoch": 2.72, "learning_rate": 4.5463368816531e-05, "loss": 62.5156, "step": 3224, "task_loss": 1.7081178426742554 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5637113622265093, "compression/movement_sparsity/importance_threshold": -0.002831576468048278, "compression/movement_sparsity/linear_layer_sparsity": 0.550865243836827, "compression/movement_sparsity/model_sparsity": 0.5319413280513241, "compression_loss": 60.3632926940918, "distillation_loss": 2.6285932064056396, "epoch": 2.73, "learning_rate": 4.546023794614903e-05, "loss": 62.5059, "step": 3225, "task_loss": 1.4834864139556885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5641970267137502, "compression/movement_sparsity/importance_threshold": -0.002828424435163681, "compression/movement_sparsity/linear_layer_sparsity": 0.5514888062591768, "compression/movement_sparsity/model_sparsity": 0.5325434691861638, "compression_loss": 60.41490173339844, "distillation_loss": 1.1217520236968994, "epoch": 2.73, "learning_rate": 4.5457107075767065e-05, "loss": 62.225, "step": 3226, "task_loss": 0.6226232051849365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5646823306483073, "compression/movement_sparsity/importance_threshold": -0.002825274742318238, "compression/movement_sparsity/linear_layer_sparsity": 0.5522911238785558, "compression/movement_sparsity/model_sparsity": 0.5333182247270994, "compression_loss": 60.466468811035156, "distillation_loss": 1.530555009841919, "epoch": 2.73, "learning_rate": 4.54539762053851e-05, "loss": 62.1491, "step": 3227, "task_loss": 0.8091095685958862 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5651672741640659, "compression/movement_sparsity/importance_threshold": -0.0028221273886430147, "compression/movement_sparsity/linear_layer_sparsity": 0.5530125717930285, "compression/movement_sparsity/model_sparsity": 0.5340148886862766, "compression_loss": 60.51799774169922, "distillation_loss": 2.3837292194366455, "epoch": 2.73, "learning_rate": 4.5450845335003135e-05, "loss": 62.9902, "step": 3228, "task_loss": 1.5531892776489258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.565651857394911, "compression/movement_sparsity/importance_threshold": -0.0028189823732690774, "compression/movement_sparsity/linear_layer_sparsity": 0.5536118566100717, "compression/movement_sparsity/model_sparsity": 0.5345935862262385, "compression_loss": 60.56949996948242, "distillation_loss": 1.5771152973175049, "epoch": 2.73, "learning_rate": 4.544771446462117e-05, "loss": 62.2818, "step": 3229, "task_loss": 0.40221303701400757 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5661360804747277, "compression/movement_sparsity/importance_threshold": -0.0028158396953274923, "compression/movement_sparsity/linear_layer_sparsity": 0.5543475777532044, "compression/movement_sparsity/model_sparsity": 0.5353040330847617, "compression_loss": 60.620967864990234, "distillation_loss": 2.158998966217041, "epoch": 2.73, "learning_rate": 4.54445835942392e-05, "loss": 62.499, "step": 3230, "task_loss": 1.5988597869873047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5666199435374015, "compression/movement_sparsity/importance_threshold": -0.002812699353949324, "compression/movement_sparsity/linear_layer_sparsity": 0.5549618512489659, "compression/movement_sparsity/model_sparsity": 0.5358972043962176, "compression_loss": 60.672340393066406, "distillation_loss": 1.1810896396636963, "epoch": 2.73, "learning_rate": 4.544145272385724e-05, "loss": 62.3654, "step": 3231, "task_loss": 1.2383344173431396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5671034467168172, "compression/movement_sparsity/importance_threshold": -0.0028095613482656398, "compression/movement_sparsity/linear_layer_sparsity": 0.5557702978905098, "compression/movement_sparsity/model_sparsity": 0.5366778784085516, "compression_loss": 60.72373580932617, "distillation_loss": 2.3745944499969482, "epoch": 2.73, "learning_rate": 4.543832185347527e-05, "loss": 63.0527, "step": 3232, "task_loss": 1.726595163345337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5675865901468602, "compression/movement_sparsity/importance_threshold": -0.0028064256774075046, "compression/movement_sparsity/linear_layer_sparsity": 0.5564531711226804, "compression/movement_sparsity/model_sparsity": 0.5373372928444333, "compression_loss": 60.77507400512695, "distillation_loss": 2.417064666748047, "epoch": 2.73, "learning_rate": 4.54351909830933e-05, "loss": 62.4877, "step": 3233, "task_loss": 1.3942477703094482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5680693739614154, "compression/movement_sparsity/importance_threshold": -0.002803292340505985, "compression/movement_sparsity/linear_layer_sparsity": 0.5571928987861389, "compression/movement_sparsity/model_sparsity": 0.5380516085869835, "compression_loss": 60.826377868652344, "distillation_loss": 1.9079704284667969, "epoch": 2.73, "learning_rate": 4.543206011271133e-05, "loss": 62.9014, "step": 3234, "task_loss": 0.8225736618041992 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5685517982943682, "compression/movement_sparsity/importance_threshold": -0.0028001613366921475, "compression/movement_sparsity/linear_layer_sparsity": 0.5577619081415546, "compression/movement_sparsity/model_sparsity": 0.5386010707205632, "compression_loss": 60.877655029296875, "distillation_loss": 2.49898362159729, "epoch": 2.73, "learning_rate": 4.542892924232937e-05, "loss": 62.6267, "step": 3235, "task_loss": 1.4038333892822266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5690338632796035, "compression/movement_sparsity/importance_threshold": -0.002797032665097058, "compression/movement_sparsity/linear_layer_sparsity": 0.558330404757762, "compression/movement_sparsity/model_sparsity": 0.5391500377291037, "compression_loss": 60.928871154785156, "distillation_loss": 1.0691715478897095, "epoch": 2.73, "learning_rate": 4.54257983719474e-05, "loss": 62.9354, "step": 3236, "task_loss": 0.4008980989456177 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5695155690510064, "compression/movement_sparsity/importance_threshold": -0.0027939063248517827, "compression/movement_sparsity/linear_layer_sparsity": 0.5588924027026079, "compression/movement_sparsity/model_sparsity": 0.5396927293156362, "compression_loss": 60.980037689208984, "distillation_loss": 3.914008617401123, "epoch": 2.74, "learning_rate": 4.5422667501565435e-05, "loss": 63.5176, "step": 3237, "task_loss": 1.8881949186325073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5699969157424623, "compression/movement_sparsity/importance_threshold": -0.002790782315087387, "compression/movement_sparsity/linear_layer_sparsity": 0.5594673502935062, "compression/movement_sparsity/model_sparsity": 0.5402479256880416, "compression_loss": 61.031185150146484, "distillation_loss": 2.08986759185791, "epoch": 2.74, "learning_rate": 4.541953663118347e-05, "loss": 62.9114, "step": 3238, "task_loss": 1.3363572359085083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5704779034878561, "compression/movement_sparsity/importance_threshold": -0.0027876606349349375, "compression/movement_sparsity/linear_layer_sparsity": 0.5600556291038216, "compression/movement_sparsity/model_sparsity": 0.5408159953114653, "compression_loss": 61.08230972290039, "distillation_loss": 2.161266326904297, "epoch": 2.74, "learning_rate": 4.5416405760801505e-05, "loss": 63.0533, "step": 3239, "task_loss": 0.8706734776496887 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5709585324210729, "compression/movement_sparsity/importance_threshold": -0.002784541283525501, "compression/movement_sparsity/linear_layer_sparsity": 0.5606440748524837, "compression/movement_sparsity/model_sparsity": 0.5413842261383901, "compression_loss": 61.13338088989258, "distillation_loss": 2.07959246635437, "epoch": 2.74, "learning_rate": 4.541327489041954e-05, "loss": 63.4249, "step": 3240, "task_loss": 1.9863539934158325 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5714388026759982, "compression/movement_sparsity/importance_threshold": -0.0027814242599901408, "compression/movement_sparsity/linear_layer_sparsity": 0.56132898711732, "compression/movement_sparsity/model_sparsity": 0.5420456095598927, "compression_loss": 61.18437957763672, "distillation_loss": 1.755380392074585, "epoch": 2.74, "learning_rate": 4.541014402003757e-05, "loss": 62.6967, "step": 3241, "task_loss": 0.3939976990222931 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5719187143865168, "compression/movement_sparsity/importance_threshold": -0.0027783095634599264, "compression/movement_sparsity/linear_layer_sparsity": 0.5620198495418067, "compression/movement_sparsity/model_sparsity": 0.5427127387347568, "compression_loss": 61.23535919189453, "distillation_loss": 1.6719709634780884, "epoch": 2.74, "learning_rate": 4.540701314965561e-05, "loss": 62.5717, "step": 3242, "task_loss": 0.7836682796478271 }, { "compression/movement_sparsity/importance_regularization_factor": 0.572398267686514, "compression/movement_sparsity/importance_threshold": -0.002775197193065921, "compression/movement_sparsity/linear_layer_sparsity": 0.5626101316122848, "compression/movement_sparsity/model_sparsity": 0.543282742800194, "compression_loss": 61.28627395629883, "distillation_loss": 2.090366840362549, "epoch": 2.74, "learning_rate": 4.540388227927364e-05, "loss": 63.273, "step": 3243, "task_loss": 2.559400796890259 }, { "compression/movement_sparsity/importance_regularization_factor": 0.572877462709875, "compression/movement_sparsity/importance_threshold": -0.0027720871479391916, "compression/movement_sparsity/linear_layer_sparsity": 0.5632268257140763, "compression/movement_sparsity/model_sparsity": 0.543878251562416, "compression_loss": 61.337154388427734, "distillation_loss": 2.2300381660461426, "epoch": 2.74, "learning_rate": 4.540075140889167e-05, "loss": 63.5227, "step": 3244, "task_loss": 2.457061767578125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5733562995904848, "compression/movement_sparsity/importance_threshold": -0.002768979427210805, "compression/movement_sparsity/linear_layer_sparsity": 0.5638712077331184, "compression/movement_sparsity/model_sparsity": 0.544500497076753, "compression_loss": 61.38800811767578, "distillation_loss": 2.6272029876708984, "epoch": 2.74, "learning_rate": 4.53976205385097e-05, "loss": 63.5471, "step": 3245, "task_loss": 2.089963912963867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5738347784622286, "compression/movement_sparsity/importance_threshold": -0.002765874030011826, "compression/movement_sparsity/linear_layer_sparsity": 0.5645212060351169, "compression/movement_sparsity/model_sparsity": 0.5451281659374492, "compression_loss": 61.4388427734375, "distillation_loss": 1.8792638778686523, "epoch": 2.74, "learning_rate": 4.539448966812774e-05, "loss": 63.1737, "step": 3246, "task_loss": 1.1348532438278198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5743128994589916, "compression/movement_sparsity/importance_threshold": -0.002762770955473322, "compression/movement_sparsity/linear_layer_sparsity": 0.5650993016062712, "compression/movement_sparsity/model_sparsity": 0.5456864021473044, "compression_loss": 61.48960876464844, "distillation_loss": 2.650062084197998, "epoch": 2.74, "learning_rate": 4.539135879774577e-05, "loss": 63.2367, "step": 3247, "task_loss": 1.9830965995788574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5747906627146587, "compression/movement_sparsity/importance_threshold": -0.002759670202726358, "compression/movement_sparsity/linear_layer_sparsity": 0.5658939519858605, "compression/movement_sparsity/model_sparsity": 0.5464537538417241, "compression_loss": 61.540382385253906, "distillation_loss": 1.1694265604019165, "epoch": 2.75, "learning_rate": 4.5388227927363804e-05, "loss": 63.625, "step": 3248, "task_loss": 1.0222479104995728 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5752680683631153, "compression/movement_sparsity/importance_threshold": -0.0027565717709020004, "compression/movement_sparsity/linear_layer_sparsity": 0.5665567091472293, "compression/movement_sparsity/model_sparsity": 0.5470937432557205, "compression_loss": 61.59110641479492, "distillation_loss": 2.362687587738037, "epoch": 2.75, "learning_rate": 4.538509705698184e-05, "loss": 63.5481, "step": 3249, "task_loss": 1.6770461797714233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5757451165382462, "compression/movement_sparsity/importance_threshold": -0.0027534756591313164, "compression/movement_sparsity/linear_layer_sparsity": 0.5671450714267181, "compression/movement_sparsity/model_sparsity": 0.5476618934808947, "compression_loss": 61.64177703857422, "distillation_loss": 3.0940659046173096, "epoch": 2.75, "learning_rate": 4.5381966186599875e-05, "loss": 63.7289, "step": 3250, "task_loss": 1.8293547630310059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.576221807373937, "compression/movement_sparsity/importance_threshold": -0.00275038186654537, "compression/movement_sparsity/linear_layer_sparsity": 0.567742400680269, "compression/movement_sparsity/model_sparsity": 0.5482387026369865, "compression_loss": 61.692420959472656, "distillation_loss": 1.7292340993881226, "epoch": 2.75, "learning_rate": 4.537883531621791e-05, "loss": 63.6197, "step": 3251, "task_loss": 1.903066635131836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5766981410040727, "compression/movement_sparsity/importance_threshold": -0.002747290392275228, "compression/movement_sparsity/linear_layer_sparsity": 0.5681702278908761, "compression/movement_sparsity/model_sparsity": 0.5486518326667588, "compression_loss": 61.74302673339844, "distillation_loss": 2.62448787689209, "epoch": 2.75, "learning_rate": 4.5375704445835945e-05, "loss": 64.2308, "step": 3252, "task_loss": 3.471588373184204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5771741175625382, "compression/movement_sparsity/importance_threshold": -0.0027442012354519565, "compression/movement_sparsity/linear_layer_sparsity": 0.5687922759439362, "compression/movement_sparsity/model_sparsity": 0.5492525114555525, "compression_loss": 61.793617248535156, "distillation_loss": 1.8836873769760132, "epoch": 2.75, "learning_rate": 4.5372573575453984e-05, "loss": 63.3659, "step": 3253, "task_loss": 1.1602904796600342 }, { "compression/movement_sparsity/importance_regularization_factor": 0.577649737183219, "compression/movement_sparsity/importance_threshold": -0.0027411143952066216, "compression/movement_sparsity/linear_layer_sparsity": 0.569307841180008, "compression/movement_sparsity/model_sparsity": 0.5497503654397011, "compression_loss": 61.844139099121094, "distillation_loss": 3.1148345470428467, "epoch": 2.75, "learning_rate": 4.5369442705072015e-05, "loss": 64.5201, "step": 3254, "task_loss": 2.974893093109131 }, { "compression/movement_sparsity/importance_regularization_factor": 0.578125, "compression/movement_sparsity/importance_threshold": -0.002738029870670289, "compression/movement_sparsity/linear_layer_sparsity": 0.5699367575536263, "compression/movement_sparsity/model_sparsity": 0.5503576766011126, "compression_loss": 61.8946418762207, "distillation_loss": 2.9435007572174072, "epoch": 2.75, "learning_rate": 4.536631183469005e-05, "loss": 63.9468, "step": 3255, "task_loss": 1.6965686082839966 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5785999061467664, "compression/movement_sparsity/importance_threshold": -0.0027349476609740252, "compression/movement_sparsity/linear_layer_sparsity": 0.5705300802868516, "compression/movement_sparsity/model_sparsity": 0.5509306168731773, "compression_loss": 61.945133209228516, "distillation_loss": 1.766937255859375, "epoch": 2.75, "learning_rate": 4.536318096430808e-05, "loss": 63.6339, "step": 3256, "task_loss": 1.7983667850494385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5790744557574032, "compression/movement_sparsity/importance_threshold": -0.002731867765248896, "compression/movement_sparsity/linear_layer_sparsity": 0.5710698635073919, "compression/movement_sparsity/model_sparsity": 0.5514518568795247, "compression_loss": 61.995609283447266, "distillation_loss": 3.585628032684326, "epoch": 2.75, "learning_rate": 4.536005009392612e-05, "loss": 64.84, "step": 3257, "task_loss": 2.5402724742889404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5795486489657957, "compression/movement_sparsity/importance_threshold": -0.0027287901826259686, "compression/movement_sparsity/linear_layer_sparsity": 0.5716247665724944, "compression/movement_sparsity/model_sparsity": 0.5519876973172594, "compression_loss": 62.0460205078125, "distillation_loss": 2.716158866882324, "epoch": 2.75, "learning_rate": 4.535691922354415e-05, "loss": 64.4213, "step": 3258, "task_loss": 2.2517213821411133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5800224859058292, "compression/movement_sparsity/importance_threshold": -0.002725714912236307, "compression/movement_sparsity/linear_layer_sparsity": 0.5723706948031234, "compression/movement_sparsity/model_sparsity": 0.5527080006184227, "compression_loss": 62.09637451171875, "distillation_loss": 2.5347249507904053, "epoch": 2.75, "learning_rate": 4.535378835316218e-05, "loss": 64.2235, "step": 3259, "task_loss": 1.7294872999191284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5804959667113884, "compression/movement_sparsity/importance_threshold": -0.0027226419532109794, "compression/movement_sparsity/linear_layer_sparsity": 0.5728706632279275, "compression/movement_sparsity/model_sparsity": 0.553190793589752, "compression_loss": 62.14668273925781, "distillation_loss": 4.100587368011475, "epoch": 2.76, "learning_rate": 4.535065748278022e-05, "loss": 65.0585, "step": 3260, "task_loss": 1.7983518838882446 }, { "compression/movement_sparsity/importance_regularization_factor": 0.580969091516359, "compression/movement_sparsity/importance_threshold": -0.0027195713046810482, "compression/movement_sparsity/linear_layer_sparsity": 0.5733109273453788, "compression/movement_sparsity/model_sparsity": 0.5536159332803581, "compression_loss": 62.19697189331055, "distillation_loss": 2.922617197036743, "epoch": 2.76, "learning_rate": 4.534752661239825e-05, "loss": 63.8987, "step": 3261, "task_loss": 2.1671738624572754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5814418604546254, "compression/movement_sparsity/importance_threshold": -0.002716502965777585, "compression/movement_sparsity/linear_layer_sparsity": 0.5738399430425439, "compression/movement_sparsity/model_sparsity": 0.5541267756608829, "compression_loss": 62.24715805053711, "distillation_loss": 2.336730480194092, "epoch": 2.76, "learning_rate": 4.534439574201628e-05, "loss": 64.6455, "step": 3262, "task_loss": 2.0633602142333984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5819142736600736, "compression/movement_sparsity/importance_threshold": -0.00271343693563165, "compression/movement_sparsity/linear_layer_sparsity": 0.5743313737633208, "compression/movement_sparsity/model_sparsity": 0.5546013242245832, "compression_loss": 62.29735565185547, "distillation_loss": 1.4821069240570068, "epoch": 2.76, "learning_rate": 4.5341264871634315e-05, "loss": 64.4266, "step": 3263, "task_loss": 0.9498777985572815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.582386331266588, "compression/movement_sparsity/importance_threshold": -0.002710373213374314, "compression/movement_sparsity/linear_layer_sparsity": 0.5749502022911192, "compression/movement_sparsity/model_sparsity": 0.5551988940887125, "compression_loss": 62.347530364990234, "distillation_loss": 2.053527593612671, "epoch": 2.76, "learning_rate": 4.5338134001252354e-05, "loss": 64.4141, "step": 3264, "task_loss": 1.483485460281372 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5828580334080542, "compression/movement_sparsity/importance_threshold": -0.0027073117981366406, "compression/movement_sparsity/linear_layer_sparsity": 0.5754960310646507, "compression/movement_sparsity/model_sparsity": 0.5557259719647076, "compression_loss": 62.39763641357422, "distillation_loss": 3.317925453186035, "epoch": 2.76, "learning_rate": 4.5335003130870385e-05, "loss": 64.9831, "step": 3265, "task_loss": 2.3749139308929443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5833293802183572, "compression/movement_sparsity/importance_threshold": -0.0027042526890496956, "compression/movement_sparsity/linear_layer_sparsity": 0.5761054514241845, "compression/movement_sparsity/model_sparsity": 0.5563144568600951, "compression_loss": 62.44774627685547, "distillation_loss": 2.4049575328826904, "epoch": 2.76, "learning_rate": 4.533187226048842e-05, "loss": 65.0609, "step": 3266, "task_loss": 1.7524023056030273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.583800371831382, "compression/movement_sparsity/importance_threshold": -0.0027011958852445474, "compression/movement_sparsity/linear_layer_sparsity": 0.5767197964649518, "compression/movement_sparsity/model_sparsity": 0.5569076972587655, "compression_loss": 62.49779510498047, "distillation_loss": 2.4823760986328125, "epoch": 2.76, "learning_rate": 4.532874139010645e-05, "loss": 64.472, "step": 3267, "task_loss": 0.8759756684303284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5842710083810142, "compression/movement_sparsity/importance_threshold": -0.002698141385852258, "compression/movement_sparsity/linear_layer_sparsity": 0.5772969619510304, "compression/movement_sparsity/model_sparsity": 0.5574650353348287, "compression_loss": 62.54779052734375, "distillation_loss": 3.127028226852417, "epoch": 2.76, "learning_rate": 4.532561051972449e-05, "loss": 65.2424, "step": 3268, "task_loss": 1.8531819581985474 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5847412900011383, "compression/movement_sparsity/importance_threshold": -0.0026950891900038975, "compression/movement_sparsity/linear_layer_sparsity": 0.5778278974391848, "compression/movement_sparsity/model_sparsity": 0.5579777315556165, "compression_loss": 62.59773635864258, "distillation_loss": 2.9271931648254395, "epoch": 2.76, "learning_rate": 4.532247964934252e-05, "loss": 65.1536, "step": 3269, "task_loss": 2.770268440246582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.58521121682564, "compression/movement_sparsity/importance_threshold": -0.0026920392968305287, "compression/movement_sparsity/linear_layer_sparsity": 0.5783652481295274, "compression/movement_sparsity/model_sparsity": 0.5584966225966617, "compression_loss": 62.64766311645508, "distillation_loss": 1.4534050226211548, "epoch": 2.76, "learning_rate": 4.531934877896055e-05, "loss": 64.2924, "step": 3270, "task_loss": 1.0047919750213623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5856807889884039, "compression/movement_sparsity/importance_threshold": -0.002688991705463221, "compression/movement_sparsity/linear_layer_sparsity": 0.5789765048108769, "compression/movement_sparsity/model_sparsity": 0.5590868807305613, "compression_loss": 62.69752502441406, "distillation_loss": 1.065590500831604, "epoch": 2.76, "learning_rate": 4.531621790857858e-05, "loss": 64.9051, "step": 3271, "task_loss": 1.287510633468628 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5861500066233158, "compression/movement_sparsity/importance_threshold": -0.002685946415033037, "compression/movement_sparsity/linear_layer_sparsity": 0.5796208033607456, "compression/movement_sparsity/model_sparsity": 0.5597090456431478, "compression_loss": 62.747344970703125, "distillation_loss": 2.4845476150512695, "epoch": 2.77, "learning_rate": 4.531308703819662e-05, "loss": 65.0637, "step": 3272, "task_loss": 1.9304927587509155 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5866188698642603, "compression/movement_sparsity/importance_threshold": -0.0026829034246710457, "compression/movement_sparsity/linear_layer_sparsity": 0.580213529885589, "compression/movement_sparsity/model_sparsity": 0.5602814101884228, "compression_loss": 62.79712677001953, "distillation_loss": 1.7986047267913818, "epoch": 2.77, "learning_rate": 4.530995616781465e-05, "loss": 64.7884, "step": 3273, "task_loss": 1.6280940771102905 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5870873788451227, "compression/movement_sparsity/importance_threshold": -0.0026798627335083104, "compression/movement_sparsity/linear_layer_sparsity": 0.580656882362458, "compression/movement_sparsity/model_sparsity": 0.5607095321437996, "compression_loss": 62.84685516357422, "distillation_loss": 2.373429775238037, "epoch": 2.77, "learning_rate": 4.5306825297432685e-05, "loss": 65.159, "step": 3274, "task_loss": 0.9746283888816833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5875555336997882, "compression/movement_sparsity/importance_threshold": -0.0026768243406758998, "compression/movement_sparsity/linear_layer_sparsity": 0.5812906041756336, "compression/movement_sparsity/model_sparsity": 0.5613214836631363, "compression_loss": 62.89656448364258, "distillation_loss": 1.2804510593414307, "epoch": 2.77, "learning_rate": 4.5303694427050724e-05, "loss": 64.8138, "step": 3275, "task_loss": 1.0682469606399536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.588023334562142, "compression/movement_sparsity/importance_threshold": -0.002673788245304878, "compression/movement_sparsity/linear_layer_sparsity": 0.5817467393602083, "compression/movement_sparsity/model_sparsity": 0.5617619492008848, "compression_loss": 62.946224212646484, "distillation_loss": 2.436553955078125, "epoch": 2.77, "learning_rate": 4.5300563556668755e-05, "loss": 65.3179, "step": 3276, "task_loss": 1.62647545337677 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5884907815660689, "compression/movement_sparsity/importance_threshold": -0.002670754446526312, "compression/movement_sparsity/linear_layer_sparsity": 0.5823413379793706, "compression/movement_sparsity/model_sparsity": 0.5623361215282796, "compression_loss": 62.99586486816406, "distillation_loss": 2.4307358264923096, "epoch": 2.77, "learning_rate": 4.529743268628679e-05, "loss": 65.3572, "step": 3277, "task_loss": 0.7289491891860962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5889578748454543, "compression/movement_sparsity/importance_threshold": -0.002667722943471269, "compression/movement_sparsity/linear_layer_sparsity": 0.5828462787820788, "compression/movement_sparsity/model_sparsity": 0.5628237160610353, "compression_loss": 63.045467376708984, "distillation_loss": 1.9870669841766357, "epoch": 2.77, "learning_rate": 4.529430181590482e-05, "loss": 64.9565, "step": 3278, "task_loss": 1.1790046691894531 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5894246145341837, "compression/movement_sparsity/importance_threshold": -0.00266469373527081, "compression/movement_sparsity/linear_layer_sparsity": 0.5832902513156649, "compression/movement_sparsity/model_sparsity": 0.5632524367722734, "compression_loss": 63.095008850097656, "distillation_loss": 3.7120556831359863, "epoch": 2.77, "learning_rate": 4.529117094552286e-05, "loss": 65.7356, "step": 3279, "task_loss": 2.329238176345825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5898910007661413, "compression/movement_sparsity/importance_threshold": -0.002661666821056009, "compression/movement_sparsity/linear_layer_sparsity": 0.5839291839900974, "compression/movement_sparsity/model_sparsity": 0.5638694201437523, "compression_loss": 63.14457702636719, "distillation_loss": 3.090104579925537, "epoch": 2.77, "learning_rate": 4.528804007514089e-05, "loss": 65.1547, "step": 3280, "task_loss": 1.4001250267028809 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5903570336752134, "compression/movement_sparsity/importance_threshold": -0.002658642199957923, "compression/movement_sparsity/linear_layer_sparsity": 0.5843753624946961, "compression/movement_sparsity/model_sparsity": 0.5643002710441124, "compression_loss": 63.19406509399414, "distillation_loss": 2.8316006660461426, "epoch": 2.77, "learning_rate": 4.528490920475892e-05, "loss": 65.7256, "step": 3281, "task_loss": 1.7057539224624634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5908227133952841, "compression/movement_sparsity/importance_threshold": -0.0026556198711076263, "compression/movement_sparsity/linear_layer_sparsity": 0.5849901368054983, "compression/movement_sparsity/model_sparsity": 0.5648939259660716, "compression_loss": 63.243507385253906, "distillation_loss": 2.977006435394287, "epoch": 2.77, "learning_rate": 4.528177833437696e-05, "loss": 65.8295, "step": 3282, "task_loss": 2.558807134628296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5912880400602394, "compression/movement_sparsity/importance_threshold": -0.002652599833636179, "compression/movement_sparsity/linear_layer_sparsity": 0.5856224038702224, "compression/movement_sparsity/model_sparsity": 0.5655044727120413, "compression_loss": 63.292945861816406, "distillation_loss": 1.649764060974121, "epoch": 2.77, "learning_rate": 4.527864746399499e-05, "loss": 65.5731, "step": 3283, "task_loss": 1.8796659708023071 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5917530138039635, "compression/movement_sparsity/importance_threshold": -0.0026495820866746524, "compression/movement_sparsity/linear_layer_sparsity": 0.5862297375004197, "compression/movement_sparsity/model_sparsity": 0.5660909425636647, "compression_loss": 63.342308044433594, "distillation_loss": 2.130422830581665, "epoch": 2.78, "learning_rate": 4.527551659361303e-05, "loss": 65.7734, "step": 3284, "task_loss": 1.1817493438720703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5922176347603425, "compression/movement_sparsity/importance_threshold": -0.002646566629354107, "compression/movement_sparsity/linear_layer_sparsity": 0.5869314270691197, "compression/movement_sparsity/model_sparsity": 0.5667685269370302, "compression_loss": 63.39168930053711, "distillation_loss": 1.6436231136322021, "epoch": 2.78, "learning_rate": 4.527238572323106e-05, "loss": 65.1085, "step": 3285, "task_loss": 0.6923684477806091 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5926819030632609, "compression/movement_sparsity/importance_threshold": -0.002643553460805613, "compression/movement_sparsity/linear_layer_sparsity": 0.5875238793381076, "compression/movement_sparsity/model_sparsity": 0.5673406266479819, "compression_loss": 63.440975189208984, "distillation_loss": 2.9229183197021484, "epoch": 2.78, "learning_rate": 4.5269254852849093e-05, "loss": 65.9576, "step": 3286, "task_loss": 1.6541121006011963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5931458188466039, "compression/movement_sparsity/importance_threshold": -0.0026405425801602355, "compression/movement_sparsity/linear_layer_sparsity": 0.5880890610357122, "compression/movement_sparsity/model_sparsity": 0.5678863926155716, "compression_loss": 63.49025344848633, "distillation_loss": 2.2987334728240967, "epoch": 2.78, "learning_rate": 4.526612398246713e-05, "loss": 65.6349, "step": 3287, "task_loss": 2.3272345066070557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5936093822442572, "compression/movement_sparsity/importance_threshold": -0.002637533986549038, "compression/movement_sparsity/linear_layer_sparsity": 0.5886417223572992, "compression/movement_sparsity/model_sparsity": 0.568420068320577, "compression_loss": 63.53947067260742, "distillation_loss": 2.217540740966797, "epoch": 2.78, "learning_rate": 4.5262993112085164e-05, "loss": 65.6499, "step": 3288, "task_loss": 1.4005820751190186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5940725933901052, "compression/movement_sparsity/importance_threshold": -0.0026345276791030895, "compression/movement_sparsity/linear_layer_sparsity": 0.5894982353185833, "compression/movement_sparsity/model_sparsity": 0.5692471574266988, "compression_loss": 63.58866500854492, "distillation_loss": 1.9773398637771606, "epoch": 2.78, "learning_rate": 4.5259862241703196e-05, "loss": 65.8997, "step": 3289, "task_loss": 0.39578530192375183 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5945354524180337, "compression/movement_sparsity/importance_threshold": -0.0026315236569534536, "compression/movement_sparsity/linear_layer_sparsity": 0.5900816132960002, "compression/movement_sparsity/model_sparsity": 0.5698104945759109, "compression_loss": 63.637813568115234, "distillation_loss": 2.959913730621338, "epoch": 2.78, "learning_rate": 4.5256731371321234e-05, "loss": 65.9394, "step": 3290, "task_loss": 1.998962163925171 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5949979594619271, "compression/movement_sparsity/importance_threshold": -0.0026285219192312007, "compression/movement_sparsity/linear_layer_sparsity": 0.5907004656721339, "compression/movement_sparsity/model_sparsity": 0.5704080874691119, "compression_loss": 63.686885833740234, "distillation_loss": 2.3769102096557617, "epoch": 2.78, "learning_rate": 4.5253600500939266e-05, "loss": 65.6867, "step": 3291, "task_loss": 1.2600765228271484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5954601146556714, "compression/movement_sparsity/importance_threshold": -0.0026255224650673908, "compression/movement_sparsity/linear_layer_sparsity": 0.5911971311025028, "compression/movement_sparsity/model_sparsity": 0.570887690914026, "compression_loss": 63.735931396484375, "distillation_loss": 3.1005611419677734, "epoch": 2.78, "learning_rate": 4.52504696305573e-05, "loss": 65.8246, "step": 3292, "task_loss": 1.7342817783355713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.595921918133151, "compression/movement_sparsity/importance_threshold": -0.0026225252935930953, "compression/movement_sparsity/linear_layer_sparsity": 0.5918477256128831, "compression/movement_sparsity/model_sparsity": 0.571515935501512, "compression_loss": 63.78495788574219, "distillation_loss": 2.2531864643096924, "epoch": 2.78, "learning_rate": 4.524733876017533e-05, "loss": 66.0762, "step": 3293, "task_loss": 0.8364577293395996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5963833700282513, "compression/movement_sparsity/importance_threshold": -0.0026195304039393777, "compression/movement_sparsity/linear_layer_sparsity": 0.5924246287672739, "compression/movement_sparsity/model_sparsity": 0.5720730202577876, "compression_loss": 63.83391571044922, "distillation_loss": 2.14603328704834, "epoch": 2.78, "learning_rate": 4.524420788979337e-05, "loss": 66.3229, "step": 3294, "task_loss": 0.8522926568984985 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5968444704748577, "compression/movement_sparsity/importance_threshold": -0.002616537795237303, "compression/movement_sparsity/linear_layer_sparsity": 0.5930233054517676, "compression/movement_sparsity/model_sparsity": 0.5726511305564241, "compression_loss": 63.88285827636719, "distillation_loss": 2.0327210426330566, "epoch": 2.78, "learning_rate": 4.52410770194114e-05, "loss": 66.3872, "step": 3295, "task_loss": 2.011417865753174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5973052196068549, "compression/movement_sparsity/importance_threshold": -0.00261354746661794, "compression/movement_sparsity/linear_layer_sparsity": 0.5936228406763312, "compression/movement_sparsity/model_sparsity": 0.5732300699016378, "compression_loss": 63.931739807128906, "distillation_loss": 3.3703911304473877, "epoch": 2.79, "learning_rate": 4.523794614902943e-05, "loss": 66.2551, "step": 3296, "task_loss": 2.7907986640930176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5977656175581283, "compression/movement_sparsity/importance_threshold": -0.002610559417212353, "compression/movement_sparsity/linear_layer_sparsity": 0.5943289064145535, "compression/movement_sparsity/model_sparsity": 0.57391188010964, "compression_loss": 63.98060607910156, "distillation_loss": 2.2302701473236084, "epoch": 2.79, "learning_rate": 4.523481527864747e-05, "loss": 66.2026, "step": 3297, "task_loss": 1.9902892112731934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.598225664462563, "compression/movement_sparsity/importance_threshold": -0.0026075736461516085, "compression/movement_sparsity/linear_layer_sparsity": 0.594970820130895, "compression/movement_sparsity/model_sparsity": 0.5745317421150675, "compression_loss": 64.02938842773438, "distillation_loss": 2.882056713104248, "epoch": 2.79, "learning_rate": 4.52316844082655e-05, "loss": 66.5946, "step": 3298, "task_loss": 2.4600772857666016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5986853604540443, "compression/movement_sparsity/importance_threshold": -0.0026045901525667712, "compression/movement_sparsity/linear_layer_sparsity": 0.5955626046464951, "compression/movement_sparsity/model_sparsity": 0.5751031970120148, "compression_loss": 64.07813262939453, "distillation_loss": 1.8837769031524658, "epoch": 2.79, "learning_rate": 4.5228553537883534e-05, "loss": 66.0343, "step": 3299, "task_loss": 0.6849653124809265 }, { "compression/movement_sparsity/importance_regularization_factor": 0.599144705666457, "compression/movement_sparsity/importance_threshold": -0.00260160893558891, "compression/movement_sparsity/linear_layer_sparsity": 0.5961655978796732, "compression/movement_sparsity/model_sparsity": 0.575685475572609, "compression_loss": 64.1268081665039, "distillation_loss": 1.7914607524871826, "epoch": 2.79, "learning_rate": 4.5225422667501566e-05, "loss": 66.4506, "step": 3300, "task_loss": 1.5256916284561157 }, { "compression/movement_sparsity/importance_regularization_factor": 0.5996037002336866, "compression/movement_sparsity/importance_threshold": -0.0025986299943490873, "compression/movement_sparsity/linear_layer_sparsity": 0.5966092603849007, "compression/movement_sparsity/model_sparsity": 0.5761138969059164, "compression_loss": 64.17549133300781, "distillation_loss": 2.083982467651367, "epoch": 2.79, "learning_rate": 4.5222291797119604e-05, "loss": 66.4935, "step": 3301, "task_loss": 1.3360068798065186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6000623442896178, "compression/movement_sparsity/importance_threshold": -0.0025956533279783735, "compression/movement_sparsity/linear_layer_sparsity": 0.5973318410952987, "compression/movement_sparsity/model_sparsity": 0.576811654745994, "compression_loss": 64.22410583496094, "distillation_loss": 2.162996292114258, "epoch": 2.79, "learning_rate": 4.5219160926737636e-05, "loss": 66.7113, "step": 3302, "task_loss": 1.2777782678604126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6005206379681363, "compression/movement_sparsity/importance_threshold": -0.00259267893560783, "compression/movement_sparsity/linear_layer_sparsity": 0.5979451248851464, "compression/movement_sparsity/model_sparsity": 0.5774038703509788, "compression_loss": 64.2726821899414, "distillation_loss": 2.523104667663574, "epoch": 2.79, "learning_rate": 4.521603005635567e-05, "loss": 67.0879, "step": 3303, "task_loss": 3.2947402000427246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6009785814031265, "compression/movement_sparsity/importance_threshold": -0.0025897068163685277, "compression/movement_sparsity/linear_layer_sparsity": 0.5985515522786035, "compression/movement_sparsity/model_sparsity": 0.5779894650978817, "compression_loss": 64.32125091552734, "distillation_loss": 2.3977468013763428, "epoch": 2.79, "learning_rate": 4.52128991859737e-05, "loss": 66.5475, "step": 3304, "task_loss": 1.833799958229065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6014361747284744, "compression/movement_sparsity/importance_threshold": -0.0025867369693915276, "compression/movement_sparsity/linear_layer_sparsity": 0.5991826030782287, "compression/movement_sparsity/model_sparsity": 0.5785988373612004, "compression_loss": 64.3697509765625, "distillation_loss": 3.528712749481201, "epoch": 2.79, "learning_rate": 4.520976831559174e-05, "loss": 66.9637, "step": 3305, "task_loss": 2.0859827995300293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6018934180780646, "compression/movement_sparsity/importance_threshold": -0.002583769393807899, "compression/movement_sparsity/linear_layer_sparsity": 0.5998290599024393, "compression/movement_sparsity/model_sparsity": 0.5792230864047657, "compression_loss": 64.41822052001953, "distillation_loss": 1.7883198261260986, "epoch": 2.79, "learning_rate": 4.520663744520977e-05, "loss": 66.6326, "step": 3306, "task_loss": 0.7495938539505005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6023503115857822, "compression/movement_sparsity/importance_threshold": -0.002580804088748708, "compression/movement_sparsity/linear_layer_sparsity": 0.6003032602009823, "compression/movement_sparsity/model_sparsity": 0.579680996464243, "compression_loss": 64.46662139892578, "distillation_loss": 3.370004892349243, "epoch": 2.79, "learning_rate": 4.52035065748278e-05, "loss": 67.4209, "step": 3307, "task_loss": 2.0598011016845703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6028068553855128, "compression/movement_sparsity/importance_threshold": -0.0025778410533450185, "compression/movement_sparsity/linear_layer_sparsity": 0.6008433057532105, "compression/movement_sparsity/model_sparsity": 0.5802024897903778, "compression_loss": 64.5150375366211, "distillation_loss": 2.510342597961426, "epoch": 2.8, "learning_rate": 4.520037570444583e-05, "loss": 66.8212, "step": 3308, "task_loss": 2.0150063037872314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6032630496111409, "compression/movement_sparsity/importance_threshold": -0.0025748802867279, "compression/movement_sparsity/linear_layer_sparsity": 0.6014445699820813, "compression/movement_sparsity/model_sparsity": 0.5807830987432817, "compression_loss": 64.56340026855469, "distillation_loss": 2.504192352294922, "epoch": 2.8, "learning_rate": 4.519724483406387e-05, "loss": 67.1174, "step": 3309, "task_loss": 1.5429729223251343 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6037188943965524, "compression/movement_sparsity/importance_threshold": -0.002571921788028413, "compression/movement_sparsity/linear_layer_sparsity": 0.6021497294835634, "compression/movement_sparsity/model_sparsity": 0.5814640338465634, "compression_loss": 64.61170959472656, "distillation_loss": 2.65661883354187, "epoch": 2.8, "learning_rate": 4.5194113963681904e-05, "loss": 66.8043, "step": 3310, "task_loss": 2.86460280418396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6041743898756315, "compression/movement_sparsity/importance_threshold": -0.00256896555637763, "compression/movement_sparsity/linear_layer_sparsity": 0.6027260841262427, "compression/movement_sparsity/model_sparsity": 0.5820205889341926, "compression_loss": 64.65999603271484, "distillation_loss": 3.386810779571533, "epoch": 2.8, "learning_rate": 4.5190983093299935e-05, "loss": 67.1826, "step": 3311, "task_loss": 2.2936220169067383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6046295361822644, "compression/movement_sparsity/importance_threshold": -0.002566011590906611, "compression/movement_sparsity/linear_layer_sparsity": 0.6033348844290594, "compression/movement_sparsity/model_sparsity": 0.5826084750737186, "compression_loss": 64.70823669433594, "distillation_loss": 1.8451652526855469, "epoch": 2.8, "learning_rate": 4.5187852222917974e-05, "loss": 66.849, "step": 3312, "task_loss": 1.205755352973938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6050843334503353, "compression/movement_sparsity/importance_threshold": -0.0025630598907464284, "compression/movement_sparsity/linear_layer_sparsity": 0.603959961220699, "compression/movement_sparsity/model_sparsity": 0.5832120785546042, "compression_loss": 64.7564468383789, "distillation_loss": 1.9038958549499512, "epoch": 2.8, "learning_rate": 4.5184721352536006e-05, "loss": 67.2443, "step": 3313, "task_loss": 1.557594656944275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6055387818137299, "compression/movement_sparsity/importance_threshold": -0.0025601104550281428, "compression/movement_sparsity/linear_layer_sparsity": 0.6045320350871971, "compression/movement_sparsity/model_sparsity": 0.5837644999238831, "compression_loss": 64.80458068847656, "distillation_loss": 1.9298832416534424, "epoch": 2.8, "learning_rate": 4.518159048215404e-05, "loss": 67.268, "step": 3314, "task_loss": 1.553122878074646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6059928814063333, "compression/movement_sparsity/importance_threshold": -0.0025571632828828217, "compression/movement_sparsity/linear_layer_sparsity": 0.6050495916592642, "compression/movement_sparsity/model_sparsity": 0.5842642768355094, "compression_loss": 64.85267639160156, "distillation_loss": 1.0555038452148438, "epoch": 2.8, "learning_rate": 4.517845961177207e-05, "loss": 66.6777, "step": 3315, "task_loss": 0.5358624458312988 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6064466323620304, "compression/movement_sparsity/importance_threshold": -0.0025542183734415325, "compression/movement_sparsity/linear_layer_sparsity": 0.6055614484792012, "compression/movement_sparsity/model_sparsity": 0.5847585497990259, "compression_loss": 64.90071105957031, "distillation_loss": 2.7813076972961426, "epoch": 2.8, "learning_rate": 4.517532874139011e-05, "loss": 67.497, "step": 3316, "task_loss": 2.09861159324646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6069000348147066, "compression/movement_sparsity/importance_threshold": -0.00255127572583534, "compression/movement_sparsity/linear_layer_sparsity": 0.606194836415683, "compression/movement_sparsity/model_sparsity": 0.5853701789113603, "compression_loss": 64.94872283935547, "distillation_loss": 2.160360813140869, "epoch": 2.8, "learning_rate": 4.517219787100814e-05, "loss": 67.2143, "step": 3317, "task_loss": 1.20736563205719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6073530888982466, "compression/movement_sparsity/importance_threshold": -0.002548335339195312, "compression/movement_sparsity/linear_layer_sparsity": 0.6069726975672408, "compression/movement_sparsity/model_sparsity": 0.5861213181393814, "compression_loss": 64.99671173095703, "distillation_loss": 1.750588297843933, "epoch": 2.8, "learning_rate": 4.516906700062618e-05, "loss": 67.2046, "step": 3318, "task_loss": 1.5759496688842773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6078057947465363, "compression/movement_sparsity/importance_threshold": -0.0025453972126525106, "compression/movement_sparsity/linear_layer_sparsity": 0.6075584842265203, "compression/movement_sparsity/model_sparsity": 0.586686981224824, "compression_loss": 65.04467010498047, "distillation_loss": 1.5941390991210938, "epoch": 2.81, "learning_rate": 4.516593613024421e-05, "loss": 67.1358, "step": 3319, "task_loss": 0.9084857702255249 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6082581524934598, "compression/movement_sparsity/importance_threshold": -0.0025424613453380084, "compression/movement_sparsity/linear_layer_sparsity": 0.6080843400192618, "compression/movement_sparsity/model_sparsity": 0.5871947722533633, "compression_loss": 65.09259033203125, "distillation_loss": 3.3474481105804443, "epoch": 2.81, "learning_rate": 4.516280525986225e-05, "loss": 67.7757, "step": 3320, "task_loss": 1.278290867805481 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6087101622729034, "compression/movement_sparsity/importance_threshold": -0.0025395277363828633, "compression/movement_sparsity/linear_layer_sparsity": 0.6087106807726708, "compression/movement_sparsity/model_sparsity": 0.5877995962750431, "compression_loss": 65.14047241210938, "distillation_loss": 2.2109224796295166, "epoch": 2.81, "learning_rate": 4.515967438948028e-05, "loss": 67.1704, "step": 3321, "task_loss": 1.1743412017822266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6091618242187513, "compression/movement_sparsity/importance_threshold": -0.002536596384918149, "compression/movement_sparsity/linear_layer_sparsity": 0.609198629636498, "compression/movement_sparsity/model_sparsity": 0.5882707825942914, "compression_loss": 65.1883316040039, "distillation_loss": 1.761135220527649, "epoch": 2.81, "learning_rate": 4.515654351909831e-05, "loss": 67.3614, "step": 3322, "task_loss": 0.6446425318717957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6096131384648894, "compression/movement_sparsity/importance_threshold": -0.002533667290074925, "compression/movement_sparsity/linear_layer_sparsity": 0.6098233129306055, "compression/movement_sparsity/model_sparsity": 0.5888740060954959, "compression_loss": 65.23612213134766, "distillation_loss": 2.108562469482422, "epoch": 2.81, "learning_rate": 4.515341264871635e-05, "loss": 67.5977, "step": 3323, "task_loss": 0.8072481751441956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.610064105145202, "compression/movement_sparsity/importance_threshold": -0.002530740450984263, "compression/movement_sparsity/linear_layer_sparsity": 0.6103211827019057, "compression/movement_sparsity/model_sparsity": 0.5893547725085253, "compression_loss": 65.28392028808594, "distillation_loss": 2.2579593658447266, "epoch": 2.81, "learning_rate": 4.515028177833438e-05, "loss": 67.8732, "step": 3324, "task_loss": 0.8080736398696899 }, { "compression/movement_sparsity/importance_regularization_factor": 0.610514724393575, "compression/movement_sparsity/importance_threshold": -0.0025278158667772247, "compression/movement_sparsity/linear_layer_sparsity": 0.6109300545497282, "compression/movement_sparsity/model_sparsity": 0.5899427277352661, "compression_loss": 65.33168029785156, "distillation_loss": 1.5418342351913452, "epoch": 2.81, "learning_rate": 4.5147150907952414e-05, "loss": 67.5141, "step": 3325, "task_loss": 1.0486868619918823 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6109649963438932, "compression/movement_sparsity/importance_threshold": -0.0025248935365848775, "compression/movement_sparsity/linear_layer_sparsity": 0.6114967029199521, "compression/movement_sparsity/model_sparsity": 0.5904899099907586, "compression_loss": 65.37944793701172, "distillation_loss": 1.6398122310638428, "epoch": 2.81, "learning_rate": 4.5144020037570446e-05, "loss": 67.2822, "step": 3326, "task_loss": 1.0372564792633057 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6114149211300417, "compression/movement_sparsity/importance_threshold": -0.0025219734595382896, "compression/movement_sparsity/linear_layer_sparsity": 0.6121462719519156, "compression/movement_sparsity/model_sparsity": 0.5911171643281662, "compression_loss": 65.4271469116211, "distillation_loss": 2.349372386932373, "epoch": 2.81, "learning_rate": 4.5140889167188485e-05, "loss": 67.9803, "step": 3327, "task_loss": 0.5977610349655151 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6118644988859057, "compression/movement_sparsity/importance_threshold": -0.002519055634768525, "compression/movement_sparsity/linear_layer_sparsity": 0.6126322771764181, "compression/movement_sparsity/model_sparsity": 0.59158647377808, "compression_loss": 65.47483825683594, "distillation_loss": 3.3992466926574707, "epoch": 2.81, "learning_rate": 4.5137758296806516e-05, "loss": 68.0094, "step": 3328, "task_loss": 2.638631820678711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6123137297453702, "compression/movement_sparsity/importance_threshold": -0.0025161400614066506, "compression/movement_sparsity/linear_layer_sparsity": 0.6130956264824126, "compression/movement_sparsity/model_sparsity": 0.5920339056099844, "compression_loss": 65.5224838256836, "distillation_loss": 3.237884044647217, "epoch": 2.81, "learning_rate": 4.513462742642455e-05, "loss": 68.136, "step": 3329, "task_loss": 2.4843153953552246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.612762613842321, "compression/movement_sparsity/importance_threshold": -0.0025132267385837284, "compression/movement_sparsity/linear_layer_sparsity": 0.6135781975466359, "compression/movement_sparsity/model_sparsity": 0.5924998988735893, "compression_loss": 65.5700912475586, "distillation_loss": 2.353095531463623, "epoch": 2.81, "learning_rate": 4.513149655604258e-05, "loss": 68.7172, "step": 3330, "task_loss": 0.9854238033294678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6132111513106422, "compression/movement_sparsity/importance_threshold": -0.002510315665430832, "compression/movement_sparsity/linear_layer_sparsity": 0.6142397026704031, "compression/movement_sparsity/model_sparsity": 0.5931386792613274, "compression_loss": 65.61764526367188, "distillation_loss": 2.156468152999878, "epoch": 2.82, "learning_rate": 4.512836568566062e-05, "loss": 68.0464, "step": 3331, "task_loss": 1.1377259492874146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6136593422842198, "compression/movement_sparsity/importance_threshold": -0.002507406841079021, "compression/movement_sparsity/linear_layer_sparsity": 0.6146253064034116, "compression/movement_sparsity/model_sparsity": 0.5935110363198511, "compression_loss": 65.66515350341797, "distillation_loss": 3.6997361183166504, "epoch": 2.82, "learning_rate": 4.512523481527865e-05, "loss": 67.9997, "step": 3332, "task_loss": 2.392578601837158 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6141071868969384, "compression/movement_sparsity/importance_threshold": -0.0025045002646593645, "compression/movement_sparsity/linear_layer_sparsity": 0.615179458245953, "compression/movement_sparsity/model_sparsity": 0.5940461513418309, "compression_loss": 65.7126235961914, "distillation_loss": 2.4800405502319336, "epoch": 2.82, "learning_rate": 4.512210394489668e-05, "loss": 68.4851, "step": 3333, "task_loss": 1.9120572805404663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6145546852826838, "compression/movement_sparsity/importance_threshold": -0.002501595935302927, "compression/movement_sparsity/linear_layer_sparsity": 0.615916383730017, "compression/movement_sparsity/model_sparsity": 0.5947577611684693, "compression_loss": 65.76007843017578, "distillation_loss": 3.027963638305664, "epoch": 2.82, "learning_rate": 4.511897307451472e-05, "loss": 68.8428, "step": 3334, "task_loss": 1.814387559890747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6150018375753401, "compression/movement_sparsity/importance_threshold": -0.002498693852140777, "compression/movement_sparsity/linear_layer_sparsity": 0.6164737908703231, "compression/movement_sparsity/model_sparsity": 0.5952960196587209, "compression_loss": 65.80747985839844, "distillation_loss": 2.1184163093566895, "epoch": 2.82, "learning_rate": 4.511584220413275e-05, "loss": 68.2002, "step": 3335, "task_loss": 2.2674639225006104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6154486439087934, "compression/movement_sparsity/importance_threshold": -0.002495794014303978, "compression/movement_sparsity/linear_layer_sparsity": 0.6170730399148634, "compression/movement_sparsity/model_sparsity": 0.5958746826550756, "compression_loss": 65.85484313964844, "distillation_loss": 3.841496467590332, "epoch": 2.82, "learning_rate": 4.5112711333750784e-05, "loss": 68.8505, "step": 3336, "task_loss": 1.742767095565796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6158951044169283, "compression/movement_sparsity/importance_threshold": -0.002492896420923597, "compression/movement_sparsity/linear_layer_sparsity": 0.6176232687062526, "compression/movement_sparsity/model_sparsity": 0.5964060093947788, "compression_loss": 65.90219116210938, "distillation_loss": 2.8821778297424316, "epoch": 2.82, "learning_rate": 4.5109580463368816e-05, "loss": 68.4274, "step": 3337, "task_loss": 1.1648426055908203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6163412192336302, "compression/movement_sparsity/importance_threshold": -0.0024900010711307007, "compression/movement_sparsity/linear_layer_sparsity": 0.6182564897043875, "compression/movement_sparsity/model_sparsity": 0.5970174773036121, "compression_loss": 65.94950866699219, "distillation_loss": 3.5515241622924805, "epoch": 2.82, "learning_rate": 4.5106449592986855e-05, "loss": 68.8353, "step": 3338, "task_loss": 2.145876169204712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6167869884927843, "compression/movement_sparsity/importance_threshold": -0.002487107964056352, "compression/movement_sparsity/linear_layer_sparsity": 0.6188055141548455, "compression/movement_sparsity/model_sparsity": 0.5975476410752002, "compression_loss": 65.99678039550781, "distillation_loss": 2.3007700443267822, "epoch": 2.82, "learning_rate": 4.5103318722604886e-05, "loss": 68.3515, "step": 3339, "task_loss": 1.3141257762908936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6172324123282753, "compression/movement_sparsity/importance_threshold": -0.002484217098831623, "compression/movement_sparsity/linear_layer_sparsity": 0.6191195848062065, "compression/movement_sparsity/model_sparsity": 0.5978509224334926, "compression_loss": 66.04401397705078, "distillation_loss": 1.7962944507598877, "epoch": 2.82, "learning_rate": 4.510018785222292e-05, "loss": 67.9048, "step": 3340, "task_loss": 0.967883288860321 }, { "compression/movement_sparsity/importance_regularization_factor": 0.617677490873989, "compression/movement_sparsity/importance_threshold": -0.002481328474587572, "compression/movement_sparsity/linear_layer_sparsity": 0.6196614070594125, "compression/movement_sparsity/model_sparsity": 0.5983741314254607, "compression_loss": 66.0911865234375, "distillation_loss": 1.8133702278137207, "epoch": 2.82, "learning_rate": 4.509705698184095e-05, "loss": 68.4781, "step": 3341, "task_loss": 0.7095984816551208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6181222242638098, "compression/movement_sparsity/importance_threshold": -0.0024784420904552735, "compression/movement_sparsity/linear_layer_sparsity": 0.6202772903178049, "compression/movement_sparsity/model_sparsity": 0.5989688571992487, "compression_loss": 66.13832092285156, "distillation_loss": 2.9005703926086426, "epoch": 2.82, "learning_rate": 4.509392611145899e-05, "loss": 68.4529, "step": 3342, "task_loss": 1.4843125343322754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6185666126316237, "compression/movement_sparsity/importance_threshold": -0.002475557945565785, "compression/movement_sparsity/linear_layer_sparsity": 0.6206990123545824, "compression/movement_sparsity/model_sparsity": 0.5993760917866943, "compression_loss": 66.18545532226562, "distillation_loss": 3.5570082664489746, "epoch": 2.83, "learning_rate": 4.509079524107702e-05, "loss": 68.7283, "step": 3343, "task_loss": 2.6282997131347656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6190106561113149, "compression/movement_sparsity/importance_threshold": -0.00247267603905018, "compression/movement_sparsity/linear_layer_sparsity": 0.621275295452256, "compression/movement_sparsity/model_sparsity": 0.5999325777871086, "compression_loss": 66.23251342773438, "distillation_loss": 2.804609537124634, "epoch": 2.83, "learning_rate": 4.508766437069505e-05, "loss": 68.7345, "step": 3344, "task_loss": 1.1425471305847168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6194543548367692, "compression/movement_sparsity/importance_threshold": -0.0024697963700395197, "compression/movement_sparsity/linear_layer_sparsity": 0.6218356240136329, "compression/movement_sparsity/model_sparsity": 0.6004736573386299, "compression_loss": 66.279541015625, "distillation_loss": 2.8984150886535645, "epoch": 2.83, "learning_rate": 4.5084533500313084e-05, "loss": 69.4035, "step": 3345, "task_loss": 1.5465617179870605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6198977089418716, "compression/movement_sparsity/importance_threshold": -0.002466918937664872, "compression/movement_sparsity/linear_layer_sparsity": 0.6223957737124951, "compression/movement_sparsity/model_sparsity": 0.6010145641721143, "compression_loss": 66.3265151977539, "distillation_loss": 2.524777889251709, "epoch": 2.83, "learning_rate": 4.508140262993112e-05, "loss": 69.0911, "step": 3346, "task_loss": 1.148667573928833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6203407185605072, "compression/movement_sparsity/importance_threshold": -0.0024640437410573023, "compression/movement_sparsity/linear_layer_sparsity": 0.6230155084770339, "compression/movement_sparsity/model_sparsity": 0.6016130091409639, "compression_loss": 66.37348175048828, "distillation_loss": 2.693455696105957, "epoch": 2.83, "learning_rate": 4.5078271759549154e-05, "loss": 68.8731, "step": 3347, "task_loss": 3.222581624984741 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6207833838265611, "compression/movement_sparsity/importance_threshold": -0.0024611707793478765, "compression/movement_sparsity/linear_layer_sparsity": 0.6236262166466722, "compression/movement_sparsity/model_sparsity": 0.6022027376062171, "compression_loss": 66.4203872680664, "distillation_loss": 2.309046745300293, "epoch": 2.83, "learning_rate": 4.5075140889167186e-05, "loss": 68.6051, "step": 3348, "task_loss": 1.9122836589813232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6212257048739183, "compression/movement_sparsity/importance_threshold": -0.0024583000516676633, "compression/movement_sparsity/linear_layer_sparsity": 0.6242127306801775, "compression/movement_sparsity/model_sparsity": 0.6027691030783432, "compression_loss": 66.46729278564453, "distillation_loss": 2.0430917739868164, "epoch": 2.83, "learning_rate": 4.5072010018785224e-05, "loss": 68.9249, "step": 3349, "task_loss": 1.514889121055603 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6216676818364644, "compression/movement_sparsity/importance_threshold": -0.002455431557147724, "compression/movement_sparsity/linear_layer_sparsity": 0.6247061408127818, "compression/movement_sparsity/model_sparsity": 0.6032455630549856, "compression_loss": 66.51416015625, "distillation_loss": 1.4723927974700928, "epoch": 2.83, "learning_rate": 4.5068879148403256e-05, "loss": 69.2952, "step": 3350, "task_loss": 1.7555503845214844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.622109314848084, "compression/movement_sparsity/importance_threshold": -0.0024525652949191285, "compression/movement_sparsity/linear_layer_sparsity": 0.625326185605679, "compression/movement_sparsity/model_sparsity": 0.6038443074017659, "compression_loss": 66.56094360351562, "distillation_loss": 1.8981165885925293, "epoch": 2.83, "learning_rate": 4.5065748278021295e-05, "loss": 68.9917, "step": 3351, "task_loss": 2.696925640106201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6225506040426627, "compression/movement_sparsity/importance_threshold": -0.002449701264112941, "compression/movement_sparsity/linear_layer_sparsity": 0.6258381497431248, "compression/movement_sparsity/model_sparsity": 0.6043386839961045, "compression_loss": 66.60775756835938, "distillation_loss": 2.4088480472564697, "epoch": 2.83, "learning_rate": 4.5062617407639327e-05, "loss": 68.8295, "step": 3352, "task_loss": 1.4518018960952759 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6229915495540852, "compression/movement_sparsity/importance_threshold": -0.002446839463860229, "compression/movement_sparsity/linear_layer_sparsity": 0.6263611079631308, "compression/movement_sparsity/model_sparsity": 0.6048436769924457, "compression_loss": 66.65449523925781, "distillation_loss": 3.2394330501556396, "epoch": 2.83, "learning_rate": 4.505948653725736e-05, "loss": 69.1934, "step": 3353, "task_loss": 2.0733425617218018 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6234321515162371, "compression/movement_sparsity/importance_threshold": -0.0024439798932920564, "compression/movement_sparsity/linear_layer_sparsity": 0.6268783187343364, "compression/movement_sparsity/model_sparsity": 0.6053431199825339, "compression_loss": 66.70115661621094, "distillation_loss": 2.4981770515441895, "epoch": 2.83, "learning_rate": 4.50563556668754e-05, "loss": 69.6173, "step": 3354, "task_loss": 2.0500218868255615 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6238724100630031, "compression/movement_sparsity/importance_threshold": -0.0024411225515394916, "compression/movement_sparsity/linear_layer_sparsity": 0.6272731517730951, "compression/movement_sparsity/model_sparsity": 0.6057243892917626, "compression_loss": 66.74784851074219, "distillation_loss": 1.9185802936553955, "epoch": 2.84, "learning_rate": 4.505322479649343e-05, "loss": 68.9075, "step": 3355, "task_loss": 1.0085738897323608 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6243123253282686, "compression/movement_sparsity/importance_threshold": -0.0024382674377335994, "compression/movement_sparsity/linear_layer_sparsity": 0.6279491924572104, "compression/movement_sparsity/model_sparsity": 0.6063772058986342, "compression_loss": 66.79450988769531, "distillation_loss": 3.268648147583008, "epoch": 2.84, "learning_rate": 4.505009392611146e-05, "loss": 69.5292, "step": 3356, "task_loss": 1.997754454612732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6247518974459187, "compression/movement_sparsity/importance_threshold": -0.0024354145510054456, "compression/movement_sparsity/linear_layer_sparsity": 0.6285711451169294, "compression/movement_sparsity/model_sparsity": 0.6069777925711416, "compression_loss": 66.8410873413086, "distillation_loss": 4.033229351043701, "epoch": 2.84, "learning_rate": 4.50469630557295e-05, "loss": 69.6263, "step": 3357, "task_loss": 2.2072455883026123 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6251911265498385, "compression/movement_sparsity/importance_threshold": -0.0024325638904860968, "compression/movement_sparsity/linear_layer_sparsity": 0.6290387036540992, "compression/movement_sparsity/model_sparsity": 0.6074292890341814, "compression_loss": 66.88763427734375, "distillation_loss": 3.5038914680480957, "epoch": 2.84, "learning_rate": 4.504383218534753e-05, "loss": 69.1605, "step": 3358, "task_loss": 2.593931198120117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6256300127739135, "compression/movement_sparsity/importance_threshold": -0.0024297154553066178, "compression/movement_sparsity/linear_layer_sparsity": 0.629613734714171, "compression/movement_sparsity/model_sparsity": 0.6079845660083373, "compression_loss": 66.93414306640625, "distillation_loss": 1.6820124387741089, "epoch": 2.84, "learning_rate": 4.504070131496556e-05, "loss": 69.2504, "step": 3359, "task_loss": 1.6749732494354248 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6260685562520281, "compression/movement_sparsity/importance_threshold": -0.002426869244598078, "compression/movement_sparsity/linear_layer_sparsity": 0.6301960275923332, "compression/movement_sparsity/model_sparsity": 0.6085468553347921, "compression_loss": 66.98056030273438, "distillation_loss": 2.62892484664917, "epoch": 2.84, "learning_rate": 4.50375704445836e-05, "loss": 69.6461, "step": 3360, "task_loss": 1.5881664752960205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6265067571180682, "compression/movement_sparsity/importance_threshold": -0.0024240252574915374, "compression/movement_sparsity/linear_layer_sparsity": 0.6307162909504535, "compression/movement_sparsity/model_sparsity": 0.6090492460460437, "compression_loss": 67.0269775390625, "distillation_loss": 3.1931281089782715, "epoch": 2.84, "learning_rate": 4.503443957420163e-05, "loss": 70.0843, "step": 3361, "task_loss": 1.3010594844818115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6269446155059183, "compression/movement_sparsity/importance_threshold": -0.0024211834931180693, "compression/movement_sparsity/linear_layer_sparsity": 0.6312275873345117, "compression/movement_sparsity/model_sparsity": 0.6095429778263778, "compression_loss": 67.07335662841797, "distillation_loss": 2.3143818378448486, "epoch": 2.84, "learning_rate": 4.5031308703819665e-05, "loss": 69.5785, "step": 3362, "task_loss": 1.0440067052841187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6273821315494641, "compression/movement_sparsity/importance_threshold": -0.002418343950608734, "compression/movement_sparsity/linear_layer_sparsity": 0.6318411573043826, "compression/movement_sparsity/model_sparsity": 0.6101354697802217, "compression_loss": 67.11967468261719, "distillation_loss": 1.7654224634170532, "epoch": 2.84, "learning_rate": 4.5028177833437696e-05, "loss": 70.3491, "step": 3363, "task_loss": 0.8969709873199463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6278193053825902, "compression/movement_sparsity/importance_threshold": -0.0024155066290946006, "compression/movement_sparsity/linear_layer_sparsity": 0.632386568732047, "compression/movement_sparsity/model_sparsity": 0.610662144647464, "compression_loss": 67.1659927368164, "distillation_loss": 3.959557294845581, "epoch": 2.84, "learning_rate": 4.5025046963055735e-05, "loss": 70.177, "step": 3364, "task_loss": 2.481288433074951 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6282561371391822, "compression/movement_sparsity/importance_threshold": -0.002412671527706735, "compression/movement_sparsity/linear_layer_sparsity": 0.632972367315494, "compression/movement_sparsity/model_sparsity": 0.6112278192474424, "compression_loss": 67.21223449707031, "distillation_loss": 2.198829174041748, "epoch": 2.84, "learning_rate": 4.502191609267377e-05, "loss": 70.1771, "step": 3365, "task_loss": 2.3124141693115234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.628692626953125, "compression/movement_sparsity/importance_threshold": -0.002409838645576201, "compression/movement_sparsity/linear_layer_sparsity": 0.6335552683262056, "compression/movement_sparsity/model_sparsity": 0.6117906958152227, "compression_loss": 67.25849151611328, "distillation_loss": 2.8211686611175537, "epoch": 2.84, "learning_rate": 4.50187852222918e-05, "loss": 69.625, "step": 3366, "task_loss": 2.5853967666625977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6291287749583039, "compression/movement_sparsity/importance_threshold": -0.0024070079818340664, "compression/movement_sparsity/linear_layer_sparsity": 0.6339928137334359, "compression/movement_sparsity/model_sparsity": 0.6122132101916676, "compression_loss": 67.30467224121094, "distillation_loss": 2.827392816543579, "epoch": 2.85, "learning_rate": 4.501565435190983e-05, "loss": 70.7641, "step": 3367, "task_loss": 1.5864144563674927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6295645812886038, "compression/movement_sparsity/importance_threshold": -0.0024041795356113975, "compression/movement_sparsity/linear_layer_sparsity": 0.6344391472522138, "compression/movement_sparsity/model_sparsity": 0.6126442107809931, "compression_loss": 67.35089874267578, "distillation_loss": 1.929840326309204, "epoch": 2.85, "learning_rate": 4.501252348152787e-05, "loss": 70.4176, "step": 3368, "task_loss": 1.6096135377883911 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6300000460779098, "compression/movement_sparsity/importance_threshold": -0.002401353306039261, "compression/movement_sparsity/linear_layer_sparsity": 0.6350846501430137, "compression/movement_sparsity/model_sparsity": 0.6132675386616947, "compression_loss": 67.39704895019531, "distillation_loss": 1.6094059944152832, "epoch": 2.85, "learning_rate": 4.50093926111459e-05, "loss": 70.1524, "step": 3369, "task_loss": 1.0032832622528076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6304351694601077, "compression/movement_sparsity/importance_threshold": -0.002398529292248719, "compression/movement_sparsity/linear_layer_sparsity": 0.635456672249085, "compression/movement_sparsity/model_sparsity": 0.6136267806639486, "compression_loss": 67.44315338134766, "distillation_loss": 2.0946664810180664, "epoch": 2.85, "learning_rate": 4.500626174076393e-05, "loss": 69.6689, "step": 3370, "task_loss": 1.346983790397644 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6308699515690818, "compression/movement_sparsity/importance_threshold": -0.002395707493370843, "compression/movement_sparsity/linear_layer_sparsity": 0.636117724254482, "compression/movement_sparsity/model_sparsity": 0.6142651234993264, "compression_loss": 67.48924255371094, "distillation_loss": 2.282266616821289, "epoch": 2.85, "learning_rate": 4.500313087038197e-05, "loss": 69.7781, "step": 3371, "task_loss": 2.173170804977417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6313043925387178, "compression/movement_sparsity/importance_threshold": -0.0023928879085366945, "compression/movement_sparsity/linear_layer_sparsity": 0.6367024854353447, "compression/movement_sparsity/model_sparsity": 0.6148297963346907, "compression_loss": 67.5352783203125, "distillation_loss": 2.8902149200439453, "epoch": 2.85, "learning_rate": 4.5e-05, "loss": 70.3861, "step": 3372, "task_loss": 1.3072479963302612 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6317384925029005, "compression/movement_sparsity/importance_threshold": -0.002390070536877343, "compression/movement_sparsity/linear_layer_sparsity": 0.6373444706966919, "compression/movement_sparsity/model_sparsity": 0.615449727427333, "compression_loss": 67.5813217163086, "distillation_loss": 2.775595188140869, "epoch": 2.85, "learning_rate": 4.4996869129618035e-05, "loss": 69.6583, "step": 3373, "task_loss": 1.9536257982254028 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6321722515955155, "compression/movement_sparsity/importance_threshold": -0.0023872553775238503, "compression/movement_sparsity/linear_layer_sparsity": 0.6378270894575858, "compression/movement_sparsity/model_sparsity": 0.6159157667490812, "compression_loss": 67.62728118896484, "distillation_loss": 2.2821874618530273, "epoch": 2.85, "learning_rate": 4.4993738259236066e-05, "loss": 70.2598, "step": 3374, "task_loss": 1.5996285676956177 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6326056699504472, "compression/movement_sparsity/importance_threshold": -0.0023844424296072878, "compression/movement_sparsity/linear_layer_sparsity": 0.6384006299967032, "compression/movement_sparsity/model_sparsity": 0.6164696044062629, "compression_loss": 67.67321014404297, "distillation_loss": 4.128964424133301, "epoch": 2.85, "learning_rate": 4.4990607388854105e-05, "loss": 70.0235, "step": 3375, "task_loss": 2.488783836364746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6330387477015813, "compression/movement_sparsity/importance_threshold": -0.0023816316922587177, "compression/movement_sparsity/linear_layer_sparsity": 0.6390497578344643, "compression/movement_sparsity/model_sparsity": 0.617096432705846, "compression_loss": 67.7190933227539, "distillation_loss": 2.0634894371032715, "epoch": 2.85, "learning_rate": 4.498747651847214e-05, "loss": 70.2092, "step": 3376, "task_loss": 1.77255380153656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6334714849828029, "compression/movement_sparsity/importance_threshold": -0.002378823164609207, "compression/movement_sparsity/linear_layer_sparsity": 0.6395900299458777, "compression/movement_sparsity/model_sparsity": 0.6176181448081609, "compression_loss": 67.76494598388672, "distillation_loss": 3.531494140625, "epoch": 2.85, "learning_rate": 4.498434564809017e-05, "loss": 70.7812, "step": 3377, "task_loss": 2.050262212753296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6339038819279967, "compression/movement_sparsity/importance_threshold": -0.0023760168457898235, "compression/movement_sparsity/linear_layer_sparsity": 0.64008743467464, "compression/movement_sparsity/model_sparsity": 0.6180984621542943, "compression_loss": 67.81076049804688, "distillation_loss": 1.9626085758209229, "epoch": 2.85, "learning_rate": 4.49812147777082e-05, "loss": 69.8193, "step": 3378, "task_loss": 0.764014720916748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6343359386710488, "compression/movement_sparsity/importance_threshold": -0.002373212734931629, "compression/movement_sparsity/linear_layer_sparsity": 0.6408018473302075, "compression/movement_sparsity/model_sparsity": 0.6187883325373527, "compression_loss": 67.85653686523438, "distillation_loss": 2.856372833251953, "epoch": 2.86, "learning_rate": 4.497808390732624e-05, "loss": 70.3695, "step": 3379, "task_loss": 2.345726490020752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6347676553458432, "compression/movement_sparsity/importance_threshold": -0.0023704108311656943, "compression/movement_sparsity/linear_layer_sparsity": 0.6412526285635136, "compression/movement_sparsity/model_sparsity": 0.6192236280485295, "compression_loss": 67.90226745605469, "distillation_loss": 1.8635663986206055, "epoch": 2.86, "learning_rate": 4.497495303694427e-05, "loss": 70.2566, "step": 3380, "task_loss": 1.1365408897399902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6351990320862659, "compression/movement_sparsity/importance_threshold": -0.0023676111336230817, "compression/movement_sparsity/linear_layer_sparsity": 0.6418544055315928, "compression/movement_sparsity/model_sparsity": 0.6198047321264726, "compression_loss": 67.94792938232422, "distillation_loss": 3.010016918182373, "epoch": 2.86, "learning_rate": 4.49718221665623e-05, "loss": 70.3348, "step": 3381, "task_loss": 1.842218041419983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6356300690262014, "compression/movement_sparsity/importance_threshold": -0.0023648136414348604, "compression/movement_sparsity/linear_layer_sparsity": 0.6425611986440409, "compression/movement_sparsity/model_sparsity": 0.6204872447211582, "compression_loss": 67.99358367919922, "distillation_loss": 2.601419687271118, "epoch": 2.86, "learning_rate": 4.496869129618034e-05, "loss": 70.1476, "step": 3382, "task_loss": 1.9611475467681885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6360607662995356, "compression/movement_sparsity/importance_threshold": -0.0023620183537320927, "compression/movement_sparsity/linear_layer_sparsity": 0.6430102389488722, "compression/movement_sparsity/model_sparsity": 0.620920859110109, "compression_loss": 68.03917694091797, "distillation_loss": 1.8188140392303467, "epoch": 2.86, "learning_rate": 4.496556042579837e-05, "loss": 70.1492, "step": 3383, "task_loss": 0.8573347330093384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6364911240401527, "compression/movement_sparsity/importance_threshold": -0.0023592252696458494, "compression/movement_sparsity/linear_layer_sparsity": 0.6436239520087548, "compression/movement_sparsity/model_sparsity": 0.6215134892383825, "compression_loss": 68.08475494384766, "distillation_loss": 2.4977173805236816, "epoch": 2.86, "learning_rate": 4.4962429555416405e-05, "loss": 70.7079, "step": 3384, "task_loss": 2.1968324184417725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6369211423819386, "compression/movement_sparsity/importance_threshold": -0.002356434388307192, "compression/movement_sparsity/linear_layer_sparsity": 0.6440064196856752, "compression/movement_sparsity/model_sparsity": 0.6218828179739924, "compression_loss": 68.1302719116211, "distillation_loss": 1.9991624355316162, "epoch": 2.86, "learning_rate": 4.495929868503444e-05, "loss": 70.5471, "step": 3385, "task_loss": 1.0663713216781616 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6373508214587782, "compression/movement_sparsity/importance_threshold": -0.002353645708847189, "compression/movement_sparsity/linear_layer_sparsity": 0.6445695146539435, "compression/movement_sparsity/model_sparsity": 0.6224265688978179, "compression_loss": 68.17575073242188, "distillation_loss": 2.1396641731262207, "epoch": 2.86, "learning_rate": 4.4956167814652475e-05, "loss": 70.474, "step": 3386, "task_loss": 1.3780348300933838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6377801614045565, "compression/movement_sparsity/importance_threshold": -0.0023508592303969067, "compression/movement_sparsity/linear_layer_sparsity": 0.6450430471990989, "compression/movement_sparsity/model_sparsity": 0.6228838341432907, "compression_loss": 68.2212142944336, "distillation_loss": 3.2690134048461914, "epoch": 2.86, "learning_rate": 4.4953036944270513e-05, "loss": 71.3805, "step": 3387, "task_loss": 2.6652543544769287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6382091623531588, "compression/movement_sparsity/importance_threshold": -0.00234807495208741, "compression/movement_sparsity/linear_layer_sparsity": 0.6455895198776829, "compression/movement_sparsity/model_sparsity": 0.6234115338042188, "compression_loss": 68.26666259765625, "distillation_loss": 1.6910722255706787, "epoch": 2.86, "learning_rate": 4.4949906073888545e-05, "loss": 70.8578, "step": 3388, "task_loss": 1.295515537261963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6386378244384701, "compression/movement_sparsity/importance_threshold": -0.002345292873049766, "compression/movement_sparsity/linear_layer_sparsity": 0.6462264731402877, "compression/movement_sparsity/model_sparsity": 0.6240266057627557, "compression_loss": 68.31204223632812, "distillation_loss": 3.0106186866760254, "epoch": 2.86, "learning_rate": 4.494677520350658e-05, "loss": 71.5362, "step": 3389, "task_loss": 1.9969996213912964 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6390661477943758, "compression/movement_sparsity/importance_threshold": -0.0023425129924150388, "compression/movement_sparsity/linear_layer_sparsity": 0.6467193466853485, "compression/movement_sparsity/model_sparsity": 0.6245025475852872, "compression_loss": 68.35741424560547, "distillation_loss": 2.808363914489746, "epoch": 2.87, "learning_rate": 4.4943644333124616e-05, "loss": 70.9102, "step": 3390, "task_loss": 2.0040805339813232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6394941325547607, "compression/movement_sparsity/importance_threshold": -0.0023397353093142963, "compression/movement_sparsity/linear_layer_sparsity": 0.6472568762382056, "compression/movement_sparsity/model_sparsity": 0.6250216113443694, "compression_loss": 68.4027328491211, "distillation_loss": 3.106311082839966, "epoch": 2.87, "learning_rate": 4.494051346274265e-05, "loss": 70.959, "step": 3391, "task_loss": 1.7947921752929688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6399217788535103, "compression/movement_sparsity/importance_threshold": -0.0023369598228786034, "compression/movement_sparsity/linear_layer_sparsity": 0.6477334017494376, "compression/movement_sparsity/model_sparsity": 0.6254817667383266, "compression_loss": 68.447998046875, "distillation_loss": 2.161774158477783, "epoch": 2.87, "learning_rate": 4.493738259236068e-05, "loss": 71.1963, "step": 3392, "task_loss": 1.8395476341247559 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6403490868245094, "compression/movement_sparsity/importance_threshold": -0.0023341865322390274, "compression/movement_sparsity/linear_layer_sparsity": 0.6483543408549075, "compression/movement_sparsity/model_sparsity": 0.6260813746752916, "compression_loss": 68.49327850341797, "distillation_loss": 3.031536340713501, "epoch": 2.87, "learning_rate": 4.493425172197871e-05, "loss": 71.4443, "step": 3393, "task_loss": 1.4873019456863403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6407760566016435, "compression/movement_sparsity/importance_threshold": -0.0023314154365266324, "compression/movement_sparsity/linear_layer_sparsity": 0.6488712773702574, "compression/movement_sparsity/model_sparsity": 0.6265805528310565, "compression_loss": 68.53851318359375, "distillation_loss": 1.5267701148986816, "epoch": 2.87, "learning_rate": 4.493112085159675e-05, "loss": 70.9565, "step": 3394, "task_loss": 0.76805579662323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6412026883187972, "compression/movement_sparsity/importance_threshold": -0.0023286465348724885, "compression/movement_sparsity/linear_layer_sparsity": 0.6493846962561548, "compression/movement_sparsity/model_sparsity": 0.627076334198762, "compression_loss": 68.58367156982422, "distillation_loss": 2.982058525085449, "epoch": 2.87, "learning_rate": 4.492798998121478e-05, "loss": 71.5202, "step": 3395, "task_loss": 1.4707194566726685 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6416289821098562, "compression/movement_sparsity/importance_threshold": -0.002325879826407657, "compression/movement_sparsity/linear_layer_sparsity": 0.6499466345801624, "compression/movement_sparsity/model_sparsity": 0.6276189682126155, "compression_loss": 68.62882995605469, "distillation_loss": 2.8209359645843506, "epoch": 2.87, "learning_rate": 4.492485911083281e-05, "loss": 72.406, "step": 3396, "task_loss": 1.6325701475143433 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6420549381087054, "compression/movement_sparsity/importance_threshold": -0.0023231153102632055, "compression/movement_sparsity/linear_layer_sparsity": 0.6504521477429173, "compression/movement_sparsity/model_sparsity": 0.6281071154430893, "compression_loss": 68.67396545410156, "distillation_loss": 4.183990955352783, "epoch": 2.87, "learning_rate": 4.492172824045085e-05, "loss": 71.5157, "step": 3397, "task_loss": 1.7056059837341309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6424805564492297, "compression/movement_sparsity/importance_threshold": -0.002320352985570202, "compression/movement_sparsity/linear_layer_sparsity": 0.6509831190035745, "compression/movement_sparsity/model_sparsity": 0.6286198462074845, "compression_loss": 68.71900177001953, "distillation_loss": 3.341855049133301, "epoch": 2.87, "learning_rate": 4.491859737006888e-05, "loss": 71.5725, "step": 3398, "task_loss": 2.2298405170440674 }, { "compression/movement_sparsity/importance_regularization_factor": 0.642905837265315, "compression/movement_sparsity/importance_threshold": -0.0023175928514597086, "compression/movement_sparsity/linear_layer_sparsity": 0.6515089747963161, "compression/movement_sparsity/model_sparsity": 0.6291276372360236, "compression_loss": 68.7640609741211, "distillation_loss": 2.6204099655151367, "epoch": 2.87, "learning_rate": 4.4915466499686915e-05, "loss": 70.7993, "step": 3399, "task_loss": 1.6781377792358398 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6433307806908455, "compression/movement_sparsity/importance_threshold": -0.0023148349070627964, "compression/movement_sparsity/linear_layer_sparsity": 0.6519463532651996, "compression/movement_sparsity/model_sparsity": 0.6295499904089674, "compression_loss": 68.80907440185547, "distillation_loss": 2.452299118041992, "epoch": 2.87, "learning_rate": 4.491233562930495e-05, "loss": 71.4413, "step": 3400, "task_loss": 1.9522731304168701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6437553868597072, "compression/movement_sparsity/importance_threshold": -0.0023120791515105254, "compression/movement_sparsity/linear_layer_sparsity": 0.6524792323926787, "compression/movement_sparsity/model_sparsity": 0.6300645634990897, "compression_loss": 68.85404968261719, "distillation_loss": 2.6115052700042725, "epoch": 2.87, "learning_rate": 4.4909204758922985e-05, "loss": 71.2059, "step": 3401, "task_loss": 1.4788455963134766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6441796559057844, "compression/movement_sparsity/importance_threshold": -0.002309325583933968, "compression/movement_sparsity/linear_layer_sparsity": 0.6531279786570754, "compression/movement_sparsity/model_sparsity": 0.6306910233335274, "compression_loss": 68.89892578125, "distillation_loss": 3.0341787338256836, "epoch": 2.88, "learning_rate": 4.490607388854102e-05, "loss": 71.9553, "step": 3402, "task_loss": 1.9828135967254639 }, { "compression/movement_sparsity/importance_regularization_factor": 0.644603587962963, "compression/movement_sparsity/importance_threshold": -0.002306574203464185, "compression/movement_sparsity/linear_layer_sparsity": 0.6537347041547235, "compression/movement_sparsity/model_sparsity": 0.6312769059438252, "compression_loss": 68.94383239746094, "distillation_loss": 2.70613169670105, "epoch": 2.88, "learning_rate": 4.490294301815905e-05, "loss": 71.4442, "step": 3403, "task_loss": 2.2430622577667236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6450271831651275, "compression/movement_sparsity/importance_threshold": -0.0023038250092322467, "compression/movement_sparsity/linear_layer_sparsity": 0.6543468074519752, "compression/movement_sparsity/model_sparsity": 0.6318679816097664, "compression_loss": 68.98863983154297, "distillation_loss": 2.194183349609375, "epoch": 2.88, "learning_rate": 4.489981214777708e-05, "loss": 70.9935, "step": 3404, "task_loss": 1.800029993057251 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6454504416461635, "compression/movement_sparsity/importance_threshold": -0.0023010780003692157, "compression/movement_sparsity/linear_layer_sparsity": 0.654818360585303, "compression/movement_sparsity/model_sparsity": 0.6323233354422974, "compression_loss": 69.03343200683594, "distillation_loss": 3.737919330596924, "epoch": 2.88, "learning_rate": 4.489668127739512e-05, "loss": 72.4726, "step": 3405, "task_loss": 2.126943826675415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6458733635399561, "compression/movement_sparsity/importance_threshold": -0.00229833317600616, "compression/movement_sparsity/linear_layer_sparsity": 0.6553046996864993, "compression/movement_sparsity/model_sparsity": 0.6327929672992134, "compression_loss": 69.07816314697266, "distillation_loss": 2.2461514472961426, "epoch": 2.88, "learning_rate": 4.489355040701315e-05, "loss": 71.3879, "step": 3406, "task_loss": 1.5593235492706299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6462959489803901, "compression/movement_sparsity/importance_threshold": -0.0022955905352741446, "compression/movement_sparsity/linear_layer_sparsity": 0.655667277851803, "compression/movement_sparsity/model_sparsity": 0.633143089789118, "compression_loss": 69.12288665771484, "distillation_loss": 3.868701696395874, "epoch": 2.88, "learning_rate": 4.489041953663118e-05, "loss": 72.515, "step": 3407, "task_loss": 1.469499945640564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.646718198101351, "compression/movement_sparsity/importance_threshold": -0.002292850077304236, "compression/movement_sparsity/linear_layer_sparsity": 0.6561557752273414, "compression/movement_sparsity/model_sparsity": 0.633614805777013, "compression_loss": 69.16754913330078, "distillation_loss": 3.3368217945098877, "epoch": 2.88, "learning_rate": 4.488728866624922e-05, "loss": 71.9615, "step": 3408, "task_loss": 2.3147497177124023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6471401110367236, "compression/movement_sparsity/importance_threshold": -0.002290111801227501, "compression/movement_sparsity/linear_layer_sparsity": 0.6566304882650927, "compression/movement_sparsity/model_sparsity": 0.6340732109615295, "compression_loss": 69.21221160888672, "distillation_loss": 3.434535503387451, "epoch": 2.88, "learning_rate": 4.488415779586725e-05, "loss": 72.714, "step": 3409, "task_loss": 1.401450514793396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6475616879203937, "compression/movement_sparsity/importance_threshold": -0.0022873757061750025, "compression/movement_sparsity/linear_layer_sparsity": 0.6571728948025128, "compression/movement_sparsity/model_sparsity": 0.6345969841657516, "compression_loss": 69.25682067871094, "distillation_loss": 1.980058193206787, "epoch": 2.88, "learning_rate": 4.4881026925485285e-05, "loss": 71.7449, "step": 3410, "task_loss": 1.1698156595230103 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6479829288862455, "compression/movement_sparsity/importance_threshold": -0.002284641791277812, "compression/movement_sparsity/linear_layer_sparsity": 0.657599935018056, "compression/movement_sparsity/model_sparsity": 0.6350093542361614, "compression_loss": 69.30143737792969, "distillation_loss": 1.5256799459457397, "epoch": 2.88, "learning_rate": 4.487789605510332e-05, "loss": 71.98, "step": 3411, "task_loss": 0.9076328873634338 }, { "compression/movement_sparsity/importance_regularization_factor": 0.648403834068165, "compression/movement_sparsity/importance_threshold": -0.00228191005566699, "compression/movement_sparsity/linear_layer_sparsity": 0.658170279880247, "compression/movement_sparsity/model_sparsity": 0.6355601059977501, "compression_loss": 69.34597778320312, "distillation_loss": 4.437154769897461, "epoch": 2.88, "learning_rate": 4.4874765184721355e-05, "loss": 72.1943, "step": 3412, "task_loss": 3.125509262084961 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6488244036000367, "compression/movement_sparsity/importance_threshold": -0.0022791804984736075, "compression/movement_sparsity/linear_layer_sparsity": 0.6585898674910178, "compression/movement_sparsity/model_sparsity": 0.6359652794832884, "compression_loss": 69.3905258178711, "distillation_loss": 2.655287504196167, "epoch": 2.88, "learning_rate": 4.487163431433939e-05, "loss": 72.1571, "step": 3413, "task_loss": 1.9836407899856567 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6492446376157464, "compression/movement_sparsity/importance_threshold": -0.002276453118828725, "compression/movement_sparsity/linear_layer_sparsity": 0.6590464915664656, "compression/movement_sparsity/model_sparsity": 0.6364062171170046, "compression_loss": 69.43505096435547, "distillation_loss": 3.517354965209961, "epoch": 2.89, "learning_rate": 4.486850344395742e-05, "loss": 72.6411, "step": 3414, "task_loss": 3.0586912631988525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6496645362491783, "compression/movement_sparsity/importance_threshold": -0.0022737279158634143, "compression/movement_sparsity/linear_layer_sparsity": 0.6596695293254393, "compression/movement_sparsity/model_sparsity": 0.6370078516122694, "compression_loss": 69.47945404052734, "distillation_loss": 3.155996799468994, "epoch": 2.89, "learning_rate": 4.486537257357545e-05, "loss": 72.525, "step": 3415, "task_loss": 2.509345293045044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6500840996342183, "compression/movement_sparsity/importance_threshold": -0.0022710048887087378, "compression/movement_sparsity/linear_layer_sparsity": 0.6601180926635652, "compression/movement_sparsity/model_sparsity": 0.6374410054197883, "compression_loss": 69.52392578125, "distillation_loss": 3.7616422176361084, "epoch": 2.89, "learning_rate": 4.486224170319349e-05, "loss": 72.0826, "step": 3416, "task_loss": 1.6629836559295654 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6505033279047514, "compression/movement_sparsity/importance_threshold": -0.0022682840364957628, "compression/movement_sparsity/linear_layer_sparsity": 0.660689033734138, "compression/movement_sparsity/model_sparsity": 0.6379923329081668, "compression_loss": 69.56831359863281, "distillation_loss": 2.4246621131896973, "epoch": 2.89, "learning_rate": 4.485911083281152e-05, "loss": 72.3429, "step": 3417, "task_loss": 2.176872730255127 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6509222211946627, "compression/movement_sparsity/importance_threshold": -0.002265565358355554, "compression/movement_sparsity/linear_layer_sparsity": 0.6610667555923392, "compression/movement_sparsity/model_sparsity": 0.6383570788585304, "compression_loss": 69.61270141601562, "distillation_loss": 4.582287788391113, "epoch": 2.89, "learning_rate": 4.485597996242955e-05, "loss": 72.6946, "step": 3418, "task_loss": 2.601179361343384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6513407796378373, "compression/movement_sparsity/importance_threshold": -0.0022628488534191787, "compression/movement_sparsity/linear_layer_sparsity": 0.6615629082834998, "compression/movement_sparsity/model_sparsity": 0.6388361871784055, "compression_loss": 69.65703582763672, "distillation_loss": 3.084564208984375, "epoch": 2.89, "learning_rate": 4.485284909204759e-05, "loss": 72.4826, "step": 3419, "task_loss": 1.5617097616195679 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6517590033681602, "compression/movement_sparsity/importance_threshold": -0.0022601345208177036, "compression/movement_sparsity/linear_layer_sparsity": 0.6619666844479855, "compression/movement_sparsity/model_sparsity": 0.6392260923894802, "compression_loss": 69.70130920410156, "distillation_loss": 2.3321046829223633, "epoch": 2.89, "learning_rate": 4.484971822166562e-05, "loss": 72.7844, "step": 3420, "task_loss": 2.1889050006866455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6521768925195169, "compression/movement_sparsity/importance_threshold": -0.0022574223596821923, "compression/movement_sparsity/linear_layer_sparsity": 0.6624675710336975, "compression/movement_sparsity/model_sparsity": 0.6397097719800657, "compression_loss": 69.74561309814453, "distillation_loss": 2.635995864868164, "epoch": 2.89, "learning_rate": 4.484658735128366e-05, "loss": 72.2236, "step": 3421, "task_loss": 2.2535831928253174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6525944472257921, "compression/movement_sparsity/importance_threshold": -0.0022547123691437146, "compression/movement_sparsity/linear_layer_sparsity": 0.6630068892116999, "compression/movement_sparsity/model_sparsity": 0.640230562919517, "compression_loss": 69.789794921875, "distillation_loss": 4.548501968383789, "epoch": 2.89, "learning_rate": 4.4843456480901694e-05, "loss": 73.2014, "step": 3422, "task_loss": 2.361332893371582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6530116676208715, "compression/movement_sparsity/importance_threshold": -0.002252004548333332, "compression/movement_sparsity/linear_layer_sparsity": 0.6635275341431847, "compression/movement_sparsity/model_sparsity": 0.640733322095914, "compression_loss": 69.83395385742188, "distillation_loss": 2.717738151550293, "epoch": 2.89, "learning_rate": 4.484032561051973e-05, "loss": 72.31, "step": 3423, "task_loss": 1.2202725410461426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6534285538386396, "compression/movement_sparsity/importance_threshold": -0.0022492988963821146, "compression/movement_sparsity/linear_layer_sparsity": 0.6640912968648406, "compression/movement_sparsity/model_sparsity": 0.6412777178337441, "compression_loss": 69.87810516357422, "distillation_loss": 1.92867910861969, "epoch": 2.89, "learning_rate": 4.4837194740137764e-05, "loss": 72.7617, "step": 3424, "task_loss": 1.3874132633209229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.653845106012982, "compression/movement_sparsity/importance_threshold": -0.0022465954124211264, "compression/movement_sparsity/linear_layer_sparsity": 0.6647618524318434, "compression/movement_sparsity/model_sparsity": 0.6419252377541501, "compression_loss": 69.92219543457031, "distillation_loss": 3.5986006259918213, "epoch": 2.89, "learning_rate": 4.4834063869755796e-05, "loss": 72.3578, "step": 3425, "task_loss": 3.0255300998687744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6542613242777837, "compression/movement_sparsity/importance_threshold": -0.0022438940955814323, "compression/movement_sparsity/linear_layer_sparsity": 0.6652707639823744, "compression/movement_sparsity/model_sparsity": 0.6424166666273253, "compression_loss": 69.96624755859375, "distillation_loss": 3.467825174331665, "epoch": 2.9, "learning_rate": 4.483093299937383e-05, "loss": 72.8366, "step": 3426, "task_loss": 1.5931099653244019 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6546772087669298, "compression/movement_sparsity/importance_threshold": -0.0022411949449941005, "compression/movement_sparsity/linear_layer_sparsity": 0.6658381993476622, "compression/movement_sparsity/model_sparsity": 0.6429646088421801, "compression_loss": 70.01026153564453, "distillation_loss": 2.091411590576172, "epoch": 2.9, "learning_rate": 4.4827802128991866e-05, "loss": 73.181, "step": 3427, "task_loss": 1.1606403589248657 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6550927596143055, "compression/movement_sparsity/importance_threshold": -0.002238497959790197, "compression/movement_sparsity/linear_layer_sparsity": 0.6663934601377938, "compression/movement_sparsity/model_sparsity": 0.6435007947159888, "compression_loss": 70.05431365966797, "distillation_loss": 2.133512496948242, "epoch": 2.9, "learning_rate": 4.48246712586099e-05, "loss": 72.4935, "step": 3428, "task_loss": 1.1999799013137817 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6555079769537958, "compression/movement_sparsity/importance_threshold": -0.002235803139100787, "compression/movement_sparsity/linear_layer_sparsity": 0.6668834957309572, "compression/movement_sparsity/model_sparsity": 0.6439739960790012, "compression_loss": 70.0982666015625, "distillation_loss": 3.394122838973999, "epoch": 2.9, "learning_rate": 4.482154038822793e-05, "loss": 72.9755, "step": 3429, "task_loss": 2.196857452392578 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6559228609192862, "compression/movement_sparsity/importance_threshold": -0.002233110482056936, "compression/movement_sparsity/linear_layer_sparsity": 0.6675367135582174, "compression/movement_sparsity/model_sparsity": 0.6446047738643619, "compression_loss": 70.14221954345703, "distillation_loss": 2.7874722480773926, "epoch": 2.9, "learning_rate": 4.481840951784596e-05, "loss": 72.7271, "step": 3430, "task_loss": 1.9253556728363037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6563374116446614, "compression/movement_sparsity/importance_threshold": -0.002230419987789711, "compression/movement_sparsity/linear_layer_sparsity": 0.6679517342127836, "compression/movement_sparsity/model_sparsity": 0.6450055372826909, "compression_loss": 70.18611145019531, "distillation_loss": 3.235100746154785, "epoch": 2.9, "learning_rate": 4.4815278647464e-05, "loss": 73.0115, "step": 3431, "task_loss": 1.111924409866333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6567516292638069, "compression/movement_sparsity/importance_threshold": -0.0022277316554301773, "compression/movement_sparsity/linear_layer_sparsity": 0.6683942519979181, "compression/movement_sparsity/model_sparsity": 0.6454328532205621, "compression_loss": 70.22991180419922, "distillation_loss": 3.8443167209625244, "epoch": 2.9, "learning_rate": 4.481214777708203e-05, "loss": 73.3982, "step": 3432, "task_loss": 2.7981274127960205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6571655139106075, "compression/movement_sparsity/importance_threshold": -0.002225045484109403, "compression/movement_sparsity/linear_layer_sparsity": 0.6688004368442662, "compression/movement_sparsity/model_sparsity": 0.6458250843678673, "compression_loss": 70.27371215820312, "distillation_loss": 3.055354356765747, "epoch": 2.9, "learning_rate": 4.4809016906700063e-05, "loss": 73.1371, "step": 3433, "task_loss": 1.1091006994247437 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6575790657189489, "compression/movement_sparsity/importance_threshold": -0.0022223614729584493, "compression/movement_sparsity/linear_layer_sparsity": 0.6692659444246026, "compression/movement_sparsity/model_sparsity": 0.6462746003307505, "compression_loss": 70.31751251220703, "distillation_loss": 4.013768672943115, "epoch": 2.9, "learning_rate": 4.48058860363181e-05, "loss": 73.1921, "step": 3434, "task_loss": 2.400937557220459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6579922848227153, "compression/movement_sparsity/importance_threshold": -0.002219679621108389, "compression/movement_sparsity/linear_layer_sparsity": 0.6698191542579008, "compression/movement_sparsity/model_sparsity": 0.6468088057044024, "compression_loss": 70.36123657226562, "distillation_loss": 2.4899086952209473, "epoch": 2.9, "learning_rate": 4.4802755165936134e-05, "loss": 72.7923, "step": 3435, "task_loss": 1.623915672302246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6584051713557927, "compression/movement_sparsity/importance_threshold": -0.002216999927690284, "compression/movement_sparsity/linear_layer_sparsity": 0.6701937639083492, "compression/movement_sparsity/model_sparsity": 0.6471705463609236, "compression_loss": 70.40487670898438, "distillation_loss": 3.5528042316436768, "epoch": 2.9, "learning_rate": 4.4799624295554166e-05, "loss": 73.1294, "step": 3436, "task_loss": 2.0518741607666016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.658817725452066, "compression/movement_sparsity/importance_threshold": -0.002214322391835199, "compression/movement_sparsity/linear_layer_sparsity": 0.6706231651090842, "compression/movement_sparsity/model_sparsity": 0.6475851963094209, "compression_loss": 70.448486328125, "distillation_loss": 1.562474250793457, "epoch": 2.9, "learning_rate": 4.47964934251722e-05, "loss": 72.7367, "step": 3437, "task_loss": 1.2675844430923462 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6592299472454202, "compression/movement_sparsity/importance_threshold": -0.0022116470126742034, "compression/movement_sparsity/linear_layer_sparsity": 0.6711004060703744, "compression/movement_sparsity/model_sparsity": 0.6480460425755258, "compression_loss": 70.49211883544922, "distillation_loss": 1.9307608604431152, "epoch": 2.91, "learning_rate": 4.4793362554790236e-05, "loss": 72.8061, "step": 3438, "task_loss": 1.223360300064087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6596418368697405, "compression/movement_sparsity/importance_threshold": -0.0022089737893383615, "compression/movement_sparsity/linear_layer_sparsity": 0.6716184276849791, "compression/movement_sparsity/model_sparsity": 0.648546268554048, "compression_loss": 70.53563690185547, "distillation_loss": 3.523749351501465, "epoch": 2.91, "learning_rate": 4.479023168440827e-05, "loss": 73.1677, "step": 3439, "task_loss": 4.739435195922852 }, { "compression/movement_sparsity/importance_regularization_factor": 0.660053394458912, "compression/movement_sparsity/importance_threshold": -0.002206302720958741, "compression/movement_sparsity/linear_layer_sparsity": 0.6721127202059886, "compression/movement_sparsity/model_sparsity": 0.6490235806063391, "compression_loss": 70.57917785644531, "distillation_loss": 3.1801705360412598, "epoch": 2.91, "learning_rate": 4.47871008140263e-05, "loss": 73.27, "step": 3440, "task_loss": 1.83657705783844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6604646201468201, "compression/movement_sparsity/importance_threshold": -0.0022036338066664055, "compression/movement_sparsity/linear_layer_sparsity": 0.6724696940125034, "compression/movement_sparsity/model_sparsity": 0.6493682912644203, "compression_loss": 70.62262725830078, "distillation_loss": 2.4999682903289795, "epoch": 2.91, "learning_rate": 4.478396994364433e-05, "loss": 73.2169, "step": 3441, "task_loss": 1.9320176839828491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6608755140673495, "compression/movement_sparsity/importance_threshold": -0.002200967045592423, "compression/movement_sparsity/linear_layer_sparsity": 0.6728559297263966, "compression/movement_sparsity/model_sparsity": 0.6497412585933412, "compression_loss": 70.66606140136719, "distillation_loss": 2.2819700241088867, "epoch": 2.91, "learning_rate": 4.478083907326237e-05, "loss": 73.3376, "step": 3442, "task_loss": 0.8913378715515137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.661286076354386, "compression/movement_sparsity/importance_threshold": -0.002198302436867856, "compression/movement_sparsity/linear_layer_sparsity": 0.6732112579977777, "compression/movement_sparsity/model_sparsity": 0.6500843802454828, "compression_loss": 70.70951080322266, "distillation_loss": 3.341846227645874, "epoch": 2.91, "learning_rate": 4.47777082028804e-05, "loss": 74.2611, "step": 3443, "task_loss": 1.9620729684829712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6616963071418138, "compression/movement_sparsity/importance_threshold": -0.0021956399796237763, "compression/movement_sparsity/linear_layer_sparsity": 0.6736784826582537, "compression/movement_sparsity/model_sparsity": 0.6505355543015202, "compression_loss": 70.75286865234375, "distillation_loss": 2.993411064147949, "epoch": 2.91, "learning_rate": 4.477457733249843e-05, "loss": 74.0237, "step": 3444, "task_loss": 1.9026890993118286 }, { "compression/movement_sparsity/importance_regularization_factor": 0.662106206563519, "compression/movement_sparsity/importance_threshold": -0.0021929796729912446, "compression/movement_sparsity/linear_layer_sparsity": 0.6741267359680211, "compression/movement_sparsity/model_sparsity": 0.6509684087311086, "compression_loss": 70.79625701904297, "distillation_loss": 3.579315185546875, "epoch": 2.91, "learning_rate": 4.477144646211647e-05, "loss": 74.3112, "step": 3445, "task_loss": 1.7495791912078857 }, { "compression/movement_sparsity/importance_regularization_factor": 0.662515774753386, "compression/movement_sparsity/importance_threshold": -0.0021903215161013312, "compression/movement_sparsity/linear_layer_sparsity": 0.674488455593255, "compression/movement_sparsity/model_sparsity": 0.6513177021744361, "compression_loss": 70.83953857421875, "distillation_loss": 3.7844529151916504, "epoch": 2.91, "learning_rate": 4.4768315591734504e-05, "loss": 74.1311, "step": 3446, "task_loss": 1.8219900131225586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6629250118453004, "compression/movement_sparsity/importance_threshold": -0.0021876655080850993, "compression/movement_sparsity/linear_layer_sparsity": 0.6749504216958037, "compression/movement_sparsity/model_sparsity": 0.6517637983201882, "compression_loss": 70.8828353881836, "distillation_loss": 3.880941867828369, "epoch": 2.91, "learning_rate": 4.4765184721352536e-05, "loss": 73.4512, "step": 3447, "task_loss": 2.177865982055664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6633339179731472, "compression/movement_sparsity/importance_threshold": -0.0021850116480736153, "compression/movement_sparsity/linear_layer_sparsity": 0.6754009167490865, "compression/movement_sparsity/model_sparsity": 0.6521988174825059, "compression_loss": 70.92610168457031, "distillation_loss": 3.257991075515747, "epoch": 2.91, "learning_rate": 4.476205385097057e-05, "loss": 74.0784, "step": 3448, "task_loss": 1.5751965045928955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6637424932708114, "compression/movement_sparsity/importance_threshold": -0.0021823599351979478, "compression/movement_sparsity/linear_layer_sparsity": 0.6760032780013797, "compression/movement_sparsity/model_sparsity": 0.6527804857727029, "compression_loss": 70.96931457519531, "distillation_loss": 3.567519426345825, "epoch": 2.91, "learning_rate": 4.4758922980588606e-05, "loss": 74.0003, "step": 3449, "task_loss": 1.6534219980239868 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6641507378721785, "compression/movement_sparsity/importance_threshold": -0.002179710368589158, "compression/movement_sparsity/linear_layer_sparsity": 0.6766380014446367, "compression/movement_sparsity/model_sparsity": 0.6533934045130463, "compression_loss": 71.01254272460938, "distillation_loss": 3.3617939949035645, "epoch": 2.92, "learning_rate": 4.475579211020664e-05, "loss": 73.3356, "step": 3450, "task_loss": 1.301986813545227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6645586519111331, "compression/movement_sparsity/importance_threshold": -0.0021770629473783177, "compression/movement_sparsity/linear_layer_sparsity": 0.677238693313461, "compression/movement_sparsity/model_sparsity": 0.6539734607682322, "compression_loss": 71.0556640625, "distillation_loss": 2.8911526203155518, "epoch": 2.92, "learning_rate": 4.475266123982467e-05, "loss": 74.1227, "step": 3451, "task_loss": 2.124176025390625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6649662355215611, "compression/movement_sparsity/importance_threshold": -0.002174417670696486, "compression/movement_sparsity/linear_layer_sparsity": 0.6777174009473703, "compression/movement_sparsity/model_sparsity": 0.6544357233222398, "compression_loss": 71.09876251220703, "distillation_loss": 2.8834357261657715, "epoch": 2.92, "learning_rate": 4.474953036944271e-05, "loss": 73.9068, "step": 3452, "task_loss": 1.9531172513961792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6653734888373467, "compression/movement_sparsity/importance_threshold": -0.0021717745376747368, "compression/movement_sparsity/linear_layer_sparsity": 0.6782008305516635, "compression/movement_sparsity/model_sparsity": 0.654902545632422, "compression_loss": 71.14179992675781, "distillation_loss": 1.8043031692504883, "epoch": 2.92, "learning_rate": 4.474639949906074e-05, "loss": 74.2313, "step": 3453, "task_loss": 1.2108323574066162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6657804119923759, "compression/movement_sparsity/importance_threshold": -0.0021691335474441293, "compression/movement_sparsity/linear_layer_sparsity": 0.6786081601181045, "compression/movement_sparsity/model_sparsity": 0.6552958821751635, "compression_loss": 71.18482971191406, "distillation_loss": 3.949230194091797, "epoch": 2.92, "learning_rate": 4.474326862867878e-05, "loss": 74.5804, "step": 3454, "task_loss": 2.726175546646118 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6661870051205332, "compression/movement_sparsity/importance_threshold": -0.002166494699135735, "compression/movement_sparsity/linear_layer_sparsity": 0.6791731987256976, "compression/movement_sparsity/model_sparsity": 0.6558415099683236, "compression_loss": 71.22779083251953, "distillation_loss": 3.0629844665527344, "epoch": 2.92, "learning_rate": 4.474013775829681e-05, "loss": 75.0189, "step": 3455, "task_loss": 2.785687208175659 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6665932683557042, "compression/movement_sparsity/importance_threshold": -0.0021638579918806158, "compression/movement_sparsity/linear_layer_sparsity": 0.6794760368111457, "compression/movement_sparsity/model_sparsity": 0.6561339446338976, "compression_loss": 71.27070617675781, "distillation_loss": 3.3594069480895996, "epoch": 2.92, "learning_rate": 4.473700688791484e-05, "loss": 74.1904, "step": 3456, "task_loss": 2.962794542312622 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6669992018317739, "compression/movement_sparsity/importance_threshold": -0.0021612234248098395, "compression/movement_sparsity/linear_layer_sparsity": 0.6798061573921445, "compression/movement_sparsity/model_sparsity": 0.6564527245573695, "compression_loss": 71.31364440917969, "distillation_loss": 3.848773241043091, "epoch": 2.92, "learning_rate": 4.473387601753288e-05, "loss": 74.612, "step": 3457, "task_loss": 1.9265755414962769 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6674048056826272, "compression/movement_sparsity/importance_threshold": -0.002158590997054473, "compression/movement_sparsity/linear_layer_sparsity": 0.6802747891044015, "compression/movement_sparsity/model_sparsity": 0.6569052573286307, "compression_loss": 71.35645294189453, "distillation_loss": 2.848858594894409, "epoch": 2.92, "learning_rate": 4.473074514715091e-05, "loss": 74.0979, "step": 3458, "task_loss": 1.9163103103637695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6678100800421496, "compression/movement_sparsity/importance_threshold": -0.00215596070774558, "compression/movement_sparsity/linear_layer_sparsity": 0.6807631195415931, "compression/movement_sparsity/model_sparsity": 0.6573768121130246, "compression_loss": 71.39934539794922, "distillation_loss": 3.1114940643310547, "epoch": 2.92, "learning_rate": 4.4727614276768944e-05, "loss": 74.7155, "step": 3459, "task_loss": 1.8489569425582886 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6682150250442258, "compression/movement_sparsity/importance_threshold": -0.0021533325560142303, "compression/movement_sparsity/linear_layer_sparsity": 0.6811310397339976, "compression/movement_sparsity/model_sparsity": 0.657732093114965, "compression_loss": 71.44219207763672, "distillation_loss": 2.9500627517700195, "epoch": 2.92, "learning_rate": 4.472448340638698e-05, "loss": 74.0899, "step": 3460, "task_loss": 2.504326581954956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6686196408227416, "compression/movement_sparsity/importance_threshold": -0.0021507065409914847, "compression/movement_sparsity/linear_layer_sparsity": 0.6815817255739627, "compression/movement_sparsity/model_sparsity": 0.6581672965098555, "compression_loss": 71.48497772216797, "distillation_loss": 2.910615921020508, "epoch": 2.93, "learning_rate": 4.4721352536005014e-05, "loss": 74.4497, "step": 3461, "task_loss": 1.9382023811340332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6690239275115815, "compression/movement_sparsity/importance_threshold": -0.0021480826618084144, "compression/movement_sparsity/linear_layer_sparsity": 0.681950373140593, "compression/movement_sparsity/model_sparsity": 0.6585232798984796, "compression_loss": 71.52770233154297, "distillation_loss": 1.573062539100647, "epoch": 2.93, "learning_rate": 4.4718221665623046e-05, "loss": 73.9911, "step": 3462, "task_loss": 1.474571704864502 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6694278852446311, "compression/movement_sparsity/importance_threshold": -0.002145460917596081, "compression/movement_sparsity/linear_layer_sparsity": 0.6824965000183155, "compression/movement_sparsity/model_sparsity": 0.6590506456378695, "compression_loss": 71.5704345703125, "distillation_loss": 2.4699759483337402, "epoch": 2.93, "learning_rate": 4.471509079524108e-05, "loss": 74.537, "step": 3463, "task_loss": 0.9548800587654114 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6698315141557751, "compression/movement_sparsity/importance_threshold": -0.002142841307485554, "compression/movement_sparsity/linear_layer_sparsity": 0.6829617691152993, "compression/movement_sparsity/model_sparsity": 0.6594999313100367, "compression_loss": 71.61307525634766, "distillation_loss": 2.190371513366699, "epoch": 2.93, "learning_rate": 4.4711959924859116e-05, "loss": 73.9244, "step": 3464, "task_loss": 1.73721182346344 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6702348143788992, "compression/movement_sparsity/importance_threshold": -0.002140223830607897, "compression/movement_sparsity/linear_layer_sparsity": 0.6833710303968974, "compression/movement_sparsity/model_sparsity": 0.659895133207577, "compression_loss": 71.65573120117188, "distillation_loss": 3.6052846908569336, "epoch": 2.93, "learning_rate": 4.470882905447715e-05, "loss": 74.5782, "step": 3465, "task_loss": 1.7624422311782837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6706377860478878, "compression/movement_sparsity/importance_threshold": -0.002137608486094179, "compression/movement_sparsity/linear_layer_sparsity": 0.683831195950133, "compression/movement_sparsity/model_sparsity": 0.6603394906584241, "compression_loss": 71.69828796386719, "distillation_loss": 2.559915065765381, "epoch": 2.93, "learning_rate": 4.470569818409518e-05, "loss": 74.7307, "step": 3466, "task_loss": 1.302794098854065 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6710404292966268, "compression/movement_sparsity/importance_threshold": -0.002134995273075462, "compression/movement_sparsity/linear_layer_sparsity": 0.6841823269145062, "compression/movement_sparsity/model_sparsity": 0.6606785591939659, "compression_loss": 71.74089813232422, "distillation_loss": 2.668214797973633, "epoch": 2.93, "learning_rate": 4.470256731371321e-05, "loss": 74.373, "step": 3467, "task_loss": 1.4021415710449219 }, { "compression/movement_sparsity/importance_regularization_factor": 0.671442744259001, "compression/movement_sparsity/importance_threshold": -0.002132384190682815, "compression/movement_sparsity/linear_layer_sparsity": 0.6845842667571759, "compression/movement_sparsity/model_sparsity": 0.6610666911665283, "compression_loss": 71.78337097167969, "distillation_loss": 3.1722607612609863, "epoch": 2.93, "learning_rate": 4.469943644333125e-05, "loss": 75.2022, "step": 3468, "task_loss": 1.635369896888733 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6718447310688952, "compression/movement_sparsity/importance_threshold": -0.0021297752380473045, "compression/movement_sparsity/linear_layer_sparsity": 0.6851068314796499, "compression/movement_sparsity/model_sparsity": 0.6615713041831882, "compression_loss": 71.82585144042969, "distillation_loss": 2.8744375705718994, "epoch": 2.93, "learning_rate": 4.469630557294928e-05, "loss": 74.555, "step": 3469, "task_loss": 1.8934327363967896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6722463898601954, "compression/movement_sparsity/importance_threshold": -0.002127168414299993, "compression/movement_sparsity/linear_layer_sparsity": 0.685682017553901, "compression/movement_sparsity/model_sparsity": 0.6621267308463095, "compression_loss": 71.8682861328125, "distillation_loss": 2.4623866081237793, "epoch": 2.93, "learning_rate": 4.4693174702567314e-05, "loss": 75.0136, "step": 3470, "task_loss": 1.236510992050171 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6726477207667857, "compression/movement_sparsity/importance_threshold": -0.0021245637185719514, "compression/movement_sparsity/linear_layer_sparsity": 0.6862027817270621, "compression/movement_sparsity/model_sparsity": 0.6626296051680645, "compression_loss": 71.91072082519531, "distillation_loss": 2.1692700386047363, "epoch": 2.93, "learning_rate": 4.469004383218535e-05, "loss": 74.6641, "step": 3471, "task_loss": 1.8599106073379517 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6730487239225522, "compression/movement_sparsity/importance_threshold": -0.0021219611499942405, "compression/movement_sparsity/linear_layer_sparsity": 0.6866334230412313, "compression/movement_sparsity/model_sparsity": 0.6630454526282843, "compression_loss": 71.95309448242188, "distillation_loss": 3.546987533569336, "epoch": 2.93, "learning_rate": 4.4686912961803384e-05, "loss": 75.175, "step": 3472, "task_loss": 2.7805216312408447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6734493994613793, "compression/movement_sparsity/importance_threshold": -0.0021193607076979313, "compression/movement_sparsity/linear_layer_sparsity": 0.6870403591101404, "compression/movement_sparsity/model_sparsity": 0.6634384091913447, "compression_loss": 71.99541473388672, "distillation_loss": 2.569979190826416, "epoch": 2.94, "learning_rate": 4.4683782091421416e-05, "loss": 74.7153, "step": 3473, "task_loss": 1.892233967781067 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6738497475171527, "compression/movement_sparsity/importance_threshold": -0.0021167623908140863, "compression/movement_sparsity/linear_layer_sparsity": 0.6874443022129729, "compression/movement_sparsity/model_sparsity": 0.6638284756059205, "compression_loss": 72.03771209716797, "distillation_loss": 3.532655715942383, "epoch": 2.94, "learning_rate": 4.468065122103945e-05, "loss": 75.3389, "step": 3474, "task_loss": 1.738996982574463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6742497682237569, "compression/movement_sparsity/importance_threshold": -0.0021141661984737737, "compression/movement_sparsity/linear_layer_sparsity": 0.6879717677683452, "compression/movement_sparsity/model_sparsity": 0.664337821096792, "compression_loss": 72.08000946044922, "distillation_loss": 3.3573291301727295, "epoch": 2.94, "learning_rate": 4.4677520350657486e-05, "loss": 75.3415, "step": 3475, "task_loss": 2.1079161167144775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6746494617150778, "compression/movement_sparsity/importance_threshold": -0.0021115721298080583, "compression/movement_sparsity/linear_layer_sparsity": 0.688453444519996, "compression/movement_sparsity/model_sparsity": 0.6648029507702125, "compression_loss": 72.12224578857422, "distillation_loss": 5.576750755310059, "epoch": 2.94, "learning_rate": 4.467438948027552e-05, "loss": 75.4725, "step": 3476, "task_loss": 2.7214443683624268 }, { "compression/movement_sparsity/importance_regularization_factor": 0.675048828125, "compression/movement_sparsity/importance_threshold": -0.002108980183948006, "compression/movement_sparsity/linear_layer_sparsity": 0.6889289564769789, "compression/movement_sparsity/model_sparsity": 0.6652621274286272, "compression_loss": 72.16451263427734, "distillation_loss": 2.9775965213775635, "epoch": 2.94, "learning_rate": 4.467125860989355e-05, "loss": 74.9527, "step": 3477, "task_loss": 1.5255008935928345 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6754478675874089, "compression/movement_sparsity/importance_threshold": -0.0021063903600246833, "compression/movement_sparsity/linear_layer_sparsity": 0.6893444660224182, "compression/movement_sparsity/model_sparsity": 0.6656633629429237, "compression_loss": 72.20667266845703, "distillation_loss": 3.1609325408935547, "epoch": 2.94, "learning_rate": 4.466812773951158e-05, "loss": 75.3435, "step": 3478, "task_loss": 2.0263190269470215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6758465802361895, "compression/movement_sparsity/importance_threshold": -0.002103802657169157, "compression/movement_sparsity/linear_layer_sparsity": 0.6897173585927269, "compression/movement_sparsity/model_sparsity": 0.6660234455062907, "compression_loss": 72.24883270263672, "distillation_loss": 3.2071924209594727, "epoch": 2.94, "learning_rate": 4.466499686912962e-05, "loss": 75.159, "step": 3479, "task_loss": 2.519763231277466 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6762449662052268, "compression/movement_sparsity/importance_threshold": -0.0021012170745124924, "compression/movement_sparsity/linear_layer_sparsity": 0.6901984033634929, "compression/movement_sparsity/model_sparsity": 0.666487964909314, "compression_loss": 72.29093170166016, "distillation_loss": 3.2028112411499023, "epoch": 2.94, "learning_rate": 4.466186599874765e-05, "loss": 75.0822, "step": 3480, "task_loss": 2.3323817253112793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6766430256284065, "compression/movement_sparsity/importance_threshold": -0.002098633611185754, "compression/movement_sparsity/linear_layer_sparsity": 0.6906211389545195, "compression/movement_sparsity/model_sparsity": 0.6668961782323021, "compression_loss": 72.3329849243164, "distillation_loss": 3.336866617202759, "epoch": 2.94, "learning_rate": 4.4658735128365684e-05, "loss": 74.9926, "step": 3481, "task_loss": 2.021195411682129 }, { "compression/movement_sparsity/importance_regularization_factor": 0.677040758639613, "compression/movement_sparsity/importance_threshold": -0.002096052266320012, "compression/movement_sparsity/linear_layer_sparsity": 0.691046497862426, "compression/movement_sparsity/model_sparsity": 0.6673069247531649, "compression_loss": 72.37505340576172, "distillation_loss": 2.4574310779571533, "epoch": 2.94, "learning_rate": 4.465560425798372e-05, "loss": 75.1292, "step": 3482, "task_loss": 1.795758843421936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6774381653727322, "compression/movement_sparsity/importance_threshold": -0.002093473039046327, "compression/movement_sparsity/linear_layer_sparsity": 0.6916073630113465, "compression/movement_sparsity/model_sparsity": 0.6678485224587969, "compression_loss": 72.4170150756836, "distillation_loss": 3.2246055603027344, "epoch": 2.94, "learning_rate": 4.4652473387601754e-05, "loss": 75.6457, "step": 3483, "task_loss": 1.2611771821975708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6778352459616486, "compression/movement_sparsity/importance_threshold": -0.0020908959284957696, "compression/movement_sparsity/linear_layer_sparsity": 0.6920581203963173, "compression/movement_sparsity/model_sparsity": 0.6682837949409022, "compression_loss": 72.45894622802734, "distillation_loss": 4.288147449493408, "epoch": 2.94, "learning_rate": 4.4649342517219786e-05, "loss": 75.5265, "step": 3484, "task_loss": 1.9512368440628052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6782320005402478, "compression/movement_sparsity/importance_threshold": -0.0020883209337994027, "compression/movement_sparsity/linear_layer_sparsity": 0.6925494199512502, "compression/movement_sparsity/model_sparsity": 0.6687582168447088, "compression_loss": 72.50092315673828, "distillation_loss": 3.283330202102661, "epoch": 2.95, "learning_rate": 4.464621164683782e-05, "loss": 75.54, "step": 3485, "task_loss": 2.3533103466033936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6786284292424143, "compression/movement_sparsity/importance_threshold": -0.0020857480540882965, "compression/movement_sparsity/linear_layer_sparsity": 0.6930868183382632, "compression/movement_sparsity/model_sparsity": 0.6692771539438972, "compression_loss": 72.54277038574219, "distillation_loss": 4.275600910186768, "epoch": 2.95, "learning_rate": 4.4643080776455856e-05, "loss": 75.7037, "step": 3486, "task_loss": 2.6077089309692383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6790245322020341, "compression/movement_sparsity/importance_threshold": -0.0020831772884935115, "compression/movement_sparsity/linear_layer_sparsity": 0.6934889370434475, "compression/movement_sparsity/model_sparsity": 0.6696654586344964, "compression_loss": 72.58460998535156, "distillation_loss": 3.7982778549194336, "epoch": 2.95, "learning_rate": 4.463994990607389e-05, "loss": 75.7745, "step": 3487, "task_loss": 2.6277151107788086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6794203095529917, "compression/movement_sparsity/importance_threshold": -0.002080608636146118, "compression/movement_sparsity/linear_layer_sparsity": 0.6939291892367312, "compression/movement_sparsity/model_sparsity": 0.6700905868105668, "compression_loss": 72.6264419555664, "distillation_loss": 2.4495623111724854, "epoch": 2.95, "learning_rate": 4.463681903569193e-05, "loss": 75.5503, "step": 3488, "task_loss": 1.8995503187179565 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6798157614291724, "compression/movement_sparsity/importance_threshold": -0.0020780420961771804, "compression/movement_sparsity/linear_layer_sparsity": 0.6943676885773724, "compression/movement_sparsity/model_sparsity": 0.6705140223498752, "compression_loss": 72.6682357788086, "distillation_loss": 4.382484436035156, "epoch": 2.95, "learning_rate": 4.463368816530996e-05, "loss": 76.058, "step": 3489, "task_loss": 2.909686326980591 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6802108879644617, "compression/movement_sparsity/importance_threshold": -0.002075477667717763, "compression/movement_sparsity/linear_layer_sparsity": 0.694687959795904, "compression/movement_sparsity/model_sparsity": 0.6708232912667808, "compression_loss": 72.70996856689453, "distillation_loss": 3.7135767936706543, "epoch": 2.95, "learning_rate": 4.4630557294928e-05, "loss": 75.3828, "step": 3490, "task_loss": 1.6974540948867798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6806056892927441, "compression/movement_sparsity/importance_threshold": -0.002072915349898937, "compression/movement_sparsity/linear_layer_sparsity": 0.6950961359782473, "compression/movement_sparsity/model_sparsity": 0.6712174453415637, "compression_loss": 72.75165557861328, "distillation_loss": 3.0363383293151855, "epoch": 2.95, "learning_rate": 4.462742642454603e-05, "loss": 75.8568, "step": 3491, "task_loss": 1.5406562089920044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6810001655479054, "compression/movement_sparsity/importance_threshold": -0.0020703551418517615, "compression/movement_sparsity/linear_layer_sparsity": 0.6956482964847935, "compression/movement_sparsity/model_sparsity": 0.6717506374360657, "compression_loss": 72.79334259033203, "distillation_loss": 4.22519588470459, "epoch": 2.95, "learning_rate": 4.462429555416406e-05, "loss": 76.0876, "step": 3492, "task_loss": 1.3099629878997803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6813943168638301, "compression/movement_sparsity/importance_threshold": -0.0020677970427073096, "compression/movement_sparsity/linear_layer_sparsity": 0.6962185340294758, "compression/movement_sparsity/model_sparsity": 0.6723012855668322, "compression_loss": 72.83494567871094, "distillation_loss": 3.0311317443847656, "epoch": 2.95, "learning_rate": 4.46211646837821e-05, "loss": 75.6562, "step": 3493, "task_loss": 1.8043335676193237 }, { "compression/movement_sparsity/importance_regularization_factor": 0.681788143374404, "compression/movement_sparsity/importance_threshold": -0.0020652410515966417, "compression/movement_sparsity/linear_layer_sparsity": 0.6966221671039498, "compression/movement_sparsity/model_sparsity": 0.6726910526034774, "compression_loss": 72.8765640258789, "distillation_loss": 4.348198413848877, "epoch": 2.95, "learning_rate": 4.461803381340013e-05, "loss": 76.1556, "step": 3494, "task_loss": 3.533262252807617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6821816452135114, "compression/movement_sparsity/importance_threshold": -0.0020626871676508287, "compression/movement_sparsity/linear_layer_sparsity": 0.6970250966525332, "compression/movement_sparsity/model_sparsity": 0.6730801402825107, "compression_loss": 72.91812896728516, "distillation_loss": 2.830470561981201, "epoch": 2.95, "learning_rate": 4.461490294301816e-05, "loss": 75.842, "step": 3495, "task_loss": 1.6345503330230713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6825748225150383, "compression/movement_sparsity/importance_threshold": -0.002060135390000932, "compression/movement_sparsity/linear_layer_sparsity": 0.6975636636079746, "compression/movement_sparsity/model_sparsity": 0.6736002058062069, "compression_loss": 72.95970916748047, "distillation_loss": 4.231863975524902, "epoch": 2.95, "learning_rate": 4.4611772072636194e-05, "loss": 76.3863, "step": 3496, "task_loss": 1.8640209436416626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6829676754128695, "compression/movement_sparsity/importance_threshold": -0.002057585717778019, "compression/movement_sparsity/linear_layer_sparsity": 0.6980023060386275, "compression/movement_sparsity/model_sparsity": 0.674023779519945, "compression_loss": 73.00118255615234, "distillation_loss": 3.31850528717041, "epoch": 2.96, "learning_rate": 4.460864120225423e-05, "loss": 76.328, "step": 3497, "task_loss": 2.039365291595459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.68336020404089, "compression/movement_sparsity/importance_threshold": -0.0020550381501131575, "compression/movement_sparsity/linear_layer_sparsity": 0.6985250973202866, "compression/movement_sparsity/model_sparsity": 0.674528611312785, "compression_loss": 73.04267883300781, "distillation_loss": 2.863635540008545, "epoch": 2.96, "learning_rate": 4.4605510331872265e-05, "loss": 76.1492, "step": 3498, "task_loss": 2.1684441566467285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6837524085329851, "compression/movement_sparsity/importance_threshold": -0.002052492686137413, "compression/movement_sparsity/linear_layer_sparsity": 0.6988918727925981, "compression/movement_sparsity/model_sparsity": 0.6748827869192893, "compression_loss": 73.08415985107422, "distillation_loss": 2.8197097778320312, "epoch": 2.96, "learning_rate": 4.4602379461490297e-05, "loss": 76.2324, "step": 3499, "task_loss": 2.2878568172454834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6841442890230398, "compression/movement_sparsity/importance_threshold": -0.002049949324981851, "compression/movement_sparsity/linear_layer_sparsity": 0.6992652900262828, "compression/movement_sparsity/model_sparsity": 0.6752433761222312, "compression_loss": 73.12556457519531, "distillation_loss": 2.220219373703003, "epoch": 2.96, "learning_rate": 4.459924859110833e-05, "loss": 75.3173, "step": 3500, "task_loss": 1.094734787940979 }, { "epoch": 2.96, "eval_accuracy": 0.6942178217821782, "eval_loss": 75.53413391113281, "eval_runtime": 210.0118, "eval_samples_per_second": 120.231, "eval_steps_per_second": 0.943, "step": 3500 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6845358456449396, "compression/movement_sparsity/importance_threshold": -0.002047408065777536, "compression/movement_sparsity/linear_layer_sparsity": 0.6998073865353444, "compression/movement_sparsity/model_sparsity": 0.6757668499485225, "compression_loss": 73.16688537597656, "distillation_loss": 3.264719009399414, "epoch": 2.96, "learning_rate": 4.459611772072637e-05, "loss": 76.4093, "step": 3501, "task_loss": 1.324640154838562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6849270785325692, "compression/movement_sparsity/importance_threshold": -0.0020448689076555367, "compression/movement_sparsity/linear_layer_sparsity": 0.7002616138530973, "compression/movement_sparsity/model_sparsity": 0.676205473160544, "compression_loss": 73.20825958251953, "distillation_loss": 4.52982759475708, "epoch": 2.96, "learning_rate": 4.45929868503444e-05, "loss": 76.7892, "step": 3502, "task_loss": 2.9776782989501953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6853179878198141, "compression/movement_sparsity/importance_threshold": -0.0020423318497469166, "compression/movement_sparsity/linear_layer_sparsity": 0.7006228565116258, "compression/movement_sparsity/model_sparsity": 0.6765543060224396, "compression_loss": 73.24957275390625, "distillation_loss": 3.085653781890869, "epoch": 2.96, "learning_rate": 4.458985597996243e-05, "loss": 76.0106, "step": 3503, "task_loss": 0.9783570766448975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6857085736405589, "compression/movement_sparsity/importance_threshold": -0.0020397968911827465, "compression/movement_sparsity/linear_layer_sparsity": 0.7010978915019033, "compression/movement_sparsity/model_sparsity": 0.6770130220994225, "compression_loss": 73.29076385498047, "distillation_loss": 4.0569047927856445, "epoch": 2.96, "learning_rate": 4.458672510958046e-05, "loss": 76.4152, "step": 3504, "task_loss": 2.9156129360198975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6860988361286895, "compression/movement_sparsity/importance_threshold": -0.002037264031094086, "compression/movement_sparsity/linear_layer_sparsity": 0.7016865280372476, "compression/movement_sparsity/model_sparsity": 0.67758143715892, "compression_loss": 73.33198547363281, "distillation_loss": 3.6655173301696777, "epoch": 2.96, "learning_rate": 4.45835942391985e-05, "loss": 76.6851, "step": 3505, "task_loss": 2.2414162158966064 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6864887754180904, "compression/movement_sparsity/importance_threshold": -0.0020347332686120054, "compression/movement_sparsity/linear_layer_sparsity": 0.7020799301758901, "compression/movement_sparsity/model_sparsity": 0.6779613247238535, "compression_loss": 73.3731689453125, "distillation_loss": 2.4243812561035156, "epoch": 2.96, "learning_rate": 4.458046336881653e-05, "loss": 76.7262, "step": 3506, "task_loss": 2.2416903972625732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6868783916426471, "compression/movement_sparsity/importance_threshold": -0.0020322046028675694, "compression/movement_sparsity/linear_layer_sparsity": 0.7025590074589961, "compression/movement_sparsity/model_sparsity": 0.6784239442284707, "compression_loss": 73.41429901123047, "distillation_loss": 2.3440120220184326, "epoch": 2.96, "learning_rate": 4.4577332498434564e-05, "loss": 76.5178, "step": 3507, "task_loss": 1.5929702520370483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6872676849362445, "compression/movement_sparsity/importance_threshold": -0.0020296780329918445, "compression/movement_sparsity/linear_layer_sparsity": 0.7028684753816496, "compression/movement_sparsity/model_sparsity": 0.6787227809759465, "compression_loss": 73.45539855957031, "distillation_loss": 4.271585464477539, "epoch": 2.96, "learning_rate": 4.45742016280526e-05, "loss": 76.5856, "step": 3508, "task_loss": 2.353320837020874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6876566554327679, "compression/movement_sparsity/importance_threshold": -0.0020271535581158967, "compression/movement_sparsity/linear_layer_sparsity": 0.7033125552327445, "compression/movement_sparsity/model_sparsity": 0.6791516053180068, "compression_loss": 73.49652099609375, "distillation_loss": 3.085920810699463, "epoch": 2.97, "learning_rate": 4.4571070757670635e-05, "loss": 76.7522, "step": 3509, "task_loss": 2.787026882171631 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6880453032661026, "compression/movement_sparsity/importance_threshold": -0.0020246311773707906, "compression/movement_sparsity/linear_layer_sparsity": 0.7036993156100138, "compression/movement_sparsity/model_sparsity": 0.6795250792865027, "compression_loss": 73.53750610351562, "distillation_loss": 5.123775959014893, "epoch": 2.97, "learning_rate": 4.4567939887288666e-05, "loss": 77.0514, "step": 3510, "task_loss": 4.033083915710449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6884336285701332, "compression/movement_sparsity/importance_threshold": -0.0020221108898875956, "compression/movement_sparsity/linear_layer_sparsity": 0.704146948863064, "compression/movement_sparsity/model_sparsity": 0.6799573349602298, "compression_loss": 73.57854461669922, "distillation_loss": 2.349801540374756, "epoch": 2.97, "learning_rate": 4.45648090169067e-05, "loss": 76.3602, "step": 3511, "task_loss": 3.30352783203125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6888216314787456, "compression/movement_sparsity/importance_threshold": -0.002019592694797372, "compression/movement_sparsity/linear_layer_sparsity": 0.7045827891143225, "compression/movement_sparsity/model_sparsity": 0.680378202758056, "compression_loss": 73.61953735351562, "distillation_loss": 3.060920238494873, "epoch": 2.97, "learning_rate": 4.456167814652474e-05, "loss": 76.5468, "step": 3512, "task_loss": 1.7679400444030762 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6892093121258241, "compression/movement_sparsity/importance_threshold": -0.002017076591231194, "compression/movement_sparsity/linear_layer_sparsity": 0.7048806190493634, "compression/movement_sparsity/model_sparsity": 0.6806658013185963, "compression_loss": 73.66048431396484, "distillation_loss": 4.381338119506836, "epoch": 2.97, "learning_rate": 4.455854727614277e-05, "loss": 77.1019, "step": 3513, "task_loss": 2.6229097843170166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6895966706452548, "compression/movement_sparsity/importance_threshold": -0.0020145625783201185, "compression/movement_sparsity/linear_layer_sparsity": 0.7051954170749501, "compression/movement_sparsity/model_sparsity": 0.6809697850635722, "compression_loss": 73.70143127441406, "distillation_loss": 2.7975080013275146, "epoch": 2.97, "learning_rate": 4.45554164057608e-05, "loss": 76.3846, "step": 3514, "task_loss": 1.2392231225967407 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6899837071709218, "compression/movement_sparsity/importance_threshold": -0.0020120506551952197, "compression/movement_sparsity/linear_layer_sparsity": 0.705742974852789, "compression/movement_sparsity/model_sparsity": 0.6814985325472575, "compression_loss": 73.74237823486328, "distillation_loss": 2.134460926055908, "epoch": 2.97, "learning_rate": 4.455228553537883e-05, "loss": 76.7631, "step": 3515, "task_loss": 1.6460086107254028 }, { "compression/movement_sparsity/importance_regularization_factor": 0.690370421836711, "compression/movement_sparsity/importance_threshold": -0.002009540820987558, "compression/movement_sparsity/linear_layer_sparsity": 0.706163862197832, "compression/movement_sparsity/model_sparsity": 0.6819049611171973, "compression_loss": 73.78324127197266, "distillation_loss": 3.8662548065185547, "epoch": 2.97, "learning_rate": 4.454915466499687e-05, "loss": 76.4187, "step": 3516, "task_loss": 2.281369924545288 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6907568147765073, "compression/movement_sparsity/importance_threshold": -0.0020070330748282027, "compression/movement_sparsity/linear_layer_sparsity": 0.7065901154183113, "compression/movement_sparsity/model_sparsity": 0.6823165712282449, "compression_loss": 73.82408142089844, "distillation_loss": 2.3370680809020996, "epoch": 2.97, "learning_rate": 4.45460237946149e-05, "loss": 76.6981, "step": 3517, "task_loss": 1.5847147703170776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6911428861241957, "compression/movement_sparsity/importance_threshold": -0.0020045274158482186, "compression/movement_sparsity/linear_layer_sparsity": 0.706985306182099, "compression/movement_sparsity/model_sparsity": 0.6826981859735475, "compression_loss": 73.8648681640625, "distillation_loss": 2.4123175144195557, "epoch": 2.97, "learning_rate": 4.4542892924232934e-05, "loss": 76.5859, "step": 3518, "task_loss": 1.2742913961410522 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6915286360136617, "compression/movement_sparsity/importance_threshold": -0.0020020238431786705, "compression/movement_sparsity/linear_layer_sparsity": 0.7074553687944724, "compression/movement_sparsity/model_sparsity": 0.6831521004891041, "compression_loss": 73.90560913085938, "distillation_loss": 2.581087589263916, "epoch": 2.97, "learning_rate": 4.453976205385097e-05, "loss": 76.9698, "step": 3519, "task_loss": 1.7041782140731812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6919140645787898, "compression/movement_sparsity/importance_threshold": -0.0019995223559506276, "compression/movement_sparsity/linear_layer_sparsity": 0.7079323951207451, "compression/movement_sparsity/model_sparsity": 0.6836127394935647, "compression_loss": 73.94634246826172, "distillation_loss": 2.5981240272521973, "epoch": 2.97, "learning_rate": 4.4536631183469005e-05, "loss": 77.1259, "step": 3520, "task_loss": 1.0200672149658203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.692299171953466, "compression/movement_sparsity/importance_threshold": -0.0019970229532951522, "compression/movement_sparsity/linear_layer_sparsity": 0.7083219099807382, "compression/movement_sparsity/model_sparsity": 0.6839888733198292, "compression_loss": 73.9870376586914, "distillation_loss": 2.860292673110962, "epoch": 2.98, "learning_rate": 4.453350031308704e-05, "loss": 76.872, "step": 3521, "task_loss": 1.818167805671692 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6926839582715747, "compression/movement_sparsity/importance_threshold": -0.0019945256343433136, "compression/movement_sparsity/linear_layer_sparsity": 0.7088866981808108, "compression/movement_sparsity/model_sparsity": 0.6845342593077376, "compression_loss": 74.02765655517578, "distillation_loss": 2.705292224884033, "epoch": 2.98, "learning_rate": 4.4530369442705075e-05, "loss": 76.1496, "step": 3522, "task_loss": 1.0087873935699463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6930684236670016, "compression/movement_sparsity/importance_threshold": -0.001992030398226175, "compression/movement_sparsity/linear_layer_sparsity": 0.7092646823707001, "compression/movement_sparsity/model_sparsity": 0.6848992585778887, "compression_loss": 74.06829833984375, "distillation_loss": 2.954558849334717, "epoch": 2.98, "learning_rate": 4.452723857232311e-05, "loss": 77.0909, "step": 3523, "task_loss": 1.8050638437271118 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6934525682736313, "compression/movement_sparsity/importance_threshold": -0.001989537244074805, "compression/movement_sparsity/linear_layer_sparsity": 0.709619521751208, "compression/movement_sparsity/model_sparsity": 0.6852419081340627, "compression_loss": 74.10892486572266, "distillation_loss": 3.013004779815674, "epoch": 2.98, "learning_rate": 4.4524107701941145e-05, "loss": 76.7687, "step": 3524, "task_loss": 1.6245458126068115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6938363922253495, "compression/movement_sparsity/importance_threshold": -0.0019870461710202675, "compression/movement_sparsity/linear_layer_sparsity": 0.7100131862215384, "compression/movement_sparsity/model_sparsity": 0.6856220490187837, "compression_loss": 74.14942932128906, "distillation_loss": 3.5251286029815674, "epoch": 2.98, "learning_rate": 4.452097683155918e-05, "loss": 77.3235, "step": 3525, "task_loss": 2.137641429901123 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6942198956560408, "compression/movement_sparsity/importance_threshold": -0.0019845571781936305, "compression/movement_sparsity/linear_layer_sparsity": 0.7104578742051827, "compression/movement_sparsity/model_sparsity": 0.6860514606021695, "compression_loss": 74.18992614746094, "distillation_loss": 3.193653106689453, "epoch": 2.98, "learning_rate": 4.451784596117721e-05, "loss": 77.2529, "step": 3526, "task_loss": 2.2450642585754395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6946030786995907, "compression/movement_sparsity/importance_threshold": -0.001982070264725959, "compression/movement_sparsity/linear_layer_sparsity": 0.711056336254659, "compression/movement_sparsity/model_sparsity": 0.6866293636391617, "compression_loss": 74.23037719726562, "distillation_loss": 4.018220901489258, "epoch": 2.98, "learning_rate": 4.451471509079525e-05, "loss": 77.1756, "step": 3527, "task_loss": 2.2401669025421143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6949859414898842, "compression/movement_sparsity/importance_threshold": -0.0019795854297483192, "compression/movement_sparsity/linear_layer_sparsity": 0.7115071294121328, "compression/movement_sparsity/model_sparsity": 0.6870646706648743, "compression_loss": 74.27081298828125, "distillation_loss": 2.871913433074951, "epoch": 2.98, "learning_rate": 4.451158422041328e-05, "loss": 77.5523, "step": 3528, "task_loss": 1.8632984161376953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6953684841608067, "compression/movement_sparsity/importance_threshold": -0.0019771026723917765, "compression/movement_sparsity/linear_layer_sparsity": 0.71198872269461, "compression/movement_sparsity/model_sparsity": 0.6875297197365442, "compression_loss": 74.31121063232422, "distillation_loss": 2.115792751312256, "epoch": 2.98, "learning_rate": 4.450845335003131e-05, "loss": 77.4673, "step": 3529, "task_loss": 0.44072240591049194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.695750706846243, "compression/movement_sparsity/importance_threshold": -0.001974621991787397, "compression/movement_sparsity/linear_layer_sparsity": 0.7122646837062069, "compression/movement_sparsity/model_sparsity": 0.6877962006384373, "compression_loss": 74.3515625, "distillation_loss": 2.9175357818603516, "epoch": 2.98, "learning_rate": 4.450532247964935e-05, "loss": 77.8658, "step": 3530, "task_loss": 2.922213554382324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6961326096800781, "compression/movement_sparsity/importance_threshold": -0.0019721433870662497, "compression/movement_sparsity/linear_layer_sparsity": 0.7126099479801032, "compression/movement_sparsity/model_sparsity": 0.6881296040223682, "compression_loss": 74.39190673828125, "distillation_loss": 2.8734731674194336, "epoch": 2.98, "learning_rate": 4.450219160926738e-05, "loss": 77.4339, "step": 3531, "task_loss": 1.9276787042617798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6965141927961978, "compression/movement_sparsity/importance_threshold": -0.0019696668573593954, "compression/movement_sparsity/linear_layer_sparsity": 0.7130769103088913, "compression/movement_sparsity/model_sparsity": 0.6885805247586182, "compression_loss": 74.4322280883789, "distillation_loss": 2.6408824920654297, "epoch": 2.99, "learning_rate": 4.449906073888541e-05, "loss": 77.2825, "step": 3532, "task_loss": 1.2518035173416138 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6968954563284865, "compression/movement_sparsity/importance_threshold": -0.0019671924017979044, "compression/movement_sparsity/linear_layer_sparsity": 0.7135052383345392, "compression/movement_sparsity/model_sparsity": 0.6889941383988939, "compression_loss": 74.4725112915039, "distillation_loss": 2.8946454524993896, "epoch": 2.99, "learning_rate": 4.4495929868503445e-05, "loss": 78.0734, "step": 3533, "task_loss": 1.5071238279342651 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6972764004108299, "compression/movement_sparsity/importance_threshold": -0.0019647200195128407, "compression/movement_sparsity/linear_layer_sparsity": 0.7139514883841437, "compression/movement_sparsity/model_sparsity": 0.6894250583864688, "compression_loss": 74.51275634765625, "distillation_loss": 5.037074089050293, "epoch": 2.99, "learning_rate": 4.4492798998121483e-05, "loss": 79.4306, "step": 3534, "task_loss": 3.1248936653137207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6976570251771128, "compression/movement_sparsity/importance_threshold": -0.0019622497096352717, "compression/movement_sparsity/linear_layer_sparsity": 0.7143852419060658, "compression/movement_sparsity/model_sparsity": 0.689843911140531, "compression_loss": 74.5529556274414, "distillation_loss": 4.346115589141846, "epoch": 2.99, "learning_rate": 4.4489668127739515e-05, "loss": 77.8447, "step": 3535, "task_loss": 1.912516474723816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6980373307612205, "compression/movement_sparsity/importance_threshold": -0.0019597814712962623, "compression/movement_sparsity/linear_layer_sparsity": 0.7146896420574741, "compression/movement_sparsity/model_sparsity": 0.690137854210294, "compression_loss": 74.59321594238281, "distillation_loss": 2.6569290161132812, "epoch": 2.99, "learning_rate": 4.448653725735755e-05, "loss": 77.9436, "step": 3536, "task_loss": 1.3835315704345703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6984173172970382, "compression/movement_sparsity/importance_threshold": -0.0019573153036268783, "compression/movement_sparsity/linear_layer_sparsity": 0.7151400894140865, "compression/movement_sparsity/model_sparsity": 0.6905728273144686, "compression_loss": 74.63337707519531, "distillation_loss": 2.603781223297119, "epoch": 2.99, "learning_rate": 4.448340638697558e-05, "loss": 77.3829, "step": 3537, "task_loss": 1.6564371585845947 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6987969849184509, "compression/movement_sparsity/importance_threshold": -0.001954851205758187, "compression/movement_sparsity/linear_layer_sparsity": 0.7156041899426419, "compression/movement_sparsity/model_sparsity": 0.6910209845621279, "compression_loss": 74.67350769042969, "distillation_loss": 2.7781929969787598, "epoch": 2.99, "learning_rate": 4.448027551659362e-05, "loss": 77.4528, "step": 3538, "task_loss": 1.9029725790023804 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6991763337593437, "compression/movement_sparsity/importance_threshold": -0.0019523891768212535, "compression/movement_sparsity/linear_layer_sparsity": 0.7159700591782131, "compression/movement_sparsity/model_sparsity": 0.6913742850639119, "compression_loss": 74.71356201171875, "distillation_loss": 2.772839069366455, "epoch": 2.99, "learning_rate": 4.447714464621165e-05, "loss": 77.6149, "step": 3539, "task_loss": 2.278661012649536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6995553639536018, "compression/movement_sparsity/importance_threshold": -0.0019499292159471443, "compression/movement_sparsity/linear_layer_sparsity": 0.7163714862816744, "compression/movement_sparsity/model_sparsity": 0.691761921911435, "compression_loss": 74.75360107421875, "distillation_loss": 4.295535087585449, "epoch": 2.99, "learning_rate": 4.447401377582968e-05, "loss": 78.3678, "step": 3540, "task_loss": 1.6903754472732544 }, { "compression/movement_sparsity/importance_regularization_factor": 0.6999340756351107, "compression/movement_sparsity/importance_threshold": -0.0019474713222669242, "compression/movement_sparsity/linear_layer_sparsity": 0.7167762521520737, "compression/movement_sparsity/model_sparsity": 0.6921527828289807, "compression_loss": 74.79360961914062, "distillation_loss": 3.9907779693603516, "epoch": 2.99, "learning_rate": 4.447088290544771e-05, "loss": 77.8046, "step": 3541, "task_loss": 2.940697193145752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7003124689377547, "compression/movement_sparsity/importance_threshold": -0.0019450154949116625, "compression/movement_sparsity/linear_layer_sparsity": 0.7173506870037638, "compression/movement_sparsity/model_sparsity": 0.692707484076347, "compression_loss": 74.83362579345703, "distillation_loss": 3.0434093475341797, "epoch": 2.99, "learning_rate": 4.446775203506575e-05, "loss": 78.1133, "step": 3542, "task_loss": 2.682593584060669 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7006905439954199, "compression/movement_sparsity/importance_threshold": -0.0019425617330124205, "compression/movement_sparsity/linear_layer_sparsity": 0.7177157215476005, "compression/movement_sparsity/model_sparsity": 0.6930599785606252, "compression_loss": 74.8735580444336, "distillation_loss": 4.390281677246094, "epoch": 2.99, "learning_rate": 4.446462116468378e-05, "loss": 77.5065, "step": 3543, "task_loss": 2.8565256595611572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7010683009419907, "compression/movement_sparsity/importance_threshold": -0.0019401100357002692, "compression/movement_sparsity/linear_layer_sparsity": 0.7181882882351774, "compression/movement_sparsity/model_sparsity": 0.6935163111286987, "compression_loss": 74.91343688964844, "distillation_loss": 2.669041872024536, "epoch": 3.0, "learning_rate": 4.4461490294301815e-05, "loss": 77.8592, "step": 3544, "task_loss": 1.6854088306427002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7014457399113527, "compression/movement_sparsity/importance_threshold": -0.00193766040210627, "compression/movement_sparsity/linear_layer_sparsity": 0.718601234084575, "compression/movement_sparsity/model_sparsity": 0.6939150710177994, "compression_loss": 74.9532699584961, "distillation_loss": 3.8379297256469727, "epoch": 3.0, "learning_rate": 4.445835942391985e-05, "loss": 78.3395, "step": 3545, "task_loss": 1.7481732368469238 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7018228610373907, "compression/movement_sparsity/importance_threshold": -0.001935212831361493, "compression/movement_sparsity/linear_layer_sparsity": 0.7190161235732971, "compression/movement_sparsity/model_sparsity": 0.6943157077762346, "compression_loss": 74.99308013916016, "distillation_loss": 3.211388111114502, "epoch": 3.0, "learning_rate": 4.4455228553537885e-05, "loss": 78.0, "step": 3546, "task_loss": 1.6500170230865479 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7021996644539901, "compression/movement_sparsity/importance_threshold": -0.0019327673225970022, "compression/movement_sparsity/linear_layer_sparsity": 0.7194532516346603, "compression/movement_sparsity/model_sparsity": 0.6947378191439267, "compression_loss": 75.03292083740234, "distillation_loss": 3.938809633255005, "epoch": 3.0, "learning_rate": 4.445209768315592e-05, "loss": 78.397, "step": 3547, "task_loss": 2.9108588695526123 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7025761502950361, "compression/movement_sparsity/importance_threshold": -0.0019303238749438624, "compression/movement_sparsity/linear_layer_sparsity": 0.7198173799417567, "compression/movement_sparsity/model_sparsity": 0.6950894385234846, "compression_loss": 75.07262420654297, "distillation_loss": 3.241940498352051, "epoch": 3.0, "learning_rate": 4.444896681277395e-05, "loss": 78.5123, "step": 3548, "task_loss": 1.7612582445144653 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7029523186944135, "compression/movement_sparsity/importance_threshold": -0.0019278824875331411, "compression/movement_sparsity/linear_layer_sparsity": 0.7202036633523204, "compression/movement_sparsity/model_sparsity": 0.6954624519105488, "compression_loss": 75.1123275756836, "distillation_loss": 3.398658275604248, "epoch": 3.0, "learning_rate": 4.444583594239199e-05, "loss": 78.3085, "step": 3549, "task_loss": 2.4252264499664307 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7037037037037036, "compression/movement_sparsity/importance_threshold": -0.0019230058899632214, "compression/movement_sparsity/linear_layer_sparsity": 0.7209359622593416, "compression/movement_sparsity/model_sparsity": 0.696169594097299, "compression_loss": 75.192138671875, "distillation_loss": 4.477725982666016, "epoch": 3.0, "learning_rate": 4.444270507201002e-05, "loss": 137.9143, "step": 3550, "task_loss": 2.224550485610962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7040789205813869, "compression/movement_sparsity/importance_threshold": -0.0019205706780661509, "compression/movement_sparsity/linear_layer_sparsity": 0.7212956905485804, "compression/movement_sparsity/model_sparsity": 0.6965169646131486, "compression_loss": 75.23170471191406, "distillation_loss": 2.2986462116241455, "epoch": 3.0, "learning_rate": 4.443957420162805e-05, "loss": 77.8979, "step": 3551, "task_loss": 1.2223762273788452 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7044538205529419, "compression/movement_sparsity/importance_threshold": -0.0019181375229357662, "compression/movement_sparsity/linear_layer_sparsity": 0.7215581176299105, "compression/movement_sparsity/model_sparsity": 0.6967703765169149, "compression_loss": 75.27127838134766, "distillation_loss": 3.045870065689087, "epoch": 3.0, "learning_rate": 4.443644333124608e-05, "loss": 78.3945, "step": 3552, "task_loss": 1.491921067237854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7048284037522546, "compression/movement_sparsity/importance_threshold": -0.0019157064237031278, "compression/movement_sparsity/linear_layer_sparsity": 0.7218198531095177, "compression/movement_sparsity/model_sparsity": 0.6970231205776051, "compression_loss": 75.31088256835938, "distillation_loss": 3.336623430252075, "epoch": 3.0, "learning_rate": 4.443331246086412e-05, "loss": 78.5461, "step": 3553, "task_loss": 1.8978300094604492 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7052026703132095, "compression/movement_sparsity/importance_threshold": -0.0019132773794993058, "compression/movement_sparsity/linear_layer_sparsity": 0.7223921416110333, "compression/movement_sparsity/model_sparsity": 0.6975757492085283, "compression_loss": 75.3503189086914, "distillation_loss": 3.766842842102051, "epoch": 3.0, "learning_rate": 4.443018159048215e-05, "loss": 78.7087, "step": 3554, "task_loss": 2.5866219997406006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7055766203696922, "compression/movement_sparsity/importance_threshold": -0.0019108503894553625, "compression/movement_sparsity/linear_layer_sparsity": 0.7227893833316544, "compression/movement_sparsity/model_sparsity": 0.6979593444539876, "compression_loss": 75.3897476196289, "distillation_loss": 1.8641656637191772, "epoch": 3.01, "learning_rate": 4.442705072010019e-05, "loss": 78.059, "step": 3555, "task_loss": 1.3319326639175415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7059502540555873, "compression/movement_sparsity/importance_threshold": -0.0019084254527023678, "compression/movement_sparsity/linear_layer_sparsity": 0.7231029531679747, "compression/movement_sparsity/model_sparsity": 0.6982621422017766, "compression_loss": 75.42918395996094, "distillation_loss": 2.7768731117248535, "epoch": 3.01, "learning_rate": 4.442391984971822e-05, "loss": 78.2639, "step": 3556, "task_loss": 1.9650377035140991 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7063235715047802, "compression/movement_sparsity/importance_threshold": -0.0019060025683713868, "compression/movement_sparsity/linear_layer_sparsity": 0.7234131723131894, "compression/movement_sparsity/model_sparsity": 0.6985617043650074, "compression_loss": 75.46857452392578, "distillation_loss": 3.131680488586426, "epoch": 3.01, "learning_rate": 4.442078897933626e-05, "loss": 78.8576, "step": 3557, "task_loss": 1.8010410070419312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7066965728511565, "compression/movement_sparsity/importance_threshold": -0.0019035817355934825, "compression/movement_sparsity/linear_layer_sparsity": 0.7238358482833778, "compression/movement_sparsity/model_sparsity": 0.6989698601153165, "compression_loss": 75.50786590576172, "distillation_loss": 4.752490043640137, "epoch": 3.01, "learning_rate": 4.4417658108954294e-05, "loss": 79.4735, "step": 3558, "task_loss": 3.149632692337036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7070692582286007, "compression/movement_sparsity/importance_threshold": -0.0019011629534997241, "compression/movement_sparsity/linear_layer_sparsity": 0.7240215910426416, "compression/movement_sparsity/model_sparsity": 0.699149222039388, "compression_loss": 75.54719543457031, "distillation_loss": 2.4303901195526123, "epoch": 3.01, "learning_rate": 4.4414527238572325e-05, "loss": 78.5913, "step": 3559, "task_loss": 1.0420986413955688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7074416277709983, "compression/movement_sparsity/importance_threshold": -0.0018987462212211766, "compression/movement_sparsity/linear_layer_sparsity": 0.7243874722023804, "compression/movement_sparsity/model_sparsity": 0.6995025340557077, "compression_loss": 75.58644104003906, "distillation_loss": 5.05695915222168, "epoch": 3.01, "learning_rate": 4.4411396368190364e-05, "loss": 78.8881, "step": 3560, "task_loss": 2.9575765132904053 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7078136816122341, "compression/movement_sparsity/importance_threshold": -0.0018963315378889064, "compression/movement_sparsity/linear_layer_sparsity": 0.7246876511984456, "compression/movement_sparsity/model_sparsity": 0.6997924009797994, "compression_loss": 75.62561798095703, "distillation_loss": 3.978353500366211, "epoch": 3.01, "learning_rate": 4.4408265497808396e-05, "loss": 78.7751, "step": 3561, "task_loss": 2.339336395263672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7081854198861938, "compression/movement_sparsity/importance_threshold": -0.0018939189026339785, "compression/movement_sparsity/linear_layer_sparsity": 0.7249872459102967, "compression/movement_sparsity/model_sparsity": 0.7000817036916374, "compression_loss": 75.6648178100586, "distillation_loss": 2.4307589530944824, "epoch": 3.01, "learning_rate": 4.440513462742643e-05, "loss": 79.0102, "step": 3562, "task_loss": 1.7913055419921875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7085568427267619, "compression/movement_sparsity/importance_threshold": -0.0018915083145874603, "compression/movement_sparsity/linear_layer_sparsity": 0.7253363259178366, "compression/movement_sparsity/model_sparsity": 0.7004187917270225, "compression_loss": 75.7039566040039, "distillation_loss": 3.564100980758667, "epoch": 3.01, "learning_rate": 4.440200375704446e-05, "loss": 78.6702, "step": 3563, "task_loss": 2.6619672775268555 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7089279502678241, "compression/movement_sparsity/importance_threshold": -0.001889099772880416, "compression/movement_sparsity/linear_layer_sparsity": 0.725709456971498, "compression/movement_sparsity/model_sparsity": 0.7007791045811053, "compression_loss": 75.7430191040039, "distillation_loss": 3.104269504547119, "epoch": 3.01, "learning_rate": 4.43988728866625e-05, "loss": 78.808, "step": 3564, "task_loss": 2.1763858795166016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.709298742643265, "compression/movement_sparsity/importance_threshold": -0.0018866932766439144, "compression/movement_sparsity/linear_layer_sparsity": 0.7261858274685508, "compression/movement_sparsity/model_sparsity": 0.7012391102860972, "compression_loss": 75.78218078613281, "distillation_loss": 2.5635428428649902, "epoch": 3.01, "learning_rate": 4.439574201628053e-05, "loss": 78.5337, "step": 3565, "task_loss": 1.1061581373214722 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7096692199869703, "compression/movement_sparsity/importance_threshold": -0.001884288825009019, "compression/movement_sparsity/linear_layer_sparsity": 0.7265532468459146, "compression/movement_sparsity/model_sparsity": 0.7015939076775344, "compression_loss": 75.82121276855469, "distillation_loss": 3.045382499694824, "epoch": 3.01, "learning_rate": 4.439261114589856e-05, "loss": 78.9692, "step": 3566, "task_loss": 2.1024227142333984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7100393824328246, "compression/movement_sparsity/importance_threshold": -0.001881886417106798, "compression/movement_sparsity/linear_layer_sparsity": 0.7269741461151252, "compression/movement_sparsity/model_sparsity": 0.70200034776201, "compression_loss": 75.86022186279297, "distillation_loss": 3.1634793281555176, "epoch": 3.02, "learning_rate": 4.43894802755166e-05, "loss": 79.167, "step": 3567, "task_loss": 1.5554254055023193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7104092301147134, "compression/movement_sparsity/importance_threshold": -0.0018794860520683155, "compression/movement_sparsity/linear_layer_sparsity": 0.7273990161321587, "compression/movement_sparsity/model_sparsity": 0.7024106221869053, "compression_loss": 75.89920043945312, "distillation_loss": 2.642212390899658, "epoch": 3.02, "learning_rate": 4.438634940513463e-05, "loss": 79.3303, "step": 3568, "task_loss": 3.131093978881836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7107787631665219, "compression/movement_sparsity/importance_threshold": -0.001877087729024638, "compression/movement_sparsity/linear_layer_sparsity": 0.7276161552248077, "compression/movement_sparsity/model_sparsity": 0.7026203018837239, "compression_loss": 75.93812561035156, "distillation_loss": 4.578335762023926, "epoch": 3.02, "learning_rate": 4.4383218534752664e-05, "loss": 79.3734, "step": 3569, "task_loss": 2.2910420894622803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7111479817221347, "compression/movement_sparsity/importance_threshold": -0.0018746914471068337, "compression/movement_sparsity/linear_layer_sparsity": 0.7280066240182117, "compression/movement_sparsity/model_sparsity": 0.7029973568728519, "compression_loss": 75.97703552246094, "distillation_loss": 2.911383867263794, "epoch": 3.02, "learning_rate": 4.4380087664370695e-05, "loss": 78.8617, "step": 3570, "task_loss": 2.3898024559020996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7115168859154377, "compression/movement_sparsity/importance_threshold": -0.0018722972054459643, "compression/movement_sparsity/linear_layer_sparsity": 0.7283742222580901, "compression/movement_sparsity/model_sparsity": 0.7033523269823261, "compression_loss": 76.01588439941406, "distillation_loss": 3.844820976257324, "epoch": 3.02, "learning_rate": 4.4376956793988734e-05, "loss": 79.3759, "step": 3571, "task_loss": 3.2330288887023926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7118854758803155, "compression/movement_sparsity/importance_threshold": -0.0018699050031731005, "compression/movement_sparsity/linear_layer_sparsity": 0.7288508550868308, "compression/movement_sparsity/model_sparsity": 0.7038125860071055, "compression_loss": 76.05479431152344, "distillation_loss": 3.176790714263916, "epoch": 3.02, "learning_rate": 4.4373825923606766e-05, "loss": 79.2416, "step": 3572, "task_loss": 2.46520733833313 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7122537517506536, "compression/movement_sparsity/importance_threshold": -0.0018675148394193038, "compression/movement_sparsity/linear_layer_sparsity": 0.7291628032363526, "compression/movement_sparsity/model_sparsity": 0.7041138177780264, "compression_loss": 76.09361267089844, "distillation_loss": 5.269916534423828, "epoch": 3.02, "learning_rate": 4.43706950532248e-05, "loss": 79.6783, "step": 3573, "task_loss": 4.3259429931640625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7126217136603368, "compression/movement_sparsity/importance_threshold": -0.0018651267133156443, "compression/movement_sparsity/linear_layer_sparsity": 0.7295040609899234, "compression/movement_sparsity/model_sparsity": 0.7044433522779304, "compression_loss": 76.13240051269531, "distillation_loss": 3.723994731903076, "epoch": 3.02, "learning_rate": 4.436756418284283e-05, "loss": 79.4293, "step": 3574, "task_loss": 2.431001663208008 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7129893617432504, "compression/movement_sparsity/importance_threshold": -0.0018627406239931851, "compression/movement_sparsity/linear_layer_sparsity": 0.7299178653793907, "compression/movement_sparsity/model_sparsity": 0.7048429412136084, "compression_loss": 76.1711654663086, "distillation_loss": 3.392322063446045, "epoch": 3.02, "learning_rate": 4.436443331246087e-05, "loss": 79.5901, "step": 3575, "task_loss": 2.1494698524475098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7133566961332795, "compression/movement_sparsity/importance_threshold": -0.0018603565705829954, "compression/movement_sparsity/linear_layer_sparsity": 0.7303075233293955, "compression/movement_sparsity/model_sparsity": 0.7052192132143024, "compression_loss": 76.2099380493164, "distillation_loss": 3.890594244003296, "epoch": 3.02, "learning_rate": 4.43613024420789e-05, "loss": 79.2564, "step": 3576, "task_loss": 3.4002721309661865 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7137237169643091, "compression/movement_sparsity/importance_threshold": -0.0018579745522161392, "compression/movement_sparsity/linear_layer_sparsity": 0.7306921015839873, "compression/movement_sparsity/model_sparsity": 0.7055905800227479, "compression_loss": 76.24860382080078, "distillation_loss": 3.0819597244262695, "epoch": 3.02, "learning_rate": 4.435817157169693e-05, "loss": 79.744, "step": 3577, "task_loss": 1.1270776987075806 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7140904243702249, "compression/movement_sparsity/importance_threshold": -0.0018555945680236814, "compression/movement_sparsity/linear_layer_sparsity": 0.7309685872589601, "compression/movement_sparsity/model_sparsity": 0.7058575675642159, "compression_loss": 76.28731536865234, "distillation_loss": 3.532421588897705, "epoch": 3.02, "learning_rate": 4.435504070131496e-05, "loss": 79.3156, "step": 3578, "task_loss": 1.0204206705093384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7144568184849114, "compression/movement_sparsity/importance_threshold": -0.0018532166171366894, "compression/movement_sparsity/linear_layer_sparsity": 0.7313384749390246, "compression/movement_sparsity/model_sparsity": 0.7062147484645626, "compression_loss": 76.3259048461914, "distillation_loss": 3.268002986907959, "epoch": 3.03, "learning_rate": 4.4351909830933e-05, "loss": 79.4378, "step": 3579, "task_loss": 2.5027732849121094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7148228994422541, "compression/movement_sparsity/importance_threshold": -0.001850840698686229, "compression/movement_sparsity/linear_layer_sparsity": 0.7317478435381314, "compression/movement_sparsity/model_sparsity": 0.706610053992925, "compression_loss": 76.36453247070312, "distillation_loss": 3.6075427532196045, "epoch": 3.03, "learning_rate": 4.4348778960551033e-05, "loss": 80.0413, "step": 3580, "task_loss": 2.1746864318847656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7151886673761378, "compression/movement_sparsity/importance_threshold": -0.0018484668118033684, "compression/movement_sparsity/linear_layer_sparsity": 0.7320803489526574, "compression/movement_sparsity/model_sparsity": 0.7069311368235557, "compression_loss": 76.40311431884766, "distillation_loss": 4.750518798828125, "epoch": 3.03, "learning_rate": 4.4345648090169065e-05, "loss": 80.0529, "step": 3581, "task_loss": 2.503535747528076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7155541224204482, "compression/movement_sparsity/importance_threshold": -0.0018460949556191692, "compression/movement_sparsity/linear_layer_sparsity": 0.7325848843336661, "compression/movement_sparsity/model_sparsity": 0.7074183398620943, "compression_loss": 76.44166564941406, "distillation_loss": 4.435973167419434, "epoch": 3.03, "learning_rate": 4.4342517219787104e-05, "loss": 80.189, "step": 3582, "task_loss": 1.866907000541687 }, { "compression/movement_sparsity/importance_regularization_factor": 0.71591926470907, "compression/movement_sparsity/importance_threshold": -0.0018437251292647006, "compression/movement_sparsity/linear_layer_sparsity": 0.7328573515641456, "compression/movement_sparsity/model_sparsity": 0.7076814470049996, "compression_loss": 76.48021697998047, "distillation_loss": 3.8892765045166016, "epoch": 3.03, "learning_rate": 4.4339386349405136e-05, "loss": 80.3519, "step": 3583, "task_loss": 2.674386501312256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7162840943758886, "compression/movement_sparsity/importance_threshold": -0.0018413573318710273, "compression/movement_sparsity/linear_layer_sparsity": 0.7332012445587638, "compression/movement_sparsity/model_sparsity": 0.7080135262173142, "compression_loss": 76.51872253417969, "distillation_loss": 6.001250267028809, "epoch": 3.03, "learning_rate": 4.433625547902317e-05, "loss": 80.4919, "step": 3584, "task_loss": 3.027952194213867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7166486115547885, "compression/movement_sparsity/importance_threshold": -0.0018389915625692185, "compression/movement_sparsity/linear_layer_sparsity": 0.7335777978485367, "compression/movement_sparsity/model_sparsity": 0.70837714374317, "compression_loss": 76.55717468261719, "distillation_loss": 3.8857197761535645, "epoch": 3.03, "learning_rate": 4.43331246086412e-05, "loss": 80.5512, "step": 3585, "task_loss": 2.725917100906372 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7170128163796559, "compression/movement_sparsity/importance_threshold": -0.001836627820490335, "compression/movement_sparsity/linear_layer_sparsity": 0.7338257966870274, "compression/movement_sparsity/model_sparsity": 0.7086166230586248, "compression_loss": 76.59563446044922, "distillation_loss": 4.7574286460876465, "epoch": 3.03, "learning_rate": 4.432999373825924e-05, "loss": 79.6996, "step": 3586, "task_loss": 2.382452964782715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.717376708984375, "compression/movement_sparsity/importance_threshold": -0.0018342661047654474, "compression/movement_sparsity/linear_layer_sparsity": 0.7340709456494531, "compression/movement_sparsity/model_sparsity": 0.7088533504000247, "compression_loss": 76.63407897949219, "distillation_loss": 2.4635610580444336, "epoch": 3.03, "learning_rate": 4.432686286787727e-05, "loss": 79.8647, "step": 3587, "task_loss": 2.811174154281616 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7177402895028313, "compression/movement_sparsity/importance_threshold": -0.00183190641452562, "compression/movement_sparsity/linear_layer_sparsity": 0.7343791973070078, "compression/movement_sparsity/model_sparsity": 0.7091510126648495, "compression_loss": 76.67244720458984, "distillation_loss": 3.478736639022827, "epoch": 3.03, "learning_rate": 4.432373199749531e-05, "loss": 79.924, "step": 3588, "task_loss": 1.9901273250579834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7181035580689104, "compression/movement_sparsity/importance_threshold": -0.0018295487489019165, "compression/movement_sparsity/linear_layer_sparsity": 0.7346927909916634, "compression/movement_sparsity/model_sparsity": 0.7094538334417101, "compression_loss": 76.71080017089844, "distillation_loss": 3.0574965476989746, "epoch": 3.03, "learning_rate": 4.432060112711334e-05, "loss": 79.3303, "step": 3589, "task_loss": 1.308659315109253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7184665148164965, "compression/movement_sparsity/importance_threshold": -0.0018271931070254072, "compression/movement_sparsity/linear_layer_sparsity": 0.7350477973105183, "compression/movement_sparsity/model_sparsity": 0.7097966442013851, "compression_loss": 76.74909973144531, "distillation_loss": 3.3358030319213867, "epoch": 3.03, "learning_rate": 4.431747025673137e-05, "loss": 79.7473, "step": 3590, "task_loss": 1.9157742261886597 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7188291598794756, "compression/movement_sparsity/importance_threshold": -0.0018248394880271543, "compression/movement_sparsity/linear_layer_sparsity": 0.7355050175943478, "compression/movement_sparsity/model_sparsity": 0.710238157561891, "compression_loss": 76.78739166259766, "distillation_loss": 2.5562069416046143, "epoch": 3.04, "learning_rate": 4.431433938634941e-05, "loss": 80.1827, "step": 3591, "task_loss": 1.9325644969940186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7191914933917322, "compression/movement_sparsity/importance_threshold": -0.0018224878910382278, "compression/movement_sparsity/linear_layer_sparsity": 0.7358003673025205, "compression/movement_sparsity/model_sparsity": 0.710523361098986, "compression_loss": 76.8255615234375, "distillation_loss": 3.8533387184143066, "epoch": 3.04, "learning_rate": 4.431120851596744e-05, "loss": 79.9994, "step": 3592, "task_loss": 2.904550313949585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.719553515487152, "compression/movement_sparsity/importance_threshold": -0.0018201383151896901, "compression/movement_sparsity/linear_layer_sparsity": 0.7362236156327555, "compression/movement_sparsity/model_sparsity": 0.7109320695470132, "compression_loss": 76.86372375488281, "distillation_loss": 2.5310556888580322, "epoch": 3.04, "learning_rate": 4.430807764558548e-05, "loss": 79.8879, "step": 3593, "task_loss": 1.107737421989441 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7199152262996196, "compression/movement_sparsity/importance_threshold": -0.0018177907596126095, "compression/movement_sparsity/linear_layer_sparsity": 0.736587278897314, "compression/movement_sparsity/model_sparsity": 0.7112832398596751, "compression_loss": 76.90182495117188, "distillation_loss": 3.9728596210479736, "epoch": 3.04, "learning_rate": 4.430494677520351e-05, "loss": 80.0161, "step": 3594, "task_loss": 1.8572592735290527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7202766259630206, "compression/movement_sparsity/importance_threshold": -0.0018154452234380508, "compression/movement_sparsity/linear_layer_sparsity": 0.7370370346522035, "compression/movement_sparsity/model_sparsity": 0.7117175451207736, "compression_loss": 76.93992614746094, "distillation_loss": 3.798469066619873, "epoch": 3.04, "learning_rate": 4.4301815904821544e-05, "loss": 80.159, "step": 3595, "task_loss": 1.8126416206359863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7206377146112397, "compression/movement_sparsity/importance_threshold": -0.0018131017057970815, "compression/movement_sparsity/linear_layer_sparsity": 0.7374322850368294, "compression/movement_sparsity/model_sparsity": 0.7120992174387553, "compression_loss": 76.97797393798828, "distillation_loss": 3.0008544921875, "epoch": 3.04, "learning_rate": 4.4298685034439576e-05, "loss": 80.1493, "step": 3596, "task_loss": 1.3953158855438232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7209984923781626, "compression/movement_sparsity/importance_threshold": -0.0018107602058207647, "compression/movement_sparsity/linear_layer_sparsity": 0.7378746835802876, "compression/movement_sparsity/model_sparsity": 0.7125264182312685, "compression_loss": 77.01590728759766, "distillation_loss": 2.7801480293273926, "epoch": 3.04, "learning_rate": 4.4295554164057614e-05, "loss": 79.5896, "step": 3597, "task_loss": 3.260056972503662 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7213589593976738, "compression/movement_sparsity/importance_threshold": -0.0018084207226401706, "compression/movement_sparsity/linear_layer_sparsity": 0.7383018788100101, "compression/movement_sparsity/model_sparsity": 0.7129389379906437, "compression_loss": 77.05386352539062, "distillation_loss": 2.7070093154907227, "epoch": 3.04, "learning_rate": 4.4292423293675646e-05, "loss": 79.7216, "step": 3598, "task_loss": 2.8922271728515625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7217191158036588, "compression/movement_sparsity/importance_threshold": -0.001806083255386363, "compression/movement_sparsity/linear_layer_sparsity": 0.7386246183312423, "compression/movement_sparsity/model_sparsity": 0.7132505904164588, "compression_loss": 77.09180450439453, "distillation_loss": 3.265291690826416, "epoch": 3.04, "learning_rate": 4.428929242329368e-05, "loss": 80.7942, "step": 3599, "task_loss": 1.9105757474899292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.722078961730003, "compression/movement_sparsity/importance_threshold": -0.0018037478031904053, "compression/movement_sparsity/linear_layer_sparsity": 0.7389951499163592, "compression/movement_sparsity/model_sparsity": 0.7136083931017383, "compression_loss": 77.12969970703125, "distillation_loss": 3.6068105697631836, "epoch": 3.04, "learning_rate": 4.428616155291171e-05, "loss": 80.269, "step": 3600, "task_loss": 2.2278811931610107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7224384973105908, "compression/movement_sparsity/importance_threshold": -0.001801414365183369, "compression/movement_sparsity/linear_layer_sparsity": 0.7394678716181152, "compression/movement_sparsity/model_sparsity": 0.714064875358777, "compression_loss": 77.16756439208984, "distillation_loss": 2.752007484436035, "epoch": 3.04, "learning_rate": 4.428303068252975e-05, "loss": 79.9602, "step": 3601, "task_loss": 2.7462007999420166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7227977226793083, "compression/movement_sparsity/importance_threshold": -0.001799082940496315, "compression/movement_sparsity/linear_layer_sparsity": 0.7399603635898117, "compression/movement_sparsity/model_sparsity": 0.7145404487161632, "compression_loss": 77.2054214477539, "distillation_loss": 4.3529052734375, "epoch": 3.04, "learning_rate": 4.427989981214778e-05, "loss": 81.2421, "step": 3602, "task_loss": 2.9055633544921875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7231566379700398, "compression/movement_sparsity/importance_threshold": -0.0017967535282603139, "compression/movement_sparsity/linear_layer_sparsity": 0.7402950511270151, "compression/movement_sparsity/model_sparsity": 0.7148636387068444, "compression_loss": 77.24327087402344, "distillation_loss": 3.955024242401123, "epoch": 3.05, "learning_rate": 4.427676894176581e-05, "loss": 80.7427, "step": 3603, "task_loss": 2.586437463760376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7235152433166709, "compression/movement_sparsity/importance_threshold": -0.001794426127606428, "compression/movement_sparsity/linear_layer_sparsity": 0.7405455898132122, "compression/movement_sparsity/model_sparsity": 0.7151055706184235, "compression_loss": 77.28106689453125, "distillation_loss": 3.918201446533203, "epoch": 3.05, "learning_rate": 4.427363807138385e-05, "loss": 80.5166, "step": 3604, "task_loss": 2.3276305198669434 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7238735388530865, "compression/movement_sparsity/importance_threshold": -0.0017921007376657258, "compression/movement_sparsity/linear_layer_sparsity": 0.7408995706536504, "compression/movement_sparsity/model_sparsity": 0.7154473911280202, "compression_loss": 77.31881713867188, "distillation_loss": 2.534377336502075, "epoch": 3.05, "learning_rate": 4.427050720100188e-05, "loss": 80.1353, "step": 3605, "task_loss": 1.0602240562438965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7242315247131721, "compression/movement_sparsity/importance_threshold": -0.0017897773575692713, "compression/movement_sparsity/linear_layer_sparsity": 0.7412304901538809, "compression/movement_sparsity/model_sparsity": 0.7157669425253903, "compression_loss": 77.35657501220703, "distillation_loss": 3.181241035461426, "epoch": 3.05, "learning_rate": 4.4267376330619914e-05, "loss": 80.2621, "step": 3606, "task_loss": 1.4473333358764648 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7245892010308121, "compression/movement_sparsity/importance_threshold": -0.0017874559864481335, "compression/movement_sparsity/linear_layer_sparsity": 0.7415379071197011, "compression/movement_sparsity/model_sparsity": 0.7160637987727094, "compression_loss": 77.39427185058594, "distillation_loss": 6.215128421783447, "epoch": 3.05, "learning_rate": 4.4264245460237946e-05, "loss": 81.0523, "step": 3607, "task_loss": 2.7484066486358643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7249465679398923, "compression/movement_sparsity/importance_threshold": -0.0017851366234333766, "compression/movement_sparsity/linear_layer_sparsity": 0.7419823804683279, "compression/movement_sparsity/model_sparsity": 0.7164930030944509, "compression_loss": 77.43197631835938, "distillation_loss": 5.144721984863281, "epoch": 3.05, "learning_rate": 4.4261114589855984e-05, "loss": 80.8959, "step": 3608, "task_loss": 2.2884976863861084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.725303625574298, "compression/movement_sparsity/importance_threshold": -0.0017828192676560646, "compression/movement_sparsity/linear_layer_sparsity": 0.742273163220296, "compression/movement_sparsity/model_sparsity": 0.7167737965643366, "compression_loss": 77.46965789794922, "distillation_loss": 1.9961227178573608, "epoch": 3.05, "learning_rate": 4.4257983719474016e-05, "loss": 80.2964, "step": 3609, "task_loss": 1.066470980644226 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7256603740679137, "compression/movement_sparsity/importance_threshold": -0.0017805039182472665, "compression/movement_sparsity/linear_layer_sparsity": 0.7426088523875809, "compression/movement_sparsity/model_sparsity": 0.7170979537760245, "compression_loss": 77.50731658935547, "distillation_loss": 3.7969813346862793, "epoch": 3.05, "learning_rate": 4.425485284909205e-05, "loss": 80.6433, "step": 3610, "task_loss": 1.7196377515792847 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7260168135546251, "compression/movement_sparsity/importance_threshold": -0.0017781905743380465, "compression/movement_sparsity/linear_layer_sparsity": 0.7429856441607064, "compression/movement_sparsity/model_sparsity": 0.7174618015925962, "compression_loss": 77.54489135742188, "distillation_loss": 2.9429688453674316, "epoch": 3.05, "learning_rate": 4.425172197871008e-05, "loss": 80.1422, "step": 3611, "task_loss": 1.6612135171890259 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7263729441683169, "compression/movement_sparsity/importance_threshold": -0.001775879235059473, "compression/movement_sparsity/linear_layer_sparsity": 0.7434042778380664, "compression/movement_sparsity/model_sparsity": 0.7178660539152709, "compression_loss": 77.58243560791016, "distillation_loss": 4.491946697235107, "epoch": 3.05, "learning_rate": 4.424859110832812e-05, "loss": 80.9363, "step": 3612, "task_loss": 2.8611388206481934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7267287660428745, "compression/movement_sparsity/importance_threshold": -0.0017735698995426097, "compression/movement_sparsity/linear_layer_sparsity": 0.7436950844383698, "compression/movement_sparsity/model_sparsity": 0.7181468704142282, "compression_loss": 77.61996459960938, "distillation_loss": 2.316256284713745, "epoch": 3.05, "learning_rate": 4.424546023794615e-05, "loss": 81.0116, "step": 3613, "task_loss": 1.3472588062286377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7270842793121828, "compression/movement_sparsity/importance_threshold": -0.0017712625669185252, "compression/movement_sparsity/linear_layer_sparsity": 0.7440707195672348, "compression/movement_sparsity/model_sparsity": 0.7185096013208278, "compression_loss": 77.65746307373047, "distillation_loss": 3.6571590900421143, "epoch": 3.05, "learning_rate": 4.424232936756418e-05, "loss": 80.2153, "step": 3614, "task_loss": 2.135869026184082 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7274394841101275, "compression/movement_sparsity/importance_threshold": -0.0017689572363182807, "compression/movement_sparsity/linear_layer_sparsity": 0.7444206581148445, "compression/movement_sparsity/model_sparsity": 0.7188475184027902, "compression_loss": 77.6948471069336, "distillation_loss": 2.35872745513916, "epoch": 3.06, "learning_rate": 4.423919849718222e-05, "loss": 80.289, "step": 3615, "task_loss": 1.5825508832931519 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7277943805705929, "compression/movement_sparsity/importance_threshold": -0.001766653906872949, "compression/movement_sparsity/linear_layer_sparsity": 0.744724724389559, "compression/movement_sparsity/model_sparsity": 0.719141139065551, "compression_loss": 77.73225402832031, "distillation_loss": 2.9137988090515137, "epoch": 3.06, "learning_rate": 4.423606762680025e-05, "loss": 80.8079, "step": 3616, "task_loss": 1.3599717617034912 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7281489688274649, "compression/movement_sparsity/importance_threshold": -0.00176435257771359, "compression/movement_sparsity/linear_layer_sparsity": 0.7452265052878437, "compression/movement_sparsity/model_sparsity": 0.7196256822463211, "compression_loss": 77.76960754394531, "distillation_loss": 4.037845611572266, "epoch": 3.06, "learning_rate": 4.4232936756418284e-05, "loss": 81.0236, "step": 3617, "task_loss": 2.941525459289551 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7285032490146282, "compression/movement_sparsity/importance_threshold": -0.0017620532479712741, "compression/movement_sparsity/linear_layer_sparsity": 0.7456410012790339, "compression/movement_sparsity/model_sparsity": 0.7200259390250751, "compression_loss": 77.80686950683594, "distillation_loss": 2.7546355724334717, "epoch": 3.06, "learning_rate": 4.4229805886036316e-05, "loss": 80.5209, "step": 3618, "task_loss": 1.7317492961883545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7288572212659681, "compression/movement_sparsity/importance_threshold": -0.0017597559167770649, "compression/movement_sparsity/linear_layer_sparsity": 0.7458633512329399, "compression/movement_sparsity/model_sparsity": 0.7202406505740359, "compression_loss": 77.8442153930664, "distillation_loss": 3.032512664794922, "epoch": 3.06, "learning_rate": 4.4226675015654354e-05, "loss": 82.2569, "step": 3619, "task_loss": 1.0733007192611694 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7292108857153698, "compression/movement_sparsity/importance_threshold": -0.0017574605832620278, "compression/movement_sparsity/linear_layer_sparsity": 0.7462441733747264, "compression/movement_sparsity/model_sparsity": 0.7206083903037062, "compression_loss": 77.88151550292969, "distillation_loss": 3.2697787284851074, "epoch": 3.06, "learning_rate": 4.4223544145272386e-05, "loss": 80.582, "step": 3620, "task_loss": 1.898140549659729 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7295642424967181, "compression/movement_sparsity/importance_threshold": -0.0017551672465572323, "compression/movement_sparsity/linear_layer_sparsity": 0.7464630653200179, "compression/movement_sparsity/model_sparsity": 0.7208197626372865, "compression_loss": 77.918701171875, "distillation_loss": 4.518433570861816, "epoch": 3.06, "learning_rate": 4.422041327489042e-05, "loss": 82.3394, "step": 3621, "task_loss": 2.8371496200561523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7299172917438987, "compression/movement_sparsity/importance_threshold": -0.0017528759057937405, "compression/movement_sparsity/linear_layer_sparsity": 0.7468326548958916, "compression/movement_sparsity/model_sparsity": 0.7211766556742383, "compression_loss": 77.9559097290039, "distillation_loss": 4.580798625946045, "epoch": 3.06, "learning_rate": 4.4217282404508456e-05, "loss": 81.2667, "step": 3622, "task_loss": 2.728023052215576 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7302700335907963, "compression/movement_sparsity/importance_threshold": -0.0017505865601026207, "compression/movement_sparsity/linear_layer_sparsity": 0.7472672192612129, "compression/movement_sparsity/model_sparsity": 0.7215962914167345, "compression_loss": 77.99307250976562, "distillation_loss": 3.4792189598083496, "epoch": 3.06, "learning_rate": 4.421415153412649e-05, "loss": 81.188, "step": 3623, "task_loss": 2.0538363456726074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7306224681712964, "compression/movement_sparsity/importance_threshold": -0.0017482992086149379, "compression/movement_sparsity/linear_layer_sparsity": 0.7475362404072456, "compression/movement_sparsity/model_sparsity": 0.7218560708587952, "compression_loss": 78.03022003173828, "distillation_loss": 4.547062873840332, "epoch": 3.06, "learning_rate": 4.421102066374453e-05, "loss": 81.5491, "step": 3624, "task_loss": 3.381809711456299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7309745956192836, "compression/movement_sparsity/importance_threshold": -0.0017460138504617603, "compression/movement_sparsity/linear_layer_sparsity": 0.7478673506941582, "compression/movement_sparsity/model_sparsity": 0.722175806488738, "compression_loss": 78.06736755371094, "distillation_loss": 3.141965389251709, "epoch": 3.06, "learning_rate": 4.420788979336256e-05, "loss": 81.8488, "step": 3625, "task_loss": 1.4020389318466187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7313264160686437, "compression/movement_sparsity/importance_threshold": -0.0017437304847741503, "compression/movement_sparsity/linear_layer_sparsity": 0.7482394324210677, "compression/movement_sparsity/model_sparsity": 0.7225351060636709, "compression_loss": 78.10445404052734, "distillation_loss": 4.267266273498535, "epoch": 3.07, "learning_rate": 4.420475892298059e-05, "loss": 81.3185, "step": 3626, "task_loss": 2.6754324436187744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7316779296532612, "compression/movement_sparsity/importance_threshold": -0.001741449110683177, "compression/movement_sparsity/linear_layer_sparsity": 0.7486269559450657, "compression/movement_sparsity/model_sparsity": 0.7229093169624576, "compression_loss": 78.14151000976562, "distillation_loss": 3.3589515686035156, "epoch": 3.07, "learning_rate": 4.420162805259863e-05, "loss": 81.1748, "step": 3627, "task_loss": 1.999585509300232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7320291365070214, "compression/movement_sparsity/importance_threshold": -0.001739169727319907, "compression/movement_sparsity/linear_layer_sparsity": 0.7489918235505554, "compression/movement_sparsity/model_sparsity": 0.7232616502432347, "compression_loss": 78.17847442626953, "distillation_loss": 3.133018970489502, "epoch": 3.07, "learning_rate": 4.419849718221666e-05, "loss": 81.442, "step": 3628, "task_loss": 2.3964903354644775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.73238003676381, "compression/movement_sparsity/importance_threshold": -0.0017368923338154018, "compression/movement_sparsity/linear_layer_sparsity": 0.7492888068696942, "compression/movement_sparsity/model_sparsity": 0.7235484312717336, "compression_loss": 78.21541595458984, "distillation_loss": 3.956521987915039, "epoch": 3.07, "learning_rate": 4.419536631183469e-05, "loss": 81.5547, "step": 3629, "task_loss": 2.651926279067993 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7327306305575114, "compression/movement_sparsity/importance_threshold": -0.0017346169293007332, "compression/movement_sparsity/linear_layer_sparsity": 0.7494779360925666, "compression/movement_sparsity/model_sparsity": 0.7237310633239709, "compression_loss": 78.25236511230469, "distillation_loss": 3.3483009338378906, "epoch": 3.07, "learning_rate": 4.419223544145273e-05, "loss": 81.0787, "step": 3630, "task_loss": 1.2515430450439453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7330809180220114, "compression/movement_sparsity/importance_threshold": -0.0017323435129069616, "compression/movement_sparsity/linear_layer_sparsity": 0.7499490480316919, "compression/movement_sparsity/model_sparsity": 0.7241859911186773, "compression_loss": 78.28919982910156, "distillation_loss": 3.8082053661346436, "epoch": 3.07, "learning_rate": 4.418910457107076e-05, "loss": 81.7366, "step": 3631, "task_loss": 3.145416498184204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7334308992911944, "compression/movement_sparsity/importance_threshold": -0.001730072083765158, "compression/movement_sparsity/linear_layer_sparsity": 0.750161536698963, "compression/movement_sparsity/model_sparsity": 0.724391180146536, "compression_loss": 78.32604217529297, "distillation_loss": 4.595346927642822, "epoch": 3.07, "learning_rate": 4.4185973700688794e-05, "loss": 82.6242, "step": 3632, "task_loss": 2.02565860748291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7337805744989463, "compression/movement_sparsity/importance_threshold": -0.0017278026410063848, "compression/movement_sparsity/linear_layer_sparsity": 0.7506908862728219, "compression/movement_sparsity/model_sparsity": 0.7249023449340631, "compression_loss": 78.36288452148438, "distillation_loss": 5.433276176452637, "epoch": 3.07, "learning_rate": 4.4182842830306826e-05, "loss": 82.4369, "step": 3633, "task_loss": 2.8843348026275635 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7341299437791516, "compression/movement_sparsity/importance_threshold": -0.001725535183761711, "compression/movement_sparsity/linear_layer_sparsity": 0.751044807492422, "compression/movement_sparsity/model_sparsity": 0.7252441078709809, "compression_loss": 78.3996353149414, "distillation_loss": 4.625478744506836, "epoch": 3.07, "learning_rate": 4.4179711959924865e-05, "loss": 82.0365, "step": 3634, "task_loss": 3.2120397090911865 }, { "compression/movement_sparsity/importance_regularization_factor": 0.734479007265696, "compression/movement_sparsity/importance_threshold": -0.0017232697111621998, "compression/movement_sparsity/linear_layer_sparsity": 0.7514867409933423, "compression/movement_sparsity/model_sparsity": 0.7256708595965982, "compression_loss": 78.4363784790039, "distillation_loss": 4.03125524520874, "epoch": 3.07, "learning_rate": 4.41765810895429e-05, "loss": 81.6831, "step": 3635, "task_loss": 2.158158540725708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7348277650924642, "compression/movement_sparsity/importance_threshold": -0.0017210062223389196, "compression/movement_sparsity/linear_layer_sparsity": 0.7517871465485926, "compression/movement_sparsity/model_sparsity": 0.72596094529687, "compression_loss": 78.47310638427734, "distillation_loss": 3.0152032375335693, "epoch": 3.07, "learning_rate": 4.417345021916093e-05, "loss": 81.7813, "step": 3636, "task_loss": 1.5624889135360718 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7351762173933416, "compression/movement_sparsity/importance_threshold": -0.0017187447164229343, "compression/movement_sparsity/linear_layer_sparsity": 0.7520464614221697, "compression/movement_sparsity/model_sparsity": 0.7262113519067939, "compression_loss": 78.509765625, "distillation_loss": 2.4387426376342773, "epoch": 3.07, "learning_rate": 4.417031934877896e-05, "loss": 81.7479, "step": 3637, "task_loss": 1.45798659324646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7355243643022131, "compression/movement_sparsity/importance_threshold": -0.0017164851925453124, "compression/movement_sparsity/linear_layer_sparsity": 0.7524807276833002, "compression/movement_sparsity/model_sparsity": 0.7266306997858952, "compression_loss": 78.54641723632812, "distillation_loss": 4.960605621337891, "epoch": 3.08, "learning_rate": 4.4167188478397e-05, "loss": 82.1377, "step": 3638, "task_loss": 2.9077093601226807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7358722059529639, "compression/movement_sparsity/importance_threshold": -0.0017142276498371203, "compression/movement_sparsity/linear_layer_sparsity": 0.7527407699311031, "compression/movement_sparsity/model_sparsity": 0.7268818087825026, "compression_loss": 78.58311462402344, "distillation_loss": 3.8493051528930664, "epoch": 3.08, "learning_rate": 4.416405760801503e-05, "loss": 81.9104, "step": 3639, "task_loss": 1.7890443801879883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7362197424794796, "compression/movement_sparsity/importance_threshold": -0.0017119720874294187, "compression/movement_sparsity/linear_layer_sparsity": 0.7530880732376652, "compression/movement_sparsity/model_sparsity": 0.7272171811520544, "compression_loss": 78.6197280883789, "distillation_loss": 3.4963693618774414, "epoch": 3.08, "learning_rate": 4.416092673763306e-05, "loss": 82.2736, "step": 3640, "task_loss": 2.3061511516571045 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7365669740156446, "compression/movement_sparsity/importance_threshold": -0.0017097185044532793, "compression/movement_sparsity/linear_layer_sparsity": 0.7534272800344026, "compression/movement_sparsity/model_sparsity": 0.7275447351518017, "compression_loss": 78.65621948242188, "distillation_loss": 3.5803723335266113, "epoch": 3.08, "learning_rate": 4.41577958672511e-05, "loss": 82.3065, "step": 3641, "task_loss": 2.5603771209716797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7369139006953448, "compression/movement_sparsity/importance_threshold": -0.0017074669000397653, "compression/movement_sparsity/linear_layer_sparsity": 0.7536776994789234, "compression/movement_sparsity/model_sparsity": 0.7277865519180229, "compression_loss": 78.69279479980469, "distillation_loss": 2.1467814445495605, "epoch": 3.08, "learning_rate": 4.415466499686913e-05, "loss": 82.2824, "step": 3642, "task_loss": 2.1186931133270264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7372605226524647, "compression/movement_sparsity/importance_threshold": -0.0017052172733199441, "compression/movement_sparsity/linear_layer_sparsity": 0.7540574365214551, "compression/movement_sparsity/model_sparsity": 0.7281532438249358, "compression_loss": 78.72930145263672, "distillation_loss": 2.767716884613037, "epoch": 3.08, "learning_rate": 4.4151534126487164e-05, "loss": 82.2642, "step": 3643, "task_loss": 1.7852729558944702 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7376068400208897, "compression/movement_sparsity/importance_threshold": -0.0017029696234248815, "compression/movement_sparsity/linear_layer_sparsity": 0.754507144579674, "compression/movement_sparsity/model_sparsity": 0.7285875030278911, "compression_loss": 78.76571655273438, "distillation_loss": 2.541390895843506, "epoch": 3.08, "learning_rate": 4.4148403256105196e-05, "loss": 81.4408, "step": 3644, "task_loss": 1.2581161260604858 }, { "compression/movement_sparsity/importance_regularization_factor": 0.737952852934505, "compression/movement_sparsity/importance_threshold": -0.0017007239494856424, "compression/movement_sparsity/linear_layer_sparsity": 0.7548059165239582, "compression/movement_sparsity/model_sparsity": 0.7288760112367592, "compression_loss": 78.80220794677734, "distillation_loss": 3.9758291244506836, "epoch": 3.08, "learning_rate": 4.4145272385723235e-05, "loss": 81.8343, "step": 3645, "task_loss": 1.8818309307098389 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7382985615271958, "compression/movement_sparsity/importance_threshold": -0.0016984802506332932, "compression/movement_sparsity/linear_layer_sparsity": 0.7550658037575818, "compression/movement_sparsity/model_sparsity": 0.7291269705444011, "compression_loss": 78.83858489990234, "distillation_loss": 4.803963661193848, "epoch": 3.08, "learning_rate": 4.4142141515341267e-05, "loss": 82.8147, "step": 3646, "task_loss": 2.4096415042877197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7386439659328468, "compression/movement_sparsity/importance_threshold": -0.0016962385259989034, "compression/movement_sparsity/linear_layer_sparsity": 0.7553587924805627, "compression/movement_sparsity/model_sparsity": 0.7294098942034088, "compression_loss": 78.8749771118164, "distillation_loss": 3.9560742378234863, "epoch": 3.08, "learning_rate": 4.41390106449593e-05, "loss": 81.7611, "step": 3647, "task_loss": 2.1173205375671387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7389890662853433, "compression/movement_sparsity/importance_threshold": -0.001693998774713535, "compression/movement_sparsity/linear_layer_sparsity": 0.7557293598381823, "compression/movement_sparsity/model_sparsity": 0.7297677314322958, "compression_loss": 78.9112777709961, "distillation_loss": 3.218472719192505, "epoch": 3.08, "learning_rate": 4.413587977457733e-05, "loss": 82.3833, "step": 3648, "task_loss": 1.2338536977767944 }, { "compression/movement_sparsity/importance_regularization_factor": 0.739333862718571, "compression/movement_sparsity/importance_threshold": -0.001691760995908254, "compression/movement_sparsity/linear_layer_sparsity": 0.755987315356649, "compression/movement_sparsity/model_sparsity": 0.730016825385139, "compression_loss": 78.94752502441406, "distillation_loss": 3.422424077987671, "epoch": 3.08, "learning_rate": 4.413274890419537e-05, "loss": 82.6181, "step": 3649, "task_loss": 2.577562093734741 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7396783553664146, "compression/movement_sparsity/importance_threshold": -0.0016895251887141284, "compression/movement_sparsity/linear_layer_sparsity": 0.7562256675435223, "compression/movement_sparsity/model_sparsity": 0.7302469894411361, "compression_loss": 78.98384857177734, "distillation_loss": 4.544760227203369, "epoch": 3.09, "learning_rate": 4.41296180338134e-05, "loss": 82.7533, "step": 3650, "task_loss": 3.1080689430236816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7400225443627593, "compression/movement_sparsity/importance_threshold": -0.0016872913522622225, "compression/movement_sparsity/linear_layer_sparsity": 0.7566449808984373, "compression/movement_sparsity/model_sparsity": 0.7306518980923512, "compression_loss": 79.02005004882812, "distillation_loss": 3.372283458709717, "epoch": 3.09, "learning_rate": 4.412648716343143e-05, "loss": 82.4046, "step": 3651, "task_loss": 2.383744239807129 }, { "compression/movement_sparsity/importance_regularization_factor": 0.74036642984149, "compression/movement_sparsity/importance_threshold": -0.0016850594856836036, "compression/movement_sparsity/linear_layer_sparsity": 0.7570164783411327, "compression/movement_sparsity/model_sparsity": 0.73101063345503, "compression_loss": 79.05615234375, "distillation_loss": 5.318266868591309, "epoch": 3.09, "learning_rate": 4.412335629304947e-05, "loss": 82.9382, "step": 3652, "task_loss": 3.052161693572998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7407100119364922, "compression/movement_sparsity/importance_threshold": -0.0016828295881093375, "compression/movement_sparsity/linear_layer_sparsity": 0.7574635988549746, "compression/movement_sparsity/model_sparsity": 0.731442394003718, "compression_loss": 79.09239959716797, "distillation_loss": 3.7698822021484375, "epoch": 3.09, "learning_rate": 4.41202254226675e-05, "loss": 81.911, "step": 3653, "task_loss": 2.9255530834198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7410532907816508, "compression/movement_sparsity/importance_threshold": -0.0016806016586704909, "compression/movement_sparsity/linear_layer_sparsity": 0.7577788976956021, "compression/movement_sparsity/model_sparsity": 0.7317468613591972, "compression_loss": 79.12848663330078, "distillation_loss": 3.2716548442840576, "epoch": 3.09, "learning_rate": 4.4117094552285534e-05, "loss": 82.8463, "step": 3654, "task_loss": 3.150697708129883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7413962665108513, "compression/movement_sparsity/importance_threshold": -0.0016783756964981276, "compression/movement_sparsity/linear_layer_sparsity": 0.7581371116155511, "compression/movement_sparsity/model_sparsity": 0.732092769529001, "compression_loss": 79.16453552246094, "distillation_loss": 2.530327320098877, "epoch": 3.09, "learning_rate": 4.411396368190357e-05, "loss": 82.5172, "step": 3655, "task_loss": 1.299980878829956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7417389392579783, "compression/movement_sparsity/importance_threshold": -0.001676151700723317, "compression/movement_sparsity/linear_layer_sparsity": 0.7585439642152867, "compression/movement_sparsity/model_sparsity": 0.7324856454903107, "compression_loss": 79.20056915283203, "distillation_loss": 4.691944122314453, "epoch": 3.09, "learning_rate": 4.4110832811521605e-05, "loss": 83.4421, "step": 3656, "task_loss": 2.584045171737671 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7420813091569174, "compression/movement_sparsity/importance_threshold": -0.001673929670477122, "compression/movement_sparsity/linear_layer_sparsity": 0.7587832584114033, "compression/movement_sparsity/model_sparsity": 0.7327167191946355, "compression_loss": 79.23651885986328, "distillation_loss": 3.3267219066619873, "epoch": 3.09, "learning_rate": 4.4107701941139636e-05, "loss": 83.1716, "step": 3657, "task_loss": 2.096855401992798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7424233763415533, "compression/movement_sparsity/importance_threshold": -0.001671709604890611, "compression/movement_sparsity/linear_layer_sparsity": 0.7591695179736317, "compression/movement_sparsity/model_sparsity": 0.7330897095526282, "compression_loss": 79.27247619628906, "distillation_loss": 5.384352684020996, "epoch": 3.09, "learning_rate": 4.4104571070757675e-05, "loss": 83.2762, "step": 3658, "task_loss": 3.6306674480438232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7427651409457716, "compression/movement_sparsity/importance_threshold": -0.0016694915030948474, "compression/movement_sparsity/linear_layer_sparsity": 0.7594971464035948, "compression/movement_sparsity/model_sparsity": 0.7334060829381189, "compression_loss": 79.3084487915039, "distillation_loss": 4.304891586303711, "epoch": 3.09, "learning_rate": 4.410144020037571e-05, "loss": 84.0403, "step": 3659, "task_loss": 2.5461270809173584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7431066031034573, "compression/movement_sparsity/importance_threshold": -0.0016672753642208993, "compression/movement_sparsity/linear_layer_sparsity": 0.7596897694075845, "compression/movement_sparsity/model_sparsity": 0.733592088749344, "compression_loss": 79.34426879882812, "distillation_loss": 4.082357406616211, "epoch": 3.09, "learning_rate": 4.4098309329993745e-05, "loss": 83.1144, "step": 3660, "task_loss": 1.9327523708343506 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7434477629484952, "compression/movement_sparsity/importance_threshold": -0.0016650611873998342, "compression/movement_sparsity/linear_layer_sparsity": 0.7600288450384779, "compression/movement_sparsity/model_sparsity": 0.7339195160891976, "compression_loss": 79.38018798828125, "distillation_loss": 3.8783211708068848, "epoch": 3.09, "learning_rate": 4.409517845961178e-05, "loss": 82.8906, "step": 3661, "task_loss": 2.8718295097351074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7437886206147711, "compression/movement_sparsity/importance_threshold": -0.0016628489717627126, "compression/movement_sparsity/linear_layer_sparsity": 0.7604045278640134, "compression/movement_sparsity/model_sparsity": 0.7342822930539403, "compression_loss": 79.416015625, "distillation_loss": 4.645908355712891, "epoch": 3.1, "learning_rate": 4.409204758922981e-05, "loss": 83.2292, "step": 3662, "task_loss": 3.1600096225738525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7441291762361695, "compression/movement_sparsity/importance_threshold": -0.0016606387164406072, "compression/movement_sparsity/linear_layer_sparsity": 0.7606623760649712, "compression/movement_sparsity/model_sparsity": 0.7345312833759615, "compression_loss": 79.45179748535156, "distillation_loss": 3.173125743865967, "epoch": 3.1, "learning_rate": 4.408891671884784e-05, "loss": 83.4151, "step": 3663, "task_loss": 2.545177936553955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7444694299465761, "compression/movement_sparsity/importance_threshold": -0.0016584304205645786, "compression/movement_sparsity/linear_layer_sparsity": 0.7611267746977176, "compression/movement_sparsity/model_sparsity": 0.7349797284870156, "compression_loss": 79.48751068115234, "distillation_loss": 3.406477689743042, "epoch": 3.1, "learning_rate": 4.408578584846588e-05, "loss": 83.5961, "step": 3664, "task_loss": 3.434521436691284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7448093818798753, "compression/movement_sparsity/importance_threshold": -0.0016562240832656968, "compression/movement_sparsity/linear_layer_sparsity": 0.76156955481454, "compression/movement_sparsity/model_sparsity": 0.7354072977446744, "compression_loss": 79.52322387695312, "distillation_loss": 2.5624570846557617, "epoch": 3.1, "learning_rate": 4.408265497808391e-05, "loss": 82.5398, "step": 3665, "task_loss": 2.164654016494751 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7451490321699531, "compression/movement_sparsity/importance_threshold": -0.001654019703675025, "compression/movement_sparsity/linear_layer_sparsity": 0.7619380592911588, "compression/movement_sparsity/model_sparsity": 0.7357631429588688, "compression_loss": 79.55890655517578, "distillation_loss": 3.937011241912842, "epoch": 3.1, "learning_rate": 4.407952410770194e-05, "loss": 83.2215, "step": 3666, "task_loss": 1.3899091482162476 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7454883809506939, "compression/movement_sparsity/importance_threshold": -0.0016518172809236323, "compression/movement_sparsity/linear_layer_sparsity": 0.7623019014182318, "compression/movement_sparsity/model_sparsity": 0.7361144859895676, "compression_loss": 79.5946044921875, "distillation_loss": 4.280656337738037, "epoch": 3.1, "learning_rate": 4.407639323731998e-05, "loss": 83.6135, "step": 3667, "task_loss": 2.2733960151672363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7458274283559831, "compression/movement_sparsity/importance_threshold": -0.0016496168141425828, "compression/movement_sparsity/linear_layer_sparsity": 0.7625494113658495, "compression/movement_sparsity/model_sparsity": 0.736353493209055, "compression_loss": 79.63018798828125, "distillation_loss": 3.334233283996582, "epoch": 3.1, "learning_rate": 4.407326236693801e-05, "loss": 83.1563, "step": 3668, "task_loss": 1.1691131591796875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7461661745197063, "compression/movement_sparsity/importance_threshold": -0.0016474183024629404, "compression/movement_sparsity/linear_layer_sparsity": 0.7628626354013084, "compression/movement_sparsity/model_sparsity": 0.736655957035306, "compression_loss": 79.66580963134766, "distillation_loss": 2.3165364265441895, "epoch": 3.1, "learning_rate": 4.4070131496556045e-05, "loss": 82.3712, "step": 3669, "task_loss": 1.834657073020935 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7465046195757481, "compression/movement_sparsity/importance_threshold": -0.0016452217450157753, "compression/movement_sparsity/linear_layer_sparsity": 0.7632159365641913, "compression/movement_sparsity/model_sparsity": 0.7369971212163624, "compression_loss": 79.70130157470703, "distillation_loss": 3.246070384979248, "epoch": 3.1, "learning_rate": 4.406700062617408e-05, "loss": 83.1234, "step": 3670, "task_loss": 1.8414052724838257 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7468427636579938, "compression/movement_sparsity/importance_threshold": -0.0016430271409321505, "compression/movement_sparsity/linear_layer_sparsity": 0.7636269864709347, "compression/movement_sparsity/model_sparsity": 0.7373940502942719, "compression_loss": 79.73682403564453, "distillation_loss": 5.045353889465332, "epoch": 3.1, "learning_rate": 4.4063869755792115e-05, "loss": 83.2871, "step": 3671, "task_loss": 2.4401562213897705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7471806069003284, "compression/movement_sparsity/importance_threshold": -0.0016408344893431345, "compression/movement_sparsity/linear_layer_sparsity": 0.7640790316660103, "compression/movement_sparsity/model_sparsity": 0.7378305663462429, "compression_loss": 79.77230072021484, "distillation_loss": 4.251863479614258, "epoch": 3.1, "learning_rate": 4.406073888541015e-05, "loss": 83.6949, "step": 3672, "task_loss": 2.6148390769958496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7475181494366374, "compression/movement_sparsity/importance_threshold": -0.0016386437893797902, "compression/movement_sparsity/linear_layer_sparsity": 0.7644647665648628, "compression/movement_sparsity/model_sparsity": 0.7382030500646605, "compression_loss": 79.80770111083984, "distillation_loss": 3.5415124893188477, "epoch": 3.1, "learning_rate": 4.405760801502818e-05, "loss": 82.9713, "step": 3673, "task_loss": 1.532615303993225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7478553914008055, "compression/movement_sparsity/importance_threshold": -0.001636455040173187, "compression/movement_sparsity/linear_layer_sparsity": 0.7648526358897223, "compression/movement_sparsity/model_sparsity": 0.7385775948849853, "compression_loss": 79.8431167602539, "distillation_loss": 3.8313987255096436, "epoch": 3.11, "learning_rate": 4.405447714464621e-05, "loss": 82.9057, "step": 3674, "task_loss": 1.7587419748306274 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7481923329267182, "compression/movement_sparsity/importance_threshold": -0.001634268240854388, "compression/movement_sparsity/linear_layer_sparsity": 0.76518266107738, "compression/movement_sparsity/model_sparsity": 0.7388962826921708, "compression_loss": 79.87845611572266, "distillation_loss": 3.3148670196533203, "epoch": 3.11, "learning_rate": 4.405134627426425e-05, "loss": 83.5179, "step": 3675, "task_loss": 2.120429754257202 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7485289741482603, "compression/movement_sparsity/importance_threshold": -0.0016320833905544624, "compression/movement_sparsity/linear_layer_sparsity": 0.7654730861043191, "compression/movement_sparsity/model_sparsity": 0.7391767307259827, "compression_loss": 79.91378784179688, "distillation_loss": 4.630916595458984, "epoch": 3.11, "learning_rate": 4.404821540388228e-05, "loss": 83.8915, "step": 3676, "task_loss": 2.0117106437683105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7488653151993174, "compression/movement_sparsity/importance_threshold": -0.0016299004884044723, "compression/movement_sparsity/linear_layer_sparsity": 0.7657063228232766, "compression/movement_sparsity/model_sparsity": 0.7394019550461238, "compression_loss": 79.94906616210938, "distillation_loss": 3.438753604888916, "epoch": 3.11, "learning_rate": 4.404508453350031e-05, "loss": 83.7234, "step": 3677, "task_loss": 2.7705938816070557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.749201356213774, "compression/movement_sparsity/importance_threshold": -0.001627719533535488, "compression/movement_sparsity/linear_layer_sparsity": 0.7660492738086516, "compression/movement_sparsity/model_sparsity": 0.7397331246101106, "compression_loss": 79.98436737060547, "distillation_loss": 3.276404619216919, "epoch": 3.11, "learning_rate": 4.404195366311835e-05, "loss": 83.0831, "step": 3678, "task_loss": 2.0047852993011475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7495370973255158, "compression/movement_sparsity/importance_threshold": -0.0016255405250785735, "compression/movement_sparsity/linear_layer_sparsity": 0.76641752787775, "compression/movement_sparsity/model_sparsity": 0.7400887280190535, "compression_loss": 80.01956939697266, "distillation_loss": 3.0435895919799805, "epoch": 3.11, "learning_rate": 4.403882279273638e-05, "loss": 83.6447, "step": 3679, "task_loss": 1.2687549591064453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7498725386684278, "compression/movement_sparsity/importance_threshold": -0.0016233634621647935, "compression/movement_sparsity/linear_layer_sparsity": 0.7668000670996761, "compression/movement_sparsity/model_sparsity": 0.7404581258418781, "compression_loss": 80.0547866821289, "distillation_loss": 3.9168543815612793, "epoch": 3.11, "learning_rate": 4.4035691922354415e-05, "loss": 83.7836, "step": 3680, "task_loss": 2.455794334411621 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7502076803763951, "compression/movement_sparsity/importance_threshold": -0.0016211883439252155, "compression/movement_sparsity/linear_layer_sparsity": 0.7671961283277013, "compression/movement_sparsity/model_sparsity": 0.7408405811482938, "compression_loss": 80.09001159667969, "distillation_loss": 3.7727737426757812, "epoch": 3.11, "learning_rate": 4.403256105197245e-05, "loss": 83.486, "step": 3681, "task_loss": 2.5561752319335938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7505425225833029, "compression/movement_sparsity/importance_threshold": -0.0016190151694909054, "compression/movement_sparsity/linear_layer_sparsity": 0.7676324932423357, "compression/movement_sparsity/model_sparsity": 0.741261955585695, "compression_loss": 80.12516021728516, "distillation_loss": 3.9887285232543945, "epoch": 3.11, "learning_rate": 4.4029430181590485e-05, "loss": 84.4056, "step": 3682, "task_loss": 2.1028285026550293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.750877065423036, "compression/movement_sparsity/importance_threshold": -0.0016168439379929305, "compression/movement_sparsity/linear_layer_sparsity": 0.7680301165363213, "compression/movement_sparsity/model_sparsity": 0.7416459192962997, "compression_loss": 80.16032409667969, "distillation_loss": 3.049469232559204, "epoch": 3.11, "learning_rate": 4.402629931120852e-05, "loss": 83.8679, "step": 3683, "task_loss": 1.6021056175231934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7512113090294802, "compression/movement_sparsity/importance_threshold": -0.001614674648562353, "compression/movement_sparsity/linear_layer_sparsity": 0.768340526468218, "compression/movement_sparsity/model_sparsity": 0.7419456656921032, "compression_loss": 80.19541931152344, "distillation_loss": 2.464712619781494, "epoch": 3.11, "learning_rate": 4.402316844082655e-05, "loss": 84.0245, "step": 3684, "task_loss": 1.8267407417297363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7515452535365199, "compression/movement_sparsity/importance_threshold": -0.0016125073003302442, "compression/movement_sparsity/linear_layer_sparsity": 0.7686498513008599, "compression/movement_sparsity/model_sparsity": 0.7422443642651494, "compression_loss": 80.23050689697266, "distillation_loss": 2.8290176391601562, "epoch": 3.11, "learning_rate": 4.402003757044458e-05, "loss": 83.9359, "step": 3685, "task_loss": 1.4678432941436768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.751878899078041, "compression/movement_sparsity/importance_threshold": -0.0016103418924276651, "compression/movement_sparsity/linear_layer_sparsity": 0.769009567665931, "compression/movement_sparsity/model_sparsity": 0.7425917232664633, "compression_loss": 80.26555633544922, "distillation_loss": 4.162856578826904, "epoch": 3.12, "learning_rate": 4.401690670006262e-05, "loss": 83.7872, "step": 3686, "task_loss": 1.9911335706710815 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7522122457879278, "compression/movement_sparsity/importance_threshold": -0.001608178423985686, "compression/movement_sparsity/linear_layer_sparsity": 0.7694101481534902, "compression/movement_sparsity/model_sparsity": 0.7429785425819451, "compression_loss": 80.30059814453125, "distillation_loss": 5.2794189453125, "epoch": 3.12, "learning_rate": 4.401377582968065e-05, "loss": 83.8648, "step": 3687, "task_loss": 2.595512628555298 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7525452938000659, "compression/movement_sparsity/importance_threshold": -0.0016060168941353717, "compression/movement_sparsity/linear_layer_sparsity": 0.7697044485349109, "compression/movement_sparsity/model_sparsity": 0.7432627328398902, "compression_loss": 80.33560943603516, "distillation_loss": 3.9007928371429443, "epoch": 3.12, "learning_rate": 4.401064495929868e-05, "loss": 84.1333, "step": 3688, "task_loss": 2.3016064167022705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7528780432483407, "compression/movement_sparsity/importance_threshold": -0.0016038573020077853, "compression/movement_sparsity/linear_layer_sparsity": 0.7699603292482089, "compression/movement_sparsity/model_sparsity": 0.7435098232635052, "compression_loss": 80.37051391601562, "distillation_loss": 3.5751657485961914, "epoch": 3.12, "learning_rate": 4.400751408891672e-05, "loss": 83.4597, "step": 3689, "task_loss": 2.757793426513672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7532104942666369, "compression/movement_sparsity/importance_threshold": -0.001601699646733996, "compression/movement_sparsity/linear_layer_sparsity": 0.7702165199898654, "compression/movement_sparsity/model_sparsity": 0.7437572130650509, "compression_loss": 80.40544128417969, "distillation_loss": 3.503779411315918, "epoch": 3.12, "learning_rate": 4.400438321853475e-05, "loss": 83.8099, "step": 3690, "task_loss": 2.3303747177124023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.75354264698884, "compression/movement_sparsity/importance_threshold": -0.001599543927445068, "compression/movement_sparsity/linear_layer_sparsity": 0.7703614343991441, "compression/movement_sparsity/model_sparsity": 0.743897149218562, "compression_loss": 80.4402847290039, "distillation_loss": 3.9433696269989014, "epoch": 3.12, "learning_rate": 4.400125234815279e-05, "loss": 84.2687, "step": 3691, "task_loss": 2.2165632247924805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7538745015488346, "compression/movement_sparsity/importance_threshold": -0.0015973901432720702, "compression/movement_sparsity/linear_layer_sparsity": 0.7706575591782131, "compression/movement_sparsity/model_sparsity": 0.7441831012004836, "compression_loss": 80.47509765625, "distillation_loss": 3.4335734844207764, "epoch": 3.12, "learning_rate": 4.399812147777082e-05, "loss": 84.0797, "step": 3692, "task_loss": 2.3353664875030518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7542060580805066, "compression/movement_sparsity/importance_threshold": -0.001595238293346065, "compression/movement_sparsity/linear_layer_sparsity": 0.7709988407801192, "compression/movement_sparsity/model_sparsity": 0.7445126587294592, "compression_loss": 80.50992584228516, "distillation_loss": 3.0164272785186768, "epoch": 3.12, "learning_rate": 4.3994990607388855e-05, "loss": 83.8617, "step": 3693, "task_loss": 2.840391159057617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7545373167177403, "compression/movement_sparsity/importance_threshold": -0.0015930883767981217, "compression/movement_sparsity/linear_layer_sparsity": 0.771369467758577, "compression/movement_sparsity/model_sparsity": 0.744870553531025, "compression_loss": 80.54463958740234, "distillation_loss": 3.2417004108428955, "epoch": 3.12, "learning_rate": 4.3991859737006894e-05, "loss": 84.017, "step": 3694, "task_loss": 1.2790828943252563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7548682775944217, "compression/movement_sparsity/importance_threshold": -0.0015909403927593024, "compression/movement_sparsity/linear_layer_sparsity": 0.771666904196086, "compression/movement_sparsity/model_sparsity": 0.7451577721118842, "compression_loss": 80.57938385009766, "distillation_loss": 4.243828296661377, "epoch": 3.12, "learning_rate": 4.3988728866624925e-05, "loss": 84.1222, "step": 3695, "task_loss": 3.792475461959839 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7551989408444351, "compression/movement_sparsity/importance_threshold": -0.0015887943403606782, "compression/movement_sparsity/linear_layer_sparsity": 0.7719995765489589, "compression/movement_sparsity/model_sparsity": 0.745479016146016, "compression_loss": 80.6141357421875, "distillation_loss": 3.4061670303344727, "epoch": 3.12, "learning_rate": 4.398559799624296e-05, "loss": 83.6181, "step": 3696, "task_loss": 2.622199773788452 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7555293066016664, "compression/movement_sparsity/importance_threshold": -0.0015866502187333103, "compression/movement_sparsity/linear_layer_sparsity": 0.7723224353118675, "compression/movement_sparsity/model_sparsity": 0.745790783717189, "compression_loss": 80.64875030517578, "distillation_loss": 3.389481544494629, "epoch": 3.13, "learning_rate": 4.3982467125860996e-05, "loss": 84.8651, "step": 3697, "task_loss": 2.837864398956299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.755859375, "compression/movement_sparsity/importance_threshold": -0.001584508027008269, "compression/movement_sparsity/linear_layer_sparsity": 0.7724519913932312, "compression/movement_sparsity/model_sparsity": 0.7459158891485966, "compression_loss": 80.68341064453125, "distillation_loss": 5.482818126678467, "epoch": 3.13, "learning_rate": 4.397933625547903e-05, "loss": 84.9671, "step": 3698, "task_loss": 3.283541202545166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7561891461733214, "compression/movement_sparsity/importance_threshold": -0.0015823677643166181, "compression/movement_sparsity/linear_layer_sparsity": 0.7727140845978675, "compression/movement_sparsity/model_sparsity": 0.7461689786453607, "compression_loss": 80.718017578125, "distillation_loss": 4.195858001708984, "epoch": 3.13, "learning_rate": 4.397620538509706e-05, "loss": 84.579, "step": 3699, "task_loss": 3.3106727600097656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7565186202555162, "compression/movement_sparsity/importance_threshold": -0.0015802294297894218, "compression/movement_sparsity/linear_layer_sparsity": 0.7729689159844135, "compression/movement_sparsity/model_sparsity": 0.7464150557898258, "compression_loss": 80.75260162353516, "distillation_loss": 5.171049118041992, "epoch": 3.13, "learning_rate": 4.397307451471509e-05, "loss": 84.9316, "step": 3700, "task_loss": 2.9310286045074463 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7568477973804688, "compression/movement_sparsity/importance_threshold": -0.00157809302255775, "compression/movement_sparsity/linear_layer_sparsity": 0.7732636933325396, "compression/movement_sparsity/model_sparsity": 0.7466997066292027, "compression_loss": 80.78717041015625, "distillation_loss": 4.509161949157715, "epoch": 3.13, "learning_rate": 4.396994364433313e-05, "loss": 84.8987, "step": 3701, "task_loss": 3.5827865600585938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7571766776820649, "compression/movement_sparsity/importance_threshold": -0.0015759585417526659, "compression/movement_sparsity/linear_layer_sparsity": 0.7735390819840899, "compression/movement_sparsity/model_sparsity": 0.7469656348333776, "compression_loss": 80.82174682617188, "distillation_loss": 2.8409318923950195, "epoch": 3.13, "learning_rate": 4.396681277395116e-05, "loss": 84.0792, "step": 3702, "task_loss": 2.199186086654663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.757505261294189, "compression/movement_sparsity/importance_threshold": -0.0015738259865052378, "compression/movement_sparsity/linear_layer_sparsity": 0.7738903918109777, "compression/movement_sparsity/model_sparsity": 0.7473048760869564, "compression_loss": 80.85620880126953, "distillation_loss": 2.8194642066955566, "epoch": 3.13, "learning_rate": 4.396368190356919e-05, "loss": 84.4015, "step": 3703, "task_loss": 1.4352034330368042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.757833548350727, "compression/movement_sparsity/importance_threshold": -0.001571695355946528, "compression/movement_sparsity/linear_layer_sparsity": 0.7742059410591255, "compression/movement_sparsity/model_sparsity": 0.7476095852476873, "compression_loss": 80.89073181152344, "distillation_loss": 5.176558971405029, "epoch": 3.13, "learning_rate": 4.396055103318723e-05, "loss": 85.5846, "step": 3704, "task_loss": 3.094977378845215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7581615389855635, "compression/movement_sparsity/importance_threshold": -0.0015695666492076075, "compression/movement_sparsity/linear_layer_sparsity": 0.7744694651638782, "compression/movement_sparsity/model_sparsity": 0.7478640564887467, "compression_loss": 80.9251937866211, "distillation_loss": 3.3718836307525635, "epoch": 3.13, "learning_rate": 4.3957420162805264e-05, "loss": 85.1505, "step": 3705, "task_loss": 2.215773582458496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.758489233332584, "compression/movement_sparsity/importance_threshold": -0.0015674398654195376, "compression/movement_sparsity/linear_layer_sparsity": 0.7748650851977008, "compression/movement_sparsity/model_sparsity": 0.7482460857573379, "compression_loss": 80.95954895019531, "distillation_loss": 4.87458610534668, "epoch": 3.13, "learning_rate": 4.3954289292423295e-05, "loss": 84.498, "step": 3706, "task_loss": 2.872258424758911 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7588166315256731, "compression/movement_sparsity/importance_threshold": -0.0015653150037133884, "compression/movement_sparsity/linear_layer_sparsity": 0.7751661704305063, "compression/movement_sparsity/model_sparsity": 0.74853682778615, "compression_loss": 80.99388885498047, "distillation_loss": 3.691584587097168, "epoch": 3.13, "learning_rate": 4.395115842204133e-05, "loss": 84.5891, "step": 3707, "task_loss": 2.344271183013916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7591437336987167, "compression/movement_sparsity/importance_threshold": -0.0015631920632202222, "compression/movement_sparsity/linear_layer_sparsity": 0.7754583721584232, "compression/movement_sparsity/model_sparsity": 0.7488189914857953, "compression_loss": 81.02827453613281, "distillation_loss": 2.6076674461364746, "epoch": 3.13, "learning_rate": 4.3948027551659366e-05, "loss": 84.1358, "step": 3708, "task_loss": 0.6204352378845215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7594705399855991, "compression/movement_sparsity/importance_threshold": -0.0015610710430711099, "compression/movement_sparsity/linear_layer_sparsity": 0.7758891207901011, "compression/movement_sparsity/model_sparsity": 0.7492349425768373, "compression_loss": 81.06254577636719, "distillation_loss": 3.276416778564453, "epoch": 3.14, "learning_rate": 4.39448966812774e-05, "loss": 84.299, "step": 3709, "task_loss": 1.9272433519363403 }, { "compression/movement_sparsity/importance_regularization_factor": 0.759797050520206, "compression/movement_sparsity/importance_threshold": -0.0015589519423971137, "compression/movement_sparsity/linear_layer_sparsity": 0.776115727671853, "compression/movement_sparsity/model_sparsity": 0.7494537648150768, "compression_loss": 81.09688568115234, "distillation_loss": 3.6314175128936768, "epoch": 3.14, "learning_rate": 4.394176581089543e-05, "loss": 84.6382, "step": 3710, "task_loss": 2.0691781044006348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7601232654364226, "compression/movement_sparsity/importance_threshold": -0.0015568347603292986, "compression/movement_sparsity/linear_layer_sparsity": 0.7763680430590278, "compression/movement_sparsity/model_sparsity": 0.7496974123924892, "compression_loss": 81.13113403320312, "distillation_loss": 3.0310287475585938, "epoch": 3.14, "learning_rate": 4.393863494051346e-05, "loss": 84.8128, "step": 3711, "task_loss": 2.5577585697174072 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7604491848681336, "compression/movement_sparsity/importance_threshold": -0.0015547194959987347, "compression/movement_sparsity/linear_layer_sparsity": 0.7767791168141066, "compression/movement_sparsity/model_sparsity": 0.7500943644994702, "compression_loss": 81.16535949707031, "distillation_loss": 4.442193031311035, "epoch": 3.14, "learning_rate": 4.39355040701315e-05, "loss": 84.3362, "step": 3712, "task_loss": 1.9692318439483643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7607748089492248, "compression/movement_sparsity/importance_threshold": -0.001552606148536484, "compression/movement_sparsity/linear_layer_sparsity": 0.7770833142546651, "compression/movement_sparsity/model_sparsity": 0.7503881118221247, "compression_loss": 81.1995620727539, "distillation_loss": 3.6526315212249756, "epoch": 3.14, "learning_rate": 4.393237319974953e-05, "loss": 84.5568, "step": 3713, "task_loss": 2.017167091369629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7611001378135805, "compression/movement_sparsity/importance_threshold": -0.0015504947170736161, "compression/movement_sparsity/linear_layer_sparsity": 0.7773856157525695, "compression/movement_sparsity/model_sparsity": 0.7506800283335879, "compression_loss": 81.23367309570312, "distillation_loss": 3.1547865867614746, "epoch": 3.14, "learning_rate": 4.392924232936756e-05, "loss": 84.3721, "step": 3714, "task_loss": 1.007944107055664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7614251715950866, "compression/movement_sparsity/importance_threshold": -0.0015483852007411947, "compression/movement_sparsity/linear_layer_sparsity": 0.7777822612648089, "compression/movement_sparsity/model_sparsity": 0.7510630478522575, "compression_loss": 81.26779174804688, "distillation_loss": 3.8923192024230957, "epoch": 3.14, "learning_rate": 4.39261114589856e-05, "loss": 84.9871, "step": 3715, "task_loss": 1.9376229047775269 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7617499104276276, "compression/movement_sparsity/importance_threshold": -0.0015462775986702874, "compression/movement_sparsity/linear_layer_sparsity": 0.7780789822522597, "compression/movement_sparsity/model_sparsity": 0.7513495755609689, "compression_loss": 81.30188751220703, "distillation_loss": 3.5644702911376953, "epoch": 3.14, "learning_rate": 4.3922980588603634e-05, "loss": 84.472, "step": 3716, "task_loss": 2.471728563308716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7620743544450893, "compression/movement_sparsity/importance_threshold": -0.0015441719099919572, "compression/movement_sparsity/linear_layer_sparsity": 0.7783296163317979, "compression/movement_sparsity/model_sparsity": 0.7515915995888344, "compression_loss": 81.33592987060547, "distillation_loss": 3.27058744430542, "epoch": 3.14, "learning_rate": 4.3919849718221665e-05, "loss": 85.1086, "step": 3717, "task_loss": 1.8563690185546875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7623985037813562, "compression/movement_sparsity/importance_threshold": -0.0015420681338372752, "compression/movement_sparsity/linear_layer_sparsity": 0.778597838558599, "compression/movement_sparsity/model_sparsity": 0.7518506075569967, "compression_loss": 81.36996459960938, "distillation_loss": 3.71075439453125, "epoch": 3.14, "learning_rate": 4.39167188478397e-05, "loss": 85.3029, "step": 3718, "task_loss": 2.2643015384674072 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7627223585703138, "compression/movement_sparsity/importance_threshold": -0.0015399662693373036, "compression/movement_sparsity/linear_layer_sparsity": 0.7790141231749346, "compression/movement_sparsity/model_sparsity": 0.75225259151612, "compression_loss": 81.40399169921875, "distillation_loss": 3.370454788208008, "epoch": 3.14, "learning_rate": 4.3913587977457736e-05, "loss": 85.2853, "step": 3719, "task_loss": 2.3660244941711426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7630459189458474, "compression/movement_sparsity/importance_threshold": -0.001537866315623108, "compression/movement_sparsity/linear_layer_sparsity": 0.7793357299002415, "compression/movement_sparsity/model_sparsity": 0.7525631500610345, "compression_loss": 81.43799591064453, "distillation_loss": 3.639441967010498, "epoch": 3.14, "learning_rate": 4.391045710707577e-05, "loss": 84.6225, "step": 3720, "task_loss": 2.156487226486206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7633691850418416, "compression/movement_sparsity/importance_threshold": -0.001535768271825758, "compression/movement_sparsity/linear_layer_sparsity": 0.7796421571601478, "compression/movement_sparsity/model_sparsity": 0.7528590506018826, "compression_loss": 81.47196197509766, "distillation_loss": 4.3593573570251465, "epoch": 3.15, "learning_rate": 4.39073262366938e-05, "loss": 84.938, "step": 3721, "task_loss": 2.688934564590454 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7636921569921822, "compression/movement_sparsity/importance_threshold": -0.0015336721370763145, "compression/movement_sparsity/linear_layer_sparsity": 0.7799590657634062, "compression/movement_sparsity/model_sparsity": 0.7531650724196941, "compression_loss": 81.50591278076172, "distillation_loss": 2.511467456817627, "epoch": 3.15, "learning_rate": 4.390419536631184e-05, "loss": 84.9606, "step": 3722, "task_loss": 2.484342336654663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7640148349307536, "compression/movement_sparsity/importance_threshold": -0.0015315779105058496, "compression/movement_sparsity/linear_layer_sparsity": 0.7801509852415054, "compression/movement_sparsity/model_sparsity": 0.7533503988733072, "compression_loss": 81.53987121582031, "distillation_loss": 3.4028987884521484, "epoch": 3.15, "learning_rate": 4.390106449592987e-05, "loss": 84.608, "step": 3723, "task_loss": 1.4678404331207275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7643372189914417, "compression/movement_sparsity/importance_threshold": -0.0015294855912454247, "compression/movement_sparsity/linear_layer_sparsity": 0.7804389538899115, "compression/movement_sparsity/model_sparsity": 0.7536284749127454, "compression_loss": 81.57369995117188, "distillation_loss": 3.353816032409668, "epoch": 3.15, "learning_rate": 4.38979336255479e-05, "loss": 85.3898, "step": 3724, "task_loss": 1.4579589366912842 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7646593093081311, "compression/movement_sparsity/importance_threshold": -0.001527395178426108, "compression/movement_sparsity/linear_layer_sparsity": 0.7806845798190426, "compression/movement_sparsity/model_sparsity": 0.7538656628355772, "compression_loss": 81.60748291015625, "distillation_loss": 2.669013500213623, "epoch": 3.15, "learning_rate": 4.389480275516594e-05, "loss": 85.0796, "step": 3725, "task_loss": 1.8970823287963867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7649811060147073, "compression/movement_sparsity/importance_threshold": -0.0015253066711789637, "compression/movement_sparsity/linear_layer_sparsity": 0.781102641136356, "compression/movement_sparsity/model_sparsity": 0.7542693624605338, "compression_loss": 81.64126586914062, "distillation_loss": 6.160707473754883, "epoch": 3.15, "learning_rate": 4.389167188478397e-05, "loss": 85.7938, "step": 3726, "task_loss": 3.805743455886841 }, { "compression/movement_sparsity/importance_regularization_factor": 0.765302609245055, "compression/movement_sparsity/importance_threshold": -0.0015232200686350608, "compression/movement_sparsity/linear_layer_sparsity": 0.7813134484959904, "compression/movement_sparsity/model_sparsity": 0.7544729279388455, "compression_loss": 81.6749496459961, "distillation_loss": 2.1965315341949463, "epoch": 3.15, "learning_rate": 4.388854101440201e-05, "loss": 84.6554, "step": 3727, "task_loss": 0.6288010478019714 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7656238191330595, "compression/movement_sparsity/importance_threshold": -0.0015211353699254642, "compression/movement_sparsity/linear_layer_sparsity": 0.781695808855402, "compression/movement_sparsity/model_sparsity": 0.7548421530436332, "compression_loss": 81.70866394042969, "distillation_loss": 3.436584949493408, "epoch": 3.15, "learning_rate": 4.388541014402004e-05, "loss": 85.8365, "step": 3728, "task_loss": 2.0476255416870117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7659447358126066, "compression/movement_sparsity/importance_threshold": -0.0015190525741812363, "compression/movement_sparsity/linear_layer_sparsity": 0.7819804029643671, "compression/movement_sparsity/model_sparsity": 0.7551169704694415, "compression_loss": 81.74234008789062, "distillation_loss": 4.656819820404053, "epoch": 3.15, "learning_rate": 4.3882279273638074e-05, "loss": 86.2033, "step": 3729, "task_loss": 2.412738084793091 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7662653594175803, "compression/movement_sparsity/importance_threshold": -0.0015169716805334497, "compression/movement_sparsity/linear_layer_sparsity": 0.7820891275248709, "compression/movement_sparsity/model_sparsity": 0.7552219600068162, "compression_loss": 81.77598571777344, "distillation_loss": 4.502987861633301, "epoch": 3.15, "learning_rate": 4.387914840325611e-05, "loss": 85.5906, "step": 3730, "task_loss": 3.103522539138794 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7665856900818668, "compression/movement_sparsity/importance_threshold": -0.0015148926881131631, "compression/movement_sparsity/linear_layer_sparsity": 0.7823519600279006, "compression/movement_sparsity/model_sparsity": 0.7554757634047994, "compression_loss": 81.8095932006836, "distillation_loss": 4.26683235168457, "epoch": 3.15, "learning_rate": 4.3876017532874144e-05, "loss": 85.886, "step": 3731, "task_loss": 2.1073224544525146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7669057279393504, "compression/movement_sparsity/importance_threshold": -0.0015128155960514502, "compression/movement_sparsity/linear_layer_sparsity": 0.7826750930466649, "compression/movement_sparsity/model_sparsity": 0.7557877958102956, "compression_loss": 81.84318542480469, "distillation_loss": 5.640399932861328, "epoch": 3.15, "learning_rate": 4.3872886662492176e-05, "loss": 85.942, "step": 3732, "task_loss": 3.2473011016845703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.767225473123917, "compression/movement_sparsity/importance_threshold": -0.0015107404034793697, "compression/movement_sparsity/linear_layer_sparsity": 0.7829945414976297, "compression/movement_sparsity/model_sparsity": 0.7560962702242314, "compression_loss": 81.8768081665039, "distillation_loss": 2.6043331623077393, "epoch": 3.16, "learning_rate": 4.386975579211021e-05, "loss": 84.6139, "step": 3733, "task_loss": 1.040607213973999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7675449257694509, "compression/movement_sparsity/importance_threshold": -0.0015086671095279952, "compression/movement_sparsity/linear_layer_sparsity": 0.7832705859783999, "compression/movement_sparsity/model_sparsity": 0.756362831727875, "compression_loss": 81.91031646728516, "distillation_loss": 5.19129753112793, "epoch": 3.16, "learning_rate": 4.3866624921728246e-05, "loss": 86.4707, "step": 3734, "task_loss": 3.880425453186035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.767864086009838, "compression/movement_sparsity/importance_threshold": -0.0015065957133283854, "compression/movement_sparsity/linear_layer_sparsity": 0.7835211127404293, "compression/movement_sparsity/model_sparsity": 0.7566047521249183, "compression_loss": 81.94384765625, "distillation_loss": 4.42138671875, "epoch": 3.16, "learning_rate": 4.386349405134628e-05, "loss": 85.5201, "step": 3735, "task_loss": 2.370262384414673 }, { "compression/movement_sparsity/importance_regularization_factor": 0.768182953978963, "compression/movement_sparsity/importance_threshold": -0.0015045262140116113, "compression/movement_sparsity/linear_layer_sparsity": 0.7837655581769646, "compression/movement_sparsity/model_sparsity": 0.7568408001087065, "compression_loss": 81.97736358642578, "distillation_loss": 2.4034910202026367, "epoch": 3.16, "learning_rate": 4.386036318096431e-05, "loss": 85.39, "step": 3736, "task_loss": 1.661028504371643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7685015298107113, "compression/movement_sparsity/importance_threshold": -0.001502458610708736, "compression/movement_sparsity/linear_layer_sparsity": 0.7840809047142626, "compression/movement_sparsity/model_sparsity": 0.7571453135223288, "compression_loss": 82.0107650756836, "distillation_loss": 3.6904735565185547, "epoch": 3.16, "learning_rate": 4.385723231058234e-05, "loss": 85.2566, "step": 3737, "task_loss": 2.091057777404785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7688198136389677, "compression/movement_sparsity/importance_threshold": -0.001500392902550828, "compression/movement_sparsity/linear_layer_sparsity": 0.7842966606034659, "compression/movement_sparsity/model_sparsity": 0.7573536575329953, "compression_loss": 82.04420471191406, "distillation_loss": 4.128425121307373, "epoch": 3.16, "learning_rate": 4.385410144020038e-05, "loss": 85.6986, "step": 3738, "task_loss": 2.848574638366699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7691378055976175, "compression/movement_sparsity/importance_threshold": -0.0014983290886689529, "compression/movement_sparsity/linear_layer_sparsity": 0.7845245195228197, "compression/movement_sparsity/model_sparsity": 0.7575736887974932, "compression_loss": 82.07762145996094, "distillation_loss": 2.8554792404174805, "epoch": 3.16, "learning_rate": 4.385097056981841e-05, "loss": 85.5744, "step": 3739, "task_loss": 1.942143201828003 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7694555058205462, "compression/movement_sparsity/importance_threshold": -0.0014962671681941738, "compression/movement_sparsity/linear_layer_sparsity": 0.78488439090207, "compression/movement_sparsity/model_sparsity": 0.7579211974877724, "compression_loss": 82.11095428466797, "distillation_loss": 4.167574882507324, "epoch": 3.16, "learning_rate": 4.3847839699436444e-05, "loss": 86.0499, "step": 3740, "task_loss": 2.7877447605133057 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7697729144416383, "compression/movement_sparsity/importance_threshold": -0.0014942071402575608, "compression/movement_sparsity/linear_layer_sparsity": 0.785147771916811, "compression/movement_sparsity/model_sparsity": 0.7581755305544022, "compression_loss": 82.14427185058594, "distillation_loss": 3.002695083618164, "epoch": 3.16, "learning_rate": 4.384470882905448e-05, "loss": 85.2344, "step": 3741, "task_loss": 1.3274555206298828 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7700900315947796, "compression/movement_sparsity/importance_threshold": -0.0014921490039901762, "compression/movement_sparsity/linear_layer_sparsity": 0.7853598790107177, "compression/movement_sparsity/model_sparsity": 0.7583803511171155, "compression_loss": 82.177490234375, "distillation_loss": 5.300873756408691, "epoch": 3.16, "learning_rate": 4.3841577958672514e-05, "loss": 86.0485, "step": 3742, "task_loss": 2.6296281814575195 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7704068574138547, "compression/movement_sparsity/importance_threshold": -0.00149009275852309, "compression/movement_sparsity/linear_layer_sparsity": 0.7857001947550452, "compression/movement_sparsity/model_sparsity": 0.7587089759686917, "compression_loss": 82.21073913574219, "distillation_loss": 5.168208122253418, "epoch": 3.16, "learning_rate": 4.3838447088290546e-05, "loss": 86.468, "step": 3743, "task_loss": 2.497447967529297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7707233920327491, "compression/movement_sparsity/importance_threshold": -0.0014880384029873645, "compression/movement_sparsity/linear_layer_sparsity": 0.7859788625526953, "compression/movement_sparsity/model_sparsity": 0.7589780706702102, "compression_loss": 82.24398040771484, "distillation_loss": 2.845010280609131, "epoch": 3.16, "learning_rate": 4.383531621790858e-05, "loss": 86.0477, "step": 3744, "task_loss": 1.7355300188064575 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7710396355853476, "compression/movement_sparsity/importance_threshold": -0.001485985936514069, "compression/movement_sparsity/linear_layer_sparsity": 0.7862053263444356, "compression/movement_sparsity/model_sparsity": 0.75919675473402, "compression_loss": 82.27716064453125, "distillation_loss": 4.6645731925964355, "epoch": 3.17, "learning_rate": 4.3832185347526616e-05, "loss": 85.8447, "step": 3745, "task_loss": 2.335645914077759 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7713555882055358, "compression/movement_sparsity/importance_threshold": -0.0014839353582342656, "compression/movement_sparsity/linear_layer_sparsity": 0.7863794311160869, "compression/movement_sparsity/model_sparsity": 0.7593648784711561, "compression_loss": 82.31031036376953, "distillation_loss": 4.000650405883789, "epoch": 3.17, "learning_rate": 4.382905447714465e-05, "loss": 85.9412, "step": 3746, "task_loss": 2.5272538661956787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7716712500271983, "compression/movement_sparsity/importance_threshold": -0.0014818866672790254, "compression/movement_sparsity/linear_layer_sparsity": 0.7866471406036796, "compression/movement_sparsity/model_sparsity": 0.7596233913142794, "compression_loss": 82.34343719482422, "distillation_loss": 2.2042269706726074, "epoch": 3.17, "learning_rate": 4.382592360676268e-05, "loss": 85.6064, "step": 3747, "task_loss": 1.2427804470062256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7719866211842208, "compression/movement_sparsity/importance_threshold": -0.0014798398627794097, "compression/movement_sparsity/linear_layer_sparsity": 0.7869906401007659, "compression/movement_sparsity/model_sparsity": 0.7599550905469127, "compression_loss": 82.37654113769531, "distillation_loss": 4.138267517089844, "epoch": 3.17, "learning_rate": 4.382279273638071e-05, "loss": 85.9461, "step": 3748, "task_loss": 2.6045150756835938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7723017018104879, "compression/movement_sparsity/importance_threshold": -0.0014777949438664886, "compression/movement_sparsity/linear_layer_sparsity": 0.7872410356969514, "compression/movement_sparsity/model_sparsity": 0.7601968842840623, "compression_loss": 82.40953826904297, "distillation_loss": 3.2660770416259766, "epoch": 3.17, "learning_rate": 4.381966186599875e-05, "loss": 85.708, "step": 3749, "task_loss": 2.108560562133789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7726164920398849, "compression/movement_sparsity/importance_threshold": -0.0014757519096713252, "compression/movement_sparsity/linear_layer_sparsity": 0.7874169171695804, "compression/movement_sparsity/model_sparsity": 0.7603667236870317, "compression_loss": 82.44257354736328, "distillation_loss": 3.5812301635742188, "epoch": 3.17, "learning_rate": 4.381653099561678e-05, "loss": 85.9378, "step": 3750, "task_loss": 1.2557867765426636 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7729309920062974, "compression/movement_sparsity/importance_threshold": -0.0014737107593249853, "compression/movement_sparsity/linear_layer_sparsity": 0.7877060663105824, "compression/movement_sparsity/model_sparsity": 0.7606459396655135, "compression_loss": 82.47552490234375, "distillation_loss": 2.866525173187256, "epoch": 3.17, "learning_rate": 4.3813400125234814e-05, "loss": 85.7209, "step": 3751, "task_loss": 1.6363738775253296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7732452018436099, "compression/movement_sparsity/importance_threshold": -0.001471671491958538, "compression/movement_sparsity/linear_layer_sparsity": 0.7878946112492407, "compression/movement_sparsity/model_sparsity": 0.7608280075054968, "compression_loss": 82.50849914550781, "distillation_loss": 3.6934590339660645, "epoch": 3.17, "learning_rate": 4.381026925485285e-05, "loss": 86.2042, "step": 3752, "task_loss": 1.7451331615447998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7735591216857081, "compression/movement_sparsity/importance_threshold": -0.0014696341067030457, "compression/movement_sparsity/linear_layer_sparsity": 0.7882758149643916, "compression/movement_sparsity/model_sparsity": 0.7611961157003125, "compression_loss": 82.54141235351562, "distillation_loss": 3.4455032348632812, "epoch": 3.17, "learning_rate": 4.3807138384470884e-05, "loss": 86.4418, "step": 3753, "task_loss": 1.7903826236724854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7738727516664765, "compression/movement_sparsity/importance_threshold": -0.0014675986026895776, "compression/movement_sparsity/linear_layer_sparsity": 0.788615808756193, "compression/movement_sparsity/model_sparsity": 0.7615244296594222, "compression_loss": 82.57429504394531, "distillation_loss": 3.2792015075683594, "epoch": 3.17, "learning_rate": 4.3804007514088916e-05, "loss": 86.3827, "step": 3754, "task_loss": 1.7794232368469238 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7741860919198009, "compression/movement_sparsity/importance_threshold": -0.0014655649790491967, "compression/movement_sparsity/linear_layer_sparsity": 0.7888328047588303, "compression/movement_sparsity/model_sparsity": 0.7617339711818113, "compression_loss": 82.60720825195312, "distillation_loss": 3.953794479370117, "epoch": 3.17, "learning_rate": 4.380087664370695e-05, "loss": 86.7646, "step": 3755, "task_loss": 2.22540020942688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.774499142579566, "compression/movement_sparsity/importance_threshold": -0.0014635332349129723, "compression/movement_sparsity/linear_layer_sparsity": 0.789066625762002, "compression/movement_sparsity/model_sparsity": 0.7619597597142065, "compression_loss": 82.64002990722656, "distillation_loss": 3.447566032409668, "epoch": 3.17, "learning_rate": 4.3797745773324986e-05, "loss": 85.8306, "step": 3756, "task_loss": 1.9097049236297607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7748119037796571, "compression/movement_sparsity/importance_threshold": -0.0014615033694119674, "compression/movement_sparsity/linear_layer_sparsity": 0.7894069653546648, "compression/movement_sparsity/model_sparsity": 0.7622884075948543, "compression_loss": 82.67284393310547, "distillation_loss": 2.6962924003601074, "epoch": 3.18, "learning_rate": 4.379461490294302e-05, "loss": 86.4598, "step": 3757, "task_loss": 1.4286205768585205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7751243756539593, "compression/movement_sparsity/importance_threshold": -0.0014594753816772506, "compression/movement_sparsity/linear_layer_sparsity": 0.7896800168693584, "compression/movement_sparsity/model_sparsity": 0.7625520789500134, "compression_loss": 82.70560455322266, "distillation_loss": 3.5418663024902344, "epoch": 3.18, "learning_rate": 4.3791484032561056e-05, "loss": 85.9286, "step": 3758, "task_loss": 1.515271544456482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7754365583363576, "compression/movement_sparsity/importance_threshold": -0.0014574492708398883, "compression/movement_sparsity/linear_layer_sparsity": 0.7898275784438522, "compression/movement_sparsity/model_sparsity": 0.7626945713304709, "compression_loss": 82.73831176757812, "distillation_loss": 4.661841869354248, "epoch": 3.18, "learning_rate": 4.378835316217909e-05, "loss": 86.6856, "step": 3759, "task_loss": 2.7708866596221924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7757484519607377, "compression/movement_sparsity/importance_threshold": -0.001455425036030941, "compression/movement_sparsity/linear_layer_sparsity": 0.7900619241103999, "compression/movement_sparsity/model_sparsity": 0.762920866502441, "compression_loss": 82.77098846435547, "distillation_loss": 3.522245168685913, "epoch": 3.18, "learning_rate": 4.378522229179712e-05, "loss": 86.2876, "step": 3760, "task_loss": 2.7381577491760254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.776060056660984, "compression/movement_sparsity/importance_threshold": -0.0014534026763814824, "compression/movement_sparsity/linear_layer_sparsity": 0.7903216921023472, "compression/movement_sparsity/model_sparsity": 0.763171710664725, "compression_loss": 82.80365753173828, "distillation_loss": 4.747779846191406, "epoch": 3.18, "learning_rate": 4.378209142141516e-05, "loss": 86.5954, "step": 3761, "task_loss": 2.0556485652923584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7763713725709822, "compression/movement_sparsity/importance_threshold": -0.001451382191022572, "compression/movement_sparsity/linear_layer_sparsity": 0.7905319151777674, "compression/movement_sparsity/model_sparsity": 0.7633747119307828, "compression_loss": 82.8363037109375, "distillation_loss": 3.680877447128296, "epoch": 3.18, "learning_rate": 4.377896055103319e-05, "loss": 86.9298, "step": 3762, "task_loss": 2.350839376449585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7766823998246171, "compression/movement_sparsity/importance_threshold": -0.001449363579085281, "compression/movement_sparsity/linear_layer_sparsity": 0.7908353852441001, "compression/movement_sparsity/model_sparsity": 0.7636677568667539, "compression_loss": 82.86892700195312, "distillation_loss": 3.3014087677001953, "epoch": 3.18, "learning_rate": 4.377582968065123e-05, "loss": 86.2013, "step": 3763, "task_loss": 2.592341661453247 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7769931385557741, "compression/movement_sparsity/importance_threshold": -0.0014473468397006714, "compression/movement_sparsity/linear_layer_sparsity": 0.7911814603613958, "compression/movement_sparsity/model_sparsity": 0.7640019432391189, "compression_loss": 82.90155792236328, "distillation_loss": 3.4840621948242188, "epoch": 3.18, "learning_rate": 4.377269881026926e-05, "loss": 86.9753, "step": 3764, "task_loss": 2.1219053268432617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7773035888983381, "compression/movement_sparsity/importance_threshold": -0.0014453319719998125, "compression/movement_sparsity/linear_layer_sparsity": 0.7914384261739487, "compression/movement_sparsity/model_sparsity": 0.7642500814854912, "compression_loss": 82.93413543701172, "distillation_loss": 5.417972564697266, "epoch": 3.18, "learning_rate": 4.376956793988729e-05, "loss": 87.0307, "step": 3765, "task_loss": 2.87770414352417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7776137509861945, "compression/movement_sparsity/importance_threshold": -0.0014433189751137674, "compression/movement_sparsity/linear_layer_sparsity": 0.7917655060922003, "compression/movement_sparsity/model_sparsity": 0.7645659252023355, "compression_loss": 82.96672058105469, "distillation_loss": 5.919469356536865, "epoch": 3.18, "learning_rate": 4.3766437069505324e-05, "loss": 87.3575, "step": 3766, "task_loss": 3.9351327419281006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.777923624953228, "compression/movement_sparsity/importance_threshold": -0.0014413078481736053, "compression/movement_sparsity/linear_layer_sparsity": 0.7919911828888767, "compression/movement_sparsity/model_sparsity": 0.7647838493067829, "compression_loss": 82.99921417236328, "distillation_loss": 3.9899425506591797, "epoch": 3.18, "learning_rate": 4.376330619912336e-05, "loss": 86.3216, "step": 3767, "task_loss": 2.065713405609131 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7782332109333243, "compression/movement_sparsity/importance_threshold": -0.0014392985903103886, "compression/movement_sparsity/linear_layer_sparsity": 0.7921427032870256, "compression/movement_sparsity/model_sparsity": 0.7649301645131241, "compression_loss": 83.0317611694336, "distillation_loss": 4.887537002563477, "epoch": 3.19, "learning_rate": 4.3760175328741395e-05, "loss": 87.3933, "step": 3768, "task_loss": 3.4606988430023193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.778542509060368, "compression/movement_sparsity/importance_threshold": -0.0014372912006551872, "compression/movement_sparsity/linear_layer_sparsity": 0.792377645161955, "compression/movement_sparsity/model_sparsity": 0.765157035411884, "compression_loss": 83.06417846679688, "distillation_loss": 4.350998878479004, "epoch": 3.19, "learning_rate": 4.3757044458359426e-05, "loss": 86.5883, "step": 3769, "task_loss": 1.5512999296188354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7788515194682446, "compression/movement_sparsity/importance_threshold": -0.0014352856783390652, "compression/movement_sparsity/linear_layer_sparsity": 0.7924833767563821, "compression/movement_sparsity/model_sparsity": 0.7652591348007742, "compression_loss": 83.09668731689453, "distillation_loss": 3.2294206619262695, "epoch": 3.19, "learning_rate": 4.375391358797746e-05, "loss": 86.6319, "step": 3770, "task_loss": 1.3307068347930908 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7791602422908394, "compression/movement_sparsity/importance_threshold": -0.0014332820224930865, "compression/movement_sparsity/linear_layer_sparsity": 0.7928134377165429, "compression/movement_sparsity/model_sparsity": 0.7655778571515671, "compression_loss": 83.12906646728516, "distillation_loss": 3.541947364807129, "epoch": 3.19, "learning_rate": 4.37507827175955e-05, "loss": 86.7095, "step": 3771, "task_loss": 2.8751637935638428 }, { "compression/movement_sparsity/importance_regularization_factor": 0.779468677662037, "compression/movement_sparsity/importance_threshold": -0.0014312802322483213, "compression/movement_sparsity/linear_layer_sparsity": 0.7930600175790848, "compression/movement_sparsity/model_sparsity": 0.7658159662372623, "compression_loss": 83.16149139404297, "distillation_loss": 3.5815091133117676, "epoch": 3.19, "learning_rate": 4.374765184721353e-05, "loss": 87.1205, "step": 3772, "task_loss": 1.729410171508789 }, { "compression/movement_sparsity/importance_regularization_factor": 0.779776825715723, "compression/movement_sparsity/importance_threshold": -0.0014292803067358327, "compression/movement_sparsity/linear_layer_sparsity": 0.7933271308582959, "compression/movement_sparsity/model_sparsity": 0.7660739033535959, "compression_loss": 83.19378662109375, "distillation_loss": 4.562980651855469, "epoch": 3.19, "learning_rate": 4.374452097683156e-05, "loss": 88.028, "step": 3773, "task_loss": 2.961336851119995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7800846865857822, "compression/movement_sparsity/importance_threshold": -0.0014272822450866881, "compression/movement_sparsity/linear_layer_sparsity": 0.7935392856488731, "compression/movement_sparsity/model_sparsity": 0.7662787699744523, "compression_loss": 83.22610473632812, "distillation_loss": 3.4020814895629883, "epoch": 3.19, "learning_rate": 4.374139010644959e-05, "loss": 86.8487, "step": 3774, "task_loss": 1.8931678533554077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7803922604061002, "compression/movement_sparsity/importance_threshold": -0.0014252860464319525, "compression/movement_sparsity/linear_layer_sparsity": 0.7938007468726247, "compression/movement_sparsity/model_sparsity": 0.7665312492008193, "compression_loss": 83.25837707519531, "distillation_loss": 3.5560786724090576, "epoch": 3.19, "learning_rate": 4.373825923606763e-05, "loss": 87.2846, "step": 3775, "task_loss": 2.006788969039917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7806995473105616, "compression/movement_sparsity/importance_threshold": -0.001423291709902694, "compression/movement_sparsity/linear_layer_sparsity": 0.7940942841073168, "compression/movement_sparsity/model_sparsity": 0.7668147025284735, "compression_loss": 83.29058074951172, "distillation_loss": 5.8420305252075195, "epoch": 3.19, "learning_rate": 4.373512836568566e-05, "loss": 87.2167, "step": 3776, "task_loss": 2.6306285858154297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7810065474330519, "compression/movement_sparsity/importance_threshold": -0.001421299234629976, "compression/movement_sparsity/linear_layer_sparsity": 0.7943793313346521, "compression/movement_sparsity/model_sparsity": 0.767089957506642, "compression_loss": 83.32279205322266, "distillation_loss": 4.215404510498047, "epoch": 3.19, "learning_rate": 4.3731997495303694e-05, "loss": 87.2159, "step": 3777, "task_loss": 3.01944637298584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.781313260907456, "compression/movement_sparsity/importance_threshold": -0.0014193086197448659, "compression/movement_sparsity/linear_layer_sparsity": 0.7946254461546562, "compression/movement_sparsity/model_sparsity": 0.7673276175254413, "compression_loss": 83.35488891601562, "distillation_loss": 4.833314895629883, "epoch": 3.19, "learning_rate": 4.372886662492173e-05, "loss": 87.3687, "step": 3778, "task_loss": 2.04854154586792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7816196878676595, "compression/movement_sparsity/importance_threshold": -0.0014173198643784293, "compression/movement_sparsity/linear_layer_sparsity": 0.79485638150926, "compression/movement_sparsity/model_sparsity": 0.7675506195401741, "compression_loss": 83.38707733154297, "distillation_loss": 2.6800687313079834, "epoch": 3.19, "learning_rate": 4.3725735754539764e-05, "loss": 86.7098, "step": 3779, "task_loss": 1.555122971534729 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7819258284475468, "compression/movement_sparsity/importance_threshold": -0.0014153329676617338, "compression/movement_sparsity/linear_layer_sparsity": 0.7950181089949052, "compression/movement_sparsity/model_sparsity": 0.7677067911891555, "compression_loss": 83.41915893554688, "distillation_loss": 3.730412483215332, "epoch": 3.2, "learning_rate": 4.3722604884157796e-05, "loss": 87.3103, "step": 3780, "task_loss": 2.980163812637329 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7822316827810035, "compression/movement_sparsity/importance_threshold": -0.0014133479287258442, "compression/movement_sparsity/linear_layer_sparsity": 0.7951853096734953, "compression/movement_sparsity/model_sparsity": 0.7678682480100666, "compression_loss": 83.4511947631836, "distillation_loss": 4.245500087738037, "epoch": 3.2, "learning_rate": 4.371947401377583e-05, "loss": 87.2575, "step": 3781, "task_loss": 2.865241289138794 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7825372510019151, "compression/movement_sparsity/importance_threshold": -0.0014113647467018237, "compression/movement_sparsity/linear_layer_sparsity": 0.7954767482546834, "compression/movement_sparsity/model_sparsity": 0.768149674779421, "compression_loss": 83.48320007324219, "distillation_loss": 4.041949272155762, "epoch": 3.2, "learning_rate": 4.371634314339387e-05, "loss": 87.5327, "step": 3782, "task_loss": 2.4600138664245605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.782842533244166, "compression/movement_sparsity/importance_threshold": -0.001409383420720744, "compression/movement_sparsity/linear_layer_sparsity": 0.7956745224990919, "compression/movement_sparsity/model_sparsity": 0.7683406548701092, "compression_loss": 83.51521301269531, "distillation_loss": 4.427818298339844, "epoch": 3.2, "learning_rate": 4.37132122730119e-05, "loss": 87.7097, "step": 3783, "task_loss": 2.762794017791748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7831475296416418, "compression/movement_sparsity/importance_threshold": -0.0014074039499136666, "compression/movement_sparsity/linear_layer_sparsity": 0.7960862163108876, "compression/movement_sparsity/model_sparsity": 0.7687382057329515, "compression_loss": 83.54713439941406, "distillation_loss": 4.814364433288574, "epoch": 3.2, "learning_rate": 4.371008140262993e-05, "loss": 87.3983, "step": 3784, "task_loss": 1.9057234525680542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7834522403282274, "compression/movement_sparsity/importance_threshold": -0.0014054263334116607, "compression/movement_sparsity/linear_layer_sparsity": 0.7963014713850503, "compression/movement_sparsity/model_sparsity": 0.7689460661331146, "compression_loss": 83.57905578613281, "distillation_loss": 4.24348258972168, "epoch": 3.2, "learning_rate": 4.370695053224796e-05, "loss": 87.2738, "step": 3785, "task_loss": 2.275540828704834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7837566654378082, "compression/movement_sparsity/importance_threshold": -0.0014034505703457894, "compression/movement_sparsity/linear_layer_sparsity": 0.7966027235562027, "compression/movement_sparsity/model_sparsity": 0.7692369693654278, "compression_loss": 83.6109619140625, "distillation_loss": 4.154793739318848, "epoch": 3.2, "learning_rate": 4.3703819661866e-05, "loss": 87.2455, "step": 3786, "task_loss": 2.0068085193634033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.784060805104269, "compression/movement_sparsity/importance_threshold": -0.0014014766598471218, "compression/movement_sparsity/linear_layer_sparsity": 0.7967533496417789, "compression/movement_sparsity/model_sparsity": 0.7693824209815845, "compression_loss": 83.6427993774414, "distillation_loss": 3.389193058013916, "epoch": 3.2, "learning_rate": 4.370068879148403e-05, "loss": 87.2311, "step": 3787, "task_loss": 1.802321195602417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7843646594614955, "compression/movement_sparsity/importance_threshold": -0.0013995046010467204, "compression/movement_sparsity/linear_layer_sparsity": 0.7970430830669951, "compression/movement_sparsity/model_sparsity": 0.7696622011723202, "compression_loss": 83.67467498779297, "distillation_loss": 4.725925922393799, "epoch": 3.2, "learning_rate": 4.3697557921102064e-05, "loss": 87.6207, "step": 3788, "task_loss": 2.5372354984283447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7846682286433722, "compression/movement_sparsity/importance_threshold": -0.0013975343930756551, "compression/movement_sparsity/linear_layer_sparsity": 0.7973803700727431, "compression/movement_sparsity/model_sparsity": 0.7699879013318046, "compression_loss": 83.70652770996094, "distillation_loss": 4.379236221313477, "epoch": 3.2, "learning_rate": 4.36944270507201e-05, "loss": 87.2583, "step": 3789, "task_loss": 2.486888885498047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7849715127837844, "compression/movement_sparsity/importance_threshold": -0.00139556603506499, "compression/movement_sparsity/linear_layer_sparsity": 0.7975888641438562, "compression/movement_sparsity/model_sparsity": 0.7701892329901722, "compression_loss": 83.73831939697266, "distillation_loss": 3.489227771759033, "epoch": 3.2, "learning_rate": 4.3691296180338134e-05, "loss": 87.518, "step": 3790, "task_loss": 1.2321891784667969 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7852745120166177, "compression/movement_sparsity/importance_threshold": -0.0013935995261457898, "compression/movement_sparsity/linear_layer_sparsity": 0.7978083165250266, "compression/movement_sparsity/model_sparsity": 0.7704011465069349, "compression_loss": 83.77009582519531, "distillation_loss": 3.249908924102783, "epoch": 3.2, "learning_rate": 4.3688165309956166e-05, "loss": 87.4641, "step": 3791, "task_loss": 1.9028083086013794 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7855772264757566, "compression/movement_sparsity/importance_threshold": -0.001391634865449123, "compression/movement_sparsity/linear_layer_sparsity": 0.798085863450919, "compression/movement_sparsity/model_sparsity": 0.7706691588420886, "compression_loss": 83.8018569946289, "distillation_loss": 3.539057493209839, "epoch": 3.21, "learning_rate": 4.3685034439574205e-05, "loss": 87.12, "step": 3792, "task_loss": 1.0352365970611572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7858796562950868, "compression/movement_sparsity/importance_threshold": -0.0013896720521060536, "compression/movement_sparsity/linear_layer_sparsity": 0.7983460726370688, "compression/movement_sparsity/model_sparsity": 0.770920429042197, "compression_loss": 83.83351135253906, "distillation_loss": 2.5326294898986816, "epoch": 3.21, "learning_rate": 4.3681903569192237e-05, "loss": 87.1529, "step": 3793, "task_loss": 0.9876841306686401 }, { "compression/movement_sparsity/importance_regularization_factor": 0.786181801608493, "compression/movement_sparsity/importance_threshold": -0.001387711085247649, "compression/movement_sparsity/linear_layer_sparsity": 0.7986411838618888, "compression/movement_sparsity/model_sparsity": 0.7712054022885761, "compression_loss": 83.86515045166016, "distillation_loss": 3.568723440170288, "epoch": 3.21, "learning_rate": 4.3678772698810275e-05, "loss": 87.1439, "step": 3794, "task_loss": 1.5577876567840576 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7864836625498608, "compression/movement_sparsity/importance_threshold": -0.001385751964004974, "compression/movement_sparsity/linear_layer_sparsity": 0.798925348700819, "compression/movement_sparsity/model_sparsity": 0.7714798051910958, "compression_loss": 83.89675903320312, "distillation_loss": 3.827022075653076, "epoch": 3.21, "learning_rate": 4.367564182842831e-05, "loss": 87.7011, "step": 3795, "task_loss": 2.3786561489105225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7867852392530748, "compression/movement_sparsity/importance_threshold": -0.0013837946875090972, "compression/movement_sparsity/linear_layer_sparsity": 0.7992209965131826, "compression/movement_sparsity/model_sparsity": 0.7717652965915857, "compression_loss": 83.92828369140625, "distillation_loss": 2.4142403602600098, "epoch": 3.21, "learning_rate": 4.367251095804634e-05, "loss": 87.0077, "step": 3796, "task_loss": 1.500040888786316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7870865318520205, "compression/movement_sparsity/importance_threshold": -0.0013818392548910815, "compression/movement_sparsity/linear_layer_sparsity": 0.7993602588670018, "compression/movement_sparsity/model_sparsity": 0.7718997748551302, "compression_loss": 83.95982360839844, "distillation_loss": 3.709933280944824, "epoch": 3.21, "learning_rate": 4.366938008766438e-05, "loss": 88.1274, "step": 3797, "task_loss": 1.7202701568603516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7873875404805828, "compression/movement_sparsity/importance_threshold": -0.0013798856652819953, "compression/movement_sparsity/linear_layer_sparsity": 0.7995838966310125, "compression/movement_sparsity/model_sparsity": 0.7721157299739567, "compression_loss": 83.99130249023438, "distillation_loss": 4.8419694900512695, "epoch": 3.21, "learning_rate": 4.366624921728241e-05, "loss": 87.9126, "step": 3798, "task_loss": 1.7631484270095825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.787688265272647, "compression/movement_sparsity/importance_threshold": -0.0013779339178129043, "compression/movement_sparsity/linear_layer_sparsity": 0.7999007933101031, "compression/movement_sparsity/model_sparsity": 0.7724217402772325, "compression_loss": 84.02281188964844, "distillation_loss": 4.384052276611328, "epoch": 3.21, "learning_rate": 4.366311834690044e-05, "loss": 88.0978, "step": 3799, "task_loss": 2.076523542404175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7879887063620985, "compression/movement_sparsity/importance_threshold": -0.0013759840116148716, "compression/movement_sparsity/linear_layer_sparsity": 0.7999716586383631, "compression/movement_sparsity/model_sparsity": 0.7724901711634594, "compression_loss": 84.05427551269531, "distillation_loss": 4.88466215133667, "epoch": 3.21, "learning_rate": 4.365998747651848e-05, "loss": 87.9036, "step": 3800, "task_loss": 3.07263445854187 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7882888638828219, "compression/movement_sparsity/importance_threshold": -0.0013740359458189674, "compression/movement_sparsity/linear_layer_sparsity": 0.8002270266124527, "compression/movement_sparsity/model_sparsity": 0.7727367664620354, "compression_loss": 84.08566284179688, "distillation_loss": 3.9636590480804443, "epoch": 3.21, "learning_rate": 4.365685660613651e-05, "loss": 87.5912, "step": 3801, "task_loss": 2.128955364227295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7885887379687029, "compression/movement_sparsity/importance_threshold": -0.0013720897195562548, "compression/movement_sparsity/linear_layer_sparsity": 0.8005423016047449, "compression/movement_sparsity/model_sparsity": 0.773041210788443, "compression_loss": 84.11703491210938, "distillation_loss": 2.6379880905151367, "epoch": 3.21, "learning_rate": 4.365372573575454e-05, "loss": 87.3157, "step": 3802, "task_loss": 2.350498676300049 }, { "compression/movement_sparsity/importance_regularization_factor": 0.788888328753626, "compression/movement_sparsity/importance_threshold": -0.001370145331957802, "compression/movement_sparsity/linear_layer_sparsity": 0.8008160089486586, "compression/movement_sparsity/model_sparsity": 0.7733055154430709, "compression_loss": 84.14835357666016, "distillation_loss": 3.645094394683838, "epoch": 3.21, "learning_rate": 4.3650594865372575e-05, "loss": 87.3927, "step": 3803, "task_loss": 2.0933637619018555 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7891876363714769, "compression/movement_sparsity/importance_threshold": -0.0013682027821546733, "compression/movement_sparsity/linear_layer_sparsity": 0.8010822994603026, "compression/movement_sparsity/model_sparsity": 0.7735626580564345, "compression_loss": 84.17965698242188, "distillation_loss": 4.360992908477783, "epoch": 3.22, "learning_rate": 4.364746399499061e-05, "loss": 87.9708, "step": 3804, "task_loss": 2.8364760875701904 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7894866609561405, "compression/movement_sparsity/importance_threshold": -0.0013662620692779359, "compression/movement_sparsity/linear_layer_sparsity": 0.8013004163346978, "compression/movement_sparsity/model_sparsity": 0.7737732819451884, "compression_loss": 84.2109603881836, "distillation_loss": 3.397578239440918, "epoch": 3.22, "learning_rate": 4.3644333124608645e-05, "loss": 87.6684, "step": 3805, "task_loss": 1.9506616592407227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7897854026415021, "compression/movement_sparsity/importance_threshold": -0.0013643231924586547, "compression/movement_sparsity/linear_layer_sparsity": 0.8016602877139481, "compression/movement_sparsity/model_sparsity": 0.7741207906354676, "compression_loss": 84.24217987060547, "distillation_loss": 3.6301426887512207, "epoch": 3.22, "learning_rate": 4.364120225422668e-05, "loss": 87.7204, "step": 3806, "task_loss": 1.5780022144317627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7900838615614465, "compression/movement_sparsity/importance_threshold": -0.0013623861508278972, "compression/movement_sparsity/linear_layer_sparsity": 0.8019700656649603, "compression/movement_sparsity/model_sparsity": 0.7744199267608739, "compression_loss": 84.2734146118164, "distillation_loss": 6.274256229400635, "epoch": 3.22, "learning_rate": 4.363807138384471e-05, "loss": 88.5136, "step": 3807, "task_loss": 3.5995147228240967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7903820378498592, "compression/movement_sparsity/importance_threshold": -0.0013604509435167283, "compression/movement_sparsity/linear_layer_sparsity": 0.8021372186468798, "compression/movement_sparsity/model_sparsity": 0.7745813375236418, "compression_loss": 84.30463409423828, "distillation_loss": 3.315661907196045, "epoch": 3.22, "learning_rate": 4.363494051346275e-05, "loss": 88.2468, "step": 3808, "task_loss": 1.7942416667938232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.790679931640625, "compression/movement_sparsity/importance_threshold": -0.0013585175696562146, "compression/movement_sparsity/linear_layer_sparsity": 0.8023424812685636, "compression/movement_sparsity/model_sparsity": 0.774779548742809, "compression_loss": 84.33583068847656, "distillation_loss": 5.113038063049316, "epoch": 3.22, "learning_rate": 4.363180964308078e-05, "loss": 88.7048, "step": 3809, "task_loss": 2.903592109680176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7909775430676292, "compression/movement_sparsity/importance_threshold": -0.0013565860283774227, "compression/movement_sparsity/linear_layer_sparsity": 0.8026187761568542, "compression/movement_sparsity/model_sparsity": 0.7750463520517044, "compression_loss": 84.366943359375, "distillation_loss": 4.083876609802246, "epoch": 3.22, "learning_rate": 4.362867877269881e-05, "loss": 88.5362, "step": 3810, "task_loss": 2.5506746768951416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7912748722647573, "compression/movement_sparsity/importance_threshold": -0.0013546563188114157, "compression/movement_sparsity/linear_layer_sparsity": 0.8028087877681316, "compression/movement_sparsity/model_sparsity": 0.7752298361795903, "compression_loss": 84.39806365966797, "distillation_loss": 4.847809791564941, "epoch": 3.22, "learning_rate": 4.362554790231684e-05, "loss": 88.6045, "step": 3811, "task_loss": 2.7054810523986816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7915719193658939, "compression/movement_sparsity/importance_threshold": -0.0013527284400892637, "compression/movement_sparsity/linear_layer_sparsity": 0.8030932984079233, "compression/movement_sparsity/model_sparsity": 0.7755045730036481, "compression_loss": 84.42919921875, "distillation_loss": 2.5901336669921875, "epoch": 3.22, "learning_rate": 4.362241703193488e-05, "loss": 87.7301, "step": 3812, "task_loss": 1.3163734674453735 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7918686845049245, "compression/movement_sparsity/importance_threshold": -0.001350802391342029, "compression/movement_sparsity/linear_layer_sparsity": 0.8033755553800318, "compression/movement_sparsity/model_sparsity": 0.7757771335804406, "compression_loss": 84.4601821899414, "distillation_loss": 5.294935703277588, "epoch": 3.22, "learning_rate": 4.361928616155291e-05, "loss": 88.5054, "step": 3813, "task_loss": 2.85512375831604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7921651678157339, "compression/movement_sparsity/importance_threshold": -0.0013488781717007816, "compression/movement_sparsity/linear_layer_sparsity": 0.8036200485132375, "compression/movement_sparsity/model_sparsity": 0.7760132276223719, "compression_loss": 84.49120330810547, "distillation_loss": 3.2664361000061035, "epoch": 3.22, "learning_rate": 4.3616155291170945e-05, "loss": 87.8127, "step": 3814, "task_loss": 1.9842255115509033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7924613694322076, "compression/movement_sparsity/importance_threshold": -0.0013469557802965847, "compression/movement_sparsity/linear_layer_sparsity": 0.8039255456880684, "compression/movement_sparsity/model_sparsity": 0.7763082300294281, "compression_loss": 84.522216796875, "distillation_loss": 4.302258491516113, "epoch": 3.22, "learning_rate": 4.361302442078898e-05, "loss": 88.8326, "step": 3815, "task_loss": 3.7465360164642334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7927572894882304, "compression/movement_sparsity/importance_threshold": -0.0013450352162605048, "compression/movement_sparsity/linear_layer_sparsity": 0.8041105133764359, "compression/movement_sparsity/model_sparsity": 0.7764868435086729, "compression_loss": 84.5531234741211, "distillation_loss": 3.676725387573242, "epoch": 3.23, "learning_rate": 4.3609893550407015e-05, "loss": 87.781, "step": 3816, "task_loss": 2.7854528427124023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7930529281176879, "compression/movement_sparsity/importance_threshold": -0.0013431164787236077, "compression/movement_sparsity/linear_layer_sparsity": 0.8043774597172999, "compression/movement_sparsity/model_sparsity": 0.7767446194215053, "compression_loss": 84.58406066894531, "distillation_loss": 3.8480844497680664, "epoch": 3.23, "learning_rate": 4.360676268002505e-05, "loss": 88.8545, "step": 3817, "task_loss": 2.150320053100586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7933482854544647, "compression/movement_sparsity/importance_threshold": -0.001341199566816961, "compression/movement_sparsity/linear_layer_sparsity": 0.8045592078804058, "compression/movement_sparsity/model_sparsity": 0.7769201239760858, "compression_loss": 84.61498260498047, "distillation_loss": 3.3975605964660645, "epoch": 3.23, "learning_rate": 4.360363180964308e-05, "loss": 88.589, "step": 3818, "task_loss": 1.8679769039154053 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7936433616324464, "compression/movement_sparsity/importance_threshold": -0.0013392844796716284, "compression/movement_sparsity/linear_layer_sparsity": 0.8048567516354235, "compression/movement_sparsity/model_sparsity": 0.7772074461877669, "compression_loss": 84.64590454101562, "distillation_loss": 4.183196067810059, "epoch": 3.23, "learning_rate": 4.360050093926112e-05, "loss": 88.3384, "step": 3819, "task_loss": 2.28397798538208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7939381567855177, "compression/movement_sparsity/importance_threshold": -0.0013373712164186794, "compression/movement_sparsity/linear_layer_sparsity": 0.8051220762894891, "compression/movement_sparsity/model_sparsity": 0.7774636561237313, "compression_loss": 84.6767807006836, "distillation_loss": 3.4125051498413086, "epoch": 3.23, "learning_rate": 4.359737006887915e-05, "loss": 88.0735, "step": 3820, "task_loss": 3.5458414554595947 }, { "compression/movement_sparsity/importance_regularization_factor": 0.794232671047564, "compression/movement_sparsity/importance_threshold": -0.001335459776189177, "compression/movement_sparsity/linear_layer_sparsity": 0.8054958273998675, "compression/movement_sparsity/model_sparsity": 0.7778245677336754, "compression_loss": 84.7076644897461, "distillation_loss": 4.9235334396362305, "epoch": 3.23, "learning_rate": 4.359423919849718e-05, "loss": 88.3002, "step": 3821, "task_loss": 3.2738828659057617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7945269045524707, "compression/movement_sparsity/importance_threshold": -0.0013335501581141869, "compression/movement_sparsity/linear_layer_sparsity": 0.8057480831662042, "compression/movement_sparsity/model_sparsity": 0.7780681577384089, "compression_loss": 84.73851013183594, "distillation_loss": 3.390097141265869, "epoch": 3.23, "learning_rate": 4.359110832811521e-05, "loss": 88.4844, "step": 3822, "task_loss": 2.723334312438965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7948208574341225, "compression/movement_sparsity/importance_threshold": -0.0013316423613247775, "compression/movement_sparsity/linear_layer_sparsity": 0.8060095682382911, "compression/movement_sparsity/model_sparsity": 0.7783206599938474, "compression_loss": 84.76934051513672, "distillation_loss": 4.714110374450684, "epoch": 3.23, "learning_rate": 4.358797745773325e-05, "loss": 88.6622, "step": 3823, "task_loss": 1.773148775100708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7951145298264047, "compression/movement_sparsity/importance_threshold": -0.0013297363849520127, "compression/movement_sparsity/linear_layer_sparsity": 0.806277337346722, "compression/movement_sparsity/model_sparsity": 0.7785792304096495, "compression_loss": 84.80013275146484, "distillation_loss": 3.418839931488037, "epoch": 3.23, "learning_rate": 4.358484658735128e-05, "loss": 87.9292, "step": 3824, "task_loss": 1.603992223739624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7954079218632024, "compression/movement_sparsity/importance_threshold": -0.001327832228126961, "compression/movement_sparsity/linear_layer_sparsity": 0.8063986776765846, "compression/movement_sparsity/model_sparsity": 0.7786964023258948, "compression_loss": 84.83088684082031, "distillation_loss": 4.064653396606445, "epoch": 3.23, "learning_rate": 4.358171571696932e-05, "loss": 87.973, "step": 3825, "task_loss": 1.5593812465667725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.795701033678401, "compression/movement_sparsity/importance_threshold": -0.0013259298899806855, "compression/movement_sparsity/linear_layer_sparsity": 0.806692202987109, "compression/movement_sparsity/model_sparsity": 0.7789798441390132, "compression_loss": 84.86157989501953, "distillation_loss": 3.0239126682281494, "epoch": 3.23, "learning_rate": 4.357858484658735e-05, "loss": 88.4998, "step": 3826, "task_loss": 1.8976010084152222 }, { "compression/movement_sparsity/importance_regularization_factor": 0.795993865405885, "compression/movement_sparsity/importance_threshold": -0.0013240293696442553, "compression/movement_sparsity/linear_layer_sparsity": 0.8069478094445512, "compression/movement_sparsity/model_sparsity": 0.779226669728305, "compression_loss": 84.8922119140625, "distillation_loss": 2.9869577884674072, "epoch": 3.23, "learning_rate": 4.3575453976205385e-05, "loss": 88.8509, "step": 3827, "task_loss": 1.6134488582611084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7962864171795404, "compression/movement_sparsity/importance_threshold": -0.0013221306662487328, "compression/movement_sparsity/linear_layer_sparsity": 0.8070828864155304, "compression/movement_sparsity/model_sparsity": 0.7793571063897855, "compression_loss": 84.92285919189453, "distillation_loss": 5.790309906005859, "epoch": 3.24, "learning_rate": 4.3572323105823423e-05, "loss": 88.9342, "step": 3828, "task_loss": 3.77681827545166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7965786891332516, "compression/movement_sparsity/importance_threshold": -0.0013202337789251879, "compression/movement_sparsity/linear_layer_sparsity": 0.8073408180856617, "compression/movement_sparsity/model_sparsity": 0.7796061773135572, "compression_loss": 84.95350646972656, "distillation_loss": 4.185464382171631, "epoch": 3.24, "learning_rate": 4.3569192235441455e-05, "loss": 89.4671, "step": 3829, "task_loss": 2.0793468952178955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7968706814009041, "compression/movement_sparsity/importance_threshold": -0.0013183387068046847, "compression/movement_sparsity/linear_layer_sparsity": 0.8074765628100283, "compression/movement_sparsity/model_sparsity": 0.7797372587890423, "compression_loss": 84.98406982421875, "distillation_loss": 3.3836236000061035, "epoch": 3.24, "learning_rate": 4.3566061365059494e-05, "loss": 88.7265, "step": 3830, "task_loss": 1.9048395156860352 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7971623941163832, "compression/movement_sparsity/importance_threshold": -0.0013164454490182881, "compression/movement_sparsity/linear_layer_sparsity": 0.8077464782686338, "compression/movement_sparsity/model_sparsity": 0.7799979018212875, "compression_loss": 85.01464080810547, "distillation_loss": 2.7353830337524414, "epoch": 3.24, "learning_rate": 4.3562930494677526e-05, "loss": 88.212, "step": 3831, "task_loss": 1.4454268217086792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7974538274135735, "compression/movement_sparsity/importance_threshold": -0.0013145540046970673, "compression/movement_sparsity/linear_layer_sparsity": 0.8079936901120605, "compression/movement_sparsity/model_sparsity": 0.7802366211773799, "compression_loss": 85.04515075683594, "distillation_loss": 6.1675639152526855, "epoch": 3.24, "learning_rate": 4.355979962429556e-05, "loss": 89.0643, "step": 3832, "task_loss": 2.4130935668945312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7977449814263609, "compression/movement_sparsity/importance_threshold": -0.0013126643729720845, "compression/movement_sparsity/linear_layer_sparsity": 0.8083204719261213, "compression/movement_sparsity/model_sparsity": 0.7805521770308294, "compression_loss": 85.07565307617188, "distillation_loss": 4.024234294891357, "epoch": 3.24, "learning_rate": 4.355666875391359e-05, "loss": 88.7409, "step": 3833, "task_loss": 2.1109867095947266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7980358562886297, "compression/movement_sparsity/importance_threshold": -0.001310776552974409, "compression/movement_sparsity/linear_layer_sparsity": 0.8084153406038322, "compression/movement_sparsity/model_sparsity": 0.7806437866776107, "compression_loss": 85.1061782836914, "distillation_loss": 5.340185165405273, "epoch": 3.24, "learning_rate": 4.355353788353163e-05, "loss": 89.4317, "step": 3834, "task_loss": 4.018442153930664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7983264521342657, "compression/movement_sparsity/importance_threshold": -0.0013088905438351046, "compression/movement_sparsity/linear_layer_sparsity": 0.8086485653986221, "compression/movement_sparsity/model_sparsity": 0.7808689994832161, "compression_loss": 85.1365737915039, "distillation_loss": 4.111159324645996, "epoch": 3.24, "learning_rate": 4.355040701314966e-05, "loss": 89.0224, "step": 3835, "task_loss": 2.020982027053833 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7986167690971536, "compression/movement_sparsity/importance_threshold": -0.001307006344685239, "compression/movement_sparsity/linear_layer_sparsity": 0.8088948352328056, "compression/movement_sparsity/model_sparsity": 0.7811068091909807, "compression_loss": 85.1670150756836, "distillation_loss": 4.027480125427246, "epoch": 3.24, "learning_rate": 4.354727614276769e-05, "loss": 88.6136, "step": 3836, "task_loss": 1.8155384063720703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.798906807311179, "compression/movement_sparsity/importance_threshold": -0.001305123954655877, "compression/movement_sparsity/linear_layer_sparsity": 0.8091333066613552, "compression/movement_sparsity/model_sparsity": 0.7813370883923358, "compression_loss": 85.1974868774414, "distillation_loss": 4.765969753265381, "epoch": 3.24, "learning_rate": 4.354414527238573e-05, "loss": 88.8596, "step": 3837, "task_loss": 2.0147860050201416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7991965669102266, "compression/movement_sparsity/importance_threshold": -0.001303243372878086, "compression/movement_sparsity/linear_layer_sparsity": 0.8093694051805452, "compression/movement_sparsity/model_sparsity": 0.7815650762010677, "compression_loss": 85.22784423828125, "distillation_loss": 4.259750843048096, "epoch": 3.24, "learning_rate": 4.354101440200376e-05, "loss": 89.2729, "step": 3838, "task_loss": 2.1506028175354004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7994860480281816, "compression/movement_sparsity/importance_threshold": -0.001301364598482931, "compression/movement_sparsity/linear_layer_sparsity": 0.8097744572309679, "compression/movement_sparsity/model_sparsity": 0.7819562134674725, "compression_loss": 85.2582015991211, "distillation_loss": 4.023374557495117, "epoch": 3.24, "learning_rate": 4.353788353162179e-05, "loss": 88.8609, "step": 3839, "task_loss": 1.8266702890396118 }, { "compression/movement_sparsity/importance_regularization_factor": 0.7997752507989294, "compression/movement_sparsity/importance_threshold": -0.0012994876306014775, "compression/movement_sparsity/linear_layer_sparsity": 0.8098905032304001, "compression/movement_sparsity/model_sparsity": 0.782068272929825, "compression_loss": 85.28853607177734, "distillation_loss": 3.0768375396728516, "epoch": 3.25, "learning_rate": 4.3534752661239825e-05, "loss": 88.23, "step": 3840, "task_loss": 1.4302870035171509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8000641753563549, "compression/movement_sparsity/importance_threshold": -0.0012976124683647932, "compression/movement_sparsity/linear_layer_sparsity": 0.8101868068719836, "compression/movement_sparsity/model_sparsity": 0.7823543976297834, "compression_loss": 85.31877136230469, "distillation_loss": 6.2855329513549805, "epoch": 3.25, "learning_rate": 4.3531621790857864e-05, "loss": 89.6141, "step": 3841, "task_loss": 3.804384231567383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8003528218343435, "compression/movement_sparsity/importance_threshold": -0.001295739110903942, "compression/movement_sparsity/linear_layer_sparsity": 0.8104431884003224, "compression/movement_sparsity/model_sparsity": 0.7826019716639019, "compression_loss": 85.34904479980469, "distillation_loss": 4.333503723144531, "epoch": 3.25, "learning_rate": 4.3528490920475895e-05, "loss": 89.9863, "step": 3842, "task_loss": 2.6990628242492676 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8006411903667799, "compression/movement_sparsity/importance_threshold": -0.0012938675573499932, "compression/movement_sparsity/linear_layer_sparsity": 0.8106029603224753, "compression/movement_sparsity/model_sparsity": 0.7827562549290129, "compression_loss": 85.37931060791016, "distillation_loss": 5.459375381469727, "epoch": 3.25, "learning_rate": 4.352536005009393e-05, "loss": 89.6308, "step": 3843, "task_loss": 2.6345577239990234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8009292810875499, "compression/movement_sparsity/importance_threshold": -0.001291997806834008, "compression/movement_sparsity/linear_layer_sparsity": 0.8107678357883764, "compression/movement_sparsity/model_sparsity": 0.7829154664154441, "compression_loss": 85.4095458984375, "distillation_loss": 3.0848135948181152, "epoch": 3.25, "learning_rate": 4.352222917971196e-05, "loss": 89.5896, "step": 3844, "task_loss": 2.267569065093994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8012170941305379, "compression/movement_sparsity/importance_threshold": -0.0012901298584870576, "compression/movement_sparsity/linear_layer_sparsity": 0.8110920181338925, "compression/movement_sparsity/model_sparsity": 0.7832285121000903, "compression_loss": 85.43973541259766, "distillation_loss": 3.854504108428955, "epoch": 3.25, "learning_rate": 4.351909830933e-05, "loss": 89.4359, "step": 3845, "task_loss": 1.8507663011550903 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8015046296296298, "compression/movement_sparsity/importance_threshold": -0.0012882637114402032, "compression/movement_sparsity/linear_layer_sparsity": 0.8112738378420042, "compression/movement_sparsity/model_sparsity": 0.7834040857418854, "compression_loss": 85.4698715209961, "distillation_loss": 3.213432788848877, "epoch": 3.25, "learning_rate": 4.351596743894803e-05, "loss": 89.3834, "step": 3846, "task_loss": 1.6765257120132446 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8017918877187099, "compression/movement_sparsity/importance_threshold": -0.0012863993648245167, "compression/movement_sparsity/linear_layer_sparsity": 0.8115227786897381, "compression/movement_sparsity/model_sparsity": 0.783644474705668, "compression_loss": 85.5000228881836, "distillation_loss": 4.9170355796813965, "epoch": 3.25, "learning_rate": 4.351283656856606e-05, "loss": 89.0048, "step": 3847, "task_loss": 3.706254720687866 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8020788685316641, "compression/movement_sparsity/importance_threshold": -0.0012845368177710577, "compression/movement_sparsity/linear_layer_sparsity": 0.8116546122871202, "compression/movement_sparsity/model_sparsity": 0.7837717794134125, "compression_loss": 85.5301284790039, "distillation_loss": 3.7830729484558105, "epoch": 3.25, "learning_rate": 4.35097056981841e-05, "loss": 89.12, "step": 3848, "task_loss": 2.2166600227355957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8023655722023769, "compression/movement_sparsity/importance_threshold": -0.001282676069410898, "compression/movement_sparsity/linear_layer_sparsity": 0.811807313177865, "compression/movement_sparsity/model_sparsity": 0.7839192345587973, "compression_loss": 85.56021118164062, "distillation_loss": 4.279556751251221, "epoch": 3.25, "learning_rate": 4.350657482780213e-05, "loss": 89.842, "step": 3849, "task_loss": 3.0454111099243164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.802651998864734, "compression/movement_sparsity/importance_threshold": -0.001280817118875099, "compression/movement_sparsity/linear_layer_sparsity": 0.8121702490681978, "compression/movement_sparsity/model_sparsity": 0.7842697024847758, "compression_loss": 85.59017944335938, "distillation_loss": 3.8417630195617676, "epoch": 3.25, "learning_rate": 4.350344395742016e-05, "loss": 88.9822, "step": 3850, "task_loss": 1.399146318435669 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8029381486526203, "compression/movement_sparsity/importance_threshold": -0.0012789599652947292, "compression/movement_sparsity/linear_layer_sparsity": 0.8124042727822193, "compression/movement_sparsity/model_sparsity": 0.7844956867642795, "compression_loss": 85.62015533447266, "distillation_loss": 6.126126766204834, "epoch": 3.26, "learning_rate": 4.3500313087038195e-05, "loss": 90.0695, "step": 3851, "task_loss": 3.8916678428649902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8032240216999207, "compression/movement_sparsity/importance_threshold": -0.001277104607800855, "compression/movement_sparsity/linear_layer_sparsity": 0.8126011050173845, "compression/movement_sparsity/model_sparsity": 0.7846857572066399, "compression_loss": 85.65007781982422, "distillation_loss": 4.733682632446289, "epoch": 3.26, "learning_rate": 4.3497182216656234e-05, "loss": 89.9214, "step": 3852, "task_loss": 2.012526512145996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8035096181405209, "compression/movement_sparsity/importance_threshold": -0.0012752510455245397, "compression/movement_sparsity/linear_layer_sparsity": 0.812871342428516, "compression/movement_sparsity/model_sparsity": 0.7849467111313516, "compression_loss": 85.68001556396484, "distillation_loss": 3.6934170722961426, "epoch": 3.26, "learning_rate": 4.3494051346274265e-05, "loss": 89.5187, "step": 3853, "task_loss": 1.3038341999053955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8037949381083055, "compression/movement_sparsity/importance_threshold": -0.0012733992775968532, "compression/movement_sparsity/linear_layer_sparsity": 0.8130234590350468, "compression/movement_sparsity/model_sparsity": 0.7850936020644825, "compression_loss": 85.70986938476562, "distillation_loss": 4.145567417144775, "epoch": 3.26, "learning_rate": 4.34909204758923e-05, "loss": 90.3762, "step": 3854, "task_loss": 2.2323107719421387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.80407998173716, "compression/movement_sparsity/importance_threshold": -0.0012715493031488579, "compression/movement_sparsity/linear_layer_sparsity": 0.8132621450986138, "compression/movement_sparsity/model_sparsity": 0.7853240885274819, "compression_loss": 85.73973846435547, "distillation_loss": 3.54253888130188, "epoch": 3.26, "learning_rate": 4.348778960551033e-05, "loss": 88.8834, "step": 3855, "task_loss": 2.3486838340759277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8043647491609693, "compression/movement_sparsity/importance_threshold": -0.001269701121311622, "compression/movement_sparsity/linear_layer_sparsity": 0.8132775153506965, "compression/movement_sparsity/model_sparsity": 0.7853389307641211, "compression_loss": 85.7695541381836, "distillation_loss": 3.0010581016540527, "epoch": 3.26, "learning_rate": 4.348465873512837e-05, "loss": 89.1709, "step": 3856, "task_loss": 3.0481648445129395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8046492405136187, "compression/movement_sparsity/importance_threshold": -0.0012678547312162114, "compression/movement_sparsity/linear_layer_sparsity": 0.8134820029014839, "compression/movement_sparsity/model_sparsity": 0.7855363935384616, "compression_loss": 85.79932403564453, "distillation_loss": 3.671555757522583, "epoch": 3.26, "learning_rate": 4.34815278647464e-05, "loss": 89.1039, "step": 3857, "task_loss": 2.639066457748413 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8049334559289932, "compression/movement_sparsity/importance_threshold": -0.0012660101319936917, "compression/movement_sparsity/linear_layer_sparsity": 0.813693048744471, "compression/movement_sparsity/model_sparsity": 0.7857401893074892, "compression_loss": 85.82904052734375, "distillation_loss": 2.368487596511841, "epoch": 3.26, "learning_rate": 4.347839699436443e-05, "loss": 89.0983, "step": 3858, "task_loss": 1.7626569271087646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8052173955409783, "compression/movement_sparsity/importance_threshold": -0.0012641673227751287, "compression/movement_sparsity/linear_layer_sparsity": 0.8139710368645658, "compression/movement_sparsity/model_sparsity": 0.7860086276804673, "compression_loss": 85.85870361328125, "distillation_loss": 5.308958053588867, "epoch": 3.26, "learning_rate": 4.347526612398247e-05, "loss": 90.5179, "step": 3859, "task_loss": 2.5066640377044678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8055010594834585, "compression/movement_sparsity/importance_threshold": -0.001262326302691589, "compression/movement_sparsity/linear_layer_sparsity": 0.8141804729449221, "compression/movement_sparsity/model_sparsity": 0.7862108689871626, "compression_loss": 85.88839721679688, "distillation_loss": 4.007132530212402, "epoch": 3.26, "learning_rate": 4.34721352536005e-05, "loss": 90.2084, "step": 3860, "task_loss": 2.0864479541778564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8057844478903194, "compression/movement_sparsity/importance_threshold": -0.0012604870708741393, "compression/movement_sparsity/linear_layer_sparsity": 0.8144438658838308, "compression/movement_sparsity/model_sparsity": 0.7864652135683283, "compression_loss": 85.91801452636719, "distillation_loss": 4.0522141456604, "epoch": 3.26, "learning_rate": 4.346900438321854e-05, "loss": 89.2583, "step": 3861, "task_loss": 2.573652505874634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8060675608954462, "compression/movement_sparsity/importance_threshold": -0.0012586496264538434, "compression/movement_sparsity/linear_layer_sparsity": 0.8147086181778498, "compression/movement_sparsity/model_sparsity": 0.7867208708065745, "compression_loss": 85.94766235351562, "distillation_loss": 2.813418388366699, "epoch": 3.26, "learning_rate": 4.346587351283657e-05, "loss": 89.1213, "step": 3862, "task_loss": 1.7635328769683838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8063503986327237, "compression/movement_sparsity/importance_threshold": -0.0012568139685617697, "compression/movement_sparsity/linear_layer_sparsity": 0.8149782832289348, "compression/movement_sparsity/model_sparsity": 0.786981272033568, "compression_loss": 85.9771957397461, "distillation_loss": 3.043480396270752, "epoch": 3.27, "learning_rate": 4.346274264245461e-05, "loss": 90.2685, "step": 3863, "task_loss": 0.9518271088600159 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8066329612360374, "compression/movement_sparsity/importance_threshold": -0.0012549800963289814, "compression/movement_sparsity/linear_layer_sparsity": 0.8151450069408195, "compression/movement_sparsity/model_sparsity": 0.7871422682730472, "compression_loss": 86.0066909790039, "distillation_loss": 3.2677416801452637, "epoch": 3.27, "learning_rate": 4.345961177207264e-05, "loss": 89.5639, "step": 3864, "task_loss": 1.4678560495376587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8069152488392721, "compression/movement_sparsity/importance_threshold": -0.0012531480088865486, "compression/movement_sparsity/linear_layer_sparsity": 0.815338750816567, "compression/movement_sparsity/model_sparsity": 0.787329356450637, "compression_loss": 86.0361557006836, "distillation_loss": 3.1342668533325195, "epoch": 3.27, "learning_rate": 4.3456480901690674e-05, "loss": 89.6114, "step": 3865, "task_loss": 1.3278528451919556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8071972615763134, "compression/movement_sparsity/importance_threshold": -0.0012513177053655325, "compression/movement_sparsity/linear_layer_sparsity": 0.8154871113102924, "compression/movement_sparsity/model_sparsity": 0.7874726203049927, "compression_loss": 86.06564331054688, "distillation_loss": 2.928405523300171, "epoch": 3.27, "learning_rate": 4.3453350031308706e-05, "loss": 89.6424, "step": 3866, "task_loss": 1.8039101362228394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8074789995810459, "compression/movement_sparsity/importance_threshold": -0.0012494891848970043, "compression/movement_sparsity/linear_layer_sparsity": 0.8157053831988669, "compression/movement_sparsity/model_sparsity": 0.7876833938827117, "compression_loss": 86.09504699707031, "distillation_loss": 4.7323899269104, "epoch": 3.27, "learning_rate": 4.3450219160926744e-05, "loss": 90.1703, "step": 3867, "task_loss": 2.8476650714874268 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8077604629873552, "compression/movement_sparsity/importance_threshold": -0.0012476624466120252, "compression/movement_sparsity/linear_layer_sparsity": 0.8159702070378918, "compression/movement_sparsity/model_sparsity": 0.7879391202081727, "compression_loss": 86.12445831298828, "distillation_loss": 4.4475417137146, "epoch": 3.27, "learning_rate": 4.3447088290544776e-05, "loss": 89.8566, "step": 3868, "task_loss": 1.5177602767944336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8080416519291259, "compression/movement_sparsity/importance_threshold": -0.0012458374896416654, "compression/movement_sparsity/linear_layer_sparsity": 0.8161087897141558, "compression/movement_sparsity/model_sparsity": 0.7880729421431768, "compression_loss": 86.15388488769531, "distillation_loss": 5.564089775085449, "epoch": 3.27, "learning_rate": 4.344395742016281e-05, "loss": 90.5112, "step": 3869, "task_loss": 2.812685251235962 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8083225665402436, "compression/movement_sparsity/importance_threshold": -0.0012440143131169888, "compression/movement_sparsity/linear_layer_sparsity": 0.8163043579875515, "compression/movement_sparsity/model_sparsity": 0.7882617920447431, "compression_loss": 86.18326568603516, "distillation_loss": 6.386137008666992, "epoch": 3.27, "learning_rate": 4.344082654978084e-05, "loss": 90.4801, "step": 3870, "task_loss": 3.124272108078003 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8086032069545934, "compression/movement_sparsity/importance_threshold": -0.0012421929161690612, "compression/movement_sparsity/linear_layer_sparsity": 0.8165019295211101, "compression/movement_sparsity/model_sparsity": 0.7884525763883229, "compression_loss": 86.212646484375, "distillation_loss": 3.467851161956787, "epoch": 3.27, "learning_rate": 4.343769567939888e-05, "loss": 89.5143, "step": 3871, "task_loss": 2.3756535053253174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8088835733060602, "compression/movement_sparsity/importance_threshold": -0.00124037329792895, "compression/movement_sparsity/linear_layer_sparsity": 0.8166116914841983, "compression/movement_sparsity/model_sparsity": 0.7885585676903116, "compression_loss": 86.24186706542969, "distillation_loss": 4.594011306762695, "epoch": 3.27, "learning_rate": 4.343456480901691e-05, "loss": 90.7712, "step": 3872, "task_loss": 2.7722079753875732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8091636657285296, "compression/movement_sparsity/importance_threshold": -0.0012385554575277185, "compression/movement_sparsity/linear_layer_sparsity": 0.8168153443432512, "compression/movement_sparsity/model_sparsity": 0.7887552244471465, "compression_loss": 86.27118682861328, "distillation_loss": 4.311746597290039, "epoch": 3.27, "learning_rate": 4.343143393863494e-05, "loss": 90.0962, "step": 3873, "task_loss": 2.836796998977661 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8094434843558862, "compression/movement_sparsity/importance_threshold": -0.0012367393940964358, "compression/movement_sparsity/linear_layer_sparsity": 0.8169836062727609, "compression/movement_sparsity/model_sparsity": 0.7889177060617433, "compression_loss": 86.30043029785156, "distillation_loss": 4.756231307983398, "epoch": 3.27, "learning_rate": 4.342830306825298e-05, "loss": 90.2284, "step": 3874, "task_loss": 2.1445376873016357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8097230293220155, "compression/movement_sparsity/importance_threshold": -0.0012349251067661658, "compression/movement_sparsity/linear_layer_sparsity": 0.8170487480005556, "compression/movement_sparsity/model_sparsity": 0.7889806099707889, "compression_loss": 86.32965087890625, "distillation_loss": 3.9638192653656006, "epoch": 3.28, "learning_rate": 4.342517219787101e-05, "loss": 90.3917, "step": 3875, "task_loss": 1.8411422967910767 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8100023007608024, "compression/movement_sparsity/importance_threshold": -0.001233112594667977, "compression/movement_sparsity/linear_layer_sparsity": 0.8172858839223299, "compression/movement_sparsity/model_sparsity": 0.7892095995441349, "compression_loss": 86.35884857177734, "distillation_loss": 3.9312593936920166, "epoch": 3.28, "learning_rate": 4.3422041327489044e-05, "loss": 90.0061, "step": 3876, "task_loss": 2.442305564880371 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8102812988061322, "compression/movement_sparsity/importance_threshold": -0.0012313018569329332, "compression/movement_sparsity/linear_layer_sparsity": 0.8175040127208928, "compression/movement_sparsity/model_sparsity": 0.7894202349474244, "compression_loss": 86.38801574707031, "distillation_loss": 3.940225124359131, "epoch": 3.28, "learning_rate": 4.3418910457107076e-05, "loss": 90.1629, "step": 3877, "task_loss": 2.2102694511413574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8105600235918899, "compression/movement_sparsity/importance_threshold": -0.0012294928926921012, "compression/movement_sparsity/linear_layer_sparsity": 0.8177827401393811, "compression/movement_sparsity/model_sparsity": 0.7896893872216219, "compression_loss": 86.41719055175781, "distillation_loss": 4.7032341957092285, "epoch": 3.28, "learning_rate": 4.3415779586725114e-05, "loss": 90.3257, "step": 3878, "task_loss": 3.4721083641052246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8108384752519607, "compression/movement_sparsity/importance_threshold": -0.0012276857010765476, "compression/movement_sparsity/linear_layer_sparsity": 0.8179859518042315, "compression/movement_sparsity/model_sparsity": 0.7898856179406324, "compression_loss": 86.44628143310547, "distillation_loss": 3.692560911178589, "epoch": 3.28, "learning_rate": 4.3412648716343146e-05, "loss": 90.0874, "step": 3879, "task_loss": 1.6042381525039673 }, { "compression/movement_sparsity/importance_regularization_factor": 0.81111665392023, "compression/movement_sparsity/importance_threshold": -0.0012258802812173371, "compression/movement_sparsity/linear_layer_sparsity": 0.8182683757146869, "compression/movement_sparsity/model_sparsity": 0.7901583397209261, "compression_loss": 86.47535705566406, "distillation_loss": 4.73038387298584, "epoch": 3.28, "learning_rate": 4.340951784596118e-05, "loss": 90.6527, "step": 3880, "task_loss": 3.3917994499206543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8113945597305825, "compression/movement_sparsity/importance_threshold": -0.0012240766322455373, "compression/movement_sparsity/linear_layer_sparsity": 0.8184405726195164, "compression/movement_sparsity/model_sparsity": 0.790324621132335, "compression_loss": 86.50443267822266, "distillation_loss": 2.7732439041137695, "epoch": 3.28, "learning_rate": 4.340638697557921e-05, "loss": 89.9843, "step": 3881, "task_loss": 0.8700898885726929 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8116721928169038, "compression/movement_sparsity/importance_threshold": -0.0012222747532922122, "compression/movement_sparsity/linear_layer_sparsity": 0.8186168237413421, "compression/movement_sparsity/model_sparsity": 0.7904948174859142, "compression_loss": 86.53350830078125, "distillation_loss": 4.02028751373291, "epoch": 3.28, "learning_rate": 4.340325610519725e-05, "loss": 90.3836, "step": 3882, "task_loss": 2.2759335041046143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8119495533130786, "compression/movement_sparsity/importance_threshold": -0.0012204746434884301, "compression/movement_sparsity/linear_layer_sparsity": 0.8187957578008859, "compression/movement_sparsity/model_sparsity": 0.790667604610047, "compression_loss": 86.56245422363281, "distillation_loss": 3.8505699634552, "epoch": 3.28, "learning_rate": 4.340012523481528e-05, "loss": 90.4794, "step": 3883, "task_loss": 2.1540467739105225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8122266413529925, "compression/movement_sparsity/importance_threshold": -0.0012186763019652542, "compression/movement_sparsity/linear_layer_sparsity": 0.8189844339053881, "compression/movement_sparsity/model_sparsity": 0.790849799109924, "compression_loss": 86.5914306640625, "distillation_loss": 3.4753365516662598, "epoch": 3.28, "learning_rate": 4.339699436443331e-05, "loss": 90.2742, "step": 3884, "task_loss": 1.5623728036880493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8125034570705301, "compression/movement_sparsity/importance_threshold": -0.0012168797278537544, "compression/movement_sparsity/linear_layer_sparsity": 0.8192670486025257, "compression/movement_sparsity/model_sparsity": 0.7911227051227905, "compression_loss": 86.6203842163086, "distillation_loss": 3.873098850250244, "epoch": 3.28, "learning_rate": 4.339386349405135e-05, "loss": 90.2662, "step": 3885, "task_loss": 2.1524300575256348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.812780000599577, "compression/movement_sparsity/importance_threshold": -0.001215084920284994, "compression/movement_sparsity/linear_layer_sparsity": 0.8194799188431612, "compression/movement_sparsity/model_sparsity": 0.7913282626157946, "compression_loss": 86.64928436279297, "distillation_loss": 3.952699661254883, "epoch": 3.28, "learning_rate": 4.339073262366938e-05, "loss": 90.2874, "step": 3886, "task_loss": 2.7643043994903564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.813056272074018, "compression/movement_sparsity/importance_threshold": -0.0012132918783900404, "compression/movement_sparsity/linear_layer_sparsity": 0.8197161723765305, "compression/movement_sparsity/model_sparsity": 0.7915564001134918, "compression_loss": 86.67810821533203, "distillation_loss": 4.497793197631836, "epoch": 3.29, "learning_rate": 4.3387601753287414e-05, "loss": 90.7225, "step": 3887, "task_loss": 2.2384889125823975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8133322716277386, "compression/movement_sparsity/importance_threshold": -0.0012115006012999576, "compression/movement_sparsity/linear_layer_sparsity": 0.8198076188181299, "compression/movement_sparsity/model_sparsity": 0.7916447050885002, "compression_loss": 86.70701599121094, "distillation_loss": 3.3626246452331543, "epoch": 3.29, "learning_rate": 4.3384470882905445e-05, "loss": 90.2157, "step": 3888, "task_loss": 2.083061695098877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8136079993946235, "compression/movement_sparsity/importance_threshold": -0.001209711088145815, "compression/movement_sparsity/linear_layer_sparsity": 0.819989784327103, "compression/movement_sparsity/model_sparsity": 0.7918206126518335, "compression_loss": 86.73580169677734, "distillation_loss": 4.343718528747559, "epoch": 3.29, "learning_rate": 4.3381340012523484e-05, "loss": 90.8514, "step": 3889, "task_loss": 2.814337730407715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8138834555085583, "compression/movement_sparsity/importance_threshold": -0.0012079233380586736, "compression/movement_sparsity/linear_layer_sparsity": 0.8201335897887915, "compression/movement_sparsity/model_sparsity": 0.7919594779535155, "compression_loss": 86.76456451416016, "distillation_loss": 3.236767530441284, "epoch": 3.29, "learning_rate": 4.3378209142141516e-05, "loss": 90.9971, "step": 3890, "task_loss": 2.2612791061401367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8141586401034278, "compression/movement_sparsity/importance_threshold": -0.0012061373501696055, "compression/movement_sparsity/linear_layer_sparsity": 0.8203129888908731, "compression/movement_sparsity/model_sparsity": 0.7921327141445444, "compression_loss": 86.79326629638672, "distillation_loss": 3.86181640625, "epoch": 3.29, "learning_rate": 4.337507827175955e-05, "loss": 90.9433, "step": 3891, "task_loss": 2.404254913330078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8144335533131172, "compression/movement_sparsity/importance_threshold": -0.0012043531236096729, "compression/movement_sparsity/linear_layer_sparsity": 0.8205763341331112, "compression/movement_sparsity/model_sparsity": 0.7923870126675668, "compression_loss": 86.82193756103516, "distillation_loss": 4.063754558563232, "epoch": 3.29, "learning_rate": 4.3371947401377586e-05, "loss": 91.0126, "step": 3892, "task_loss": 2.303164005279541 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8147081952715118, "compression/movement_sparsity/importance_threshold": -0.0012025706575099424, "compression/movement_sparsity/linear_layer_sparsity": 0.8207901106104869, "compression/movement_sparsity/model_sparsity": 0.7925934452652914, "compression_loss": 86.85059356689453, "distillation_loss": 5.306834697723389, "epoch": 3.29, "learning_rate": 4.336881653099562e-05, "loss": 90.8207, "step": 3893, "task_loss": 2.6821224689483643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8149825661124966, "compression/movement_sparsity/importance_threshold": -0.0012007899510014805, "compression/movement_sparsity/linear_layer_sparsity": 0.8209526727878665, "compression/movement_sparsity/model_sparsity": 0.7927504229317783, "compression_loss": 86.8792724609375, "distillation_loss": 5.74947452545166, "epoch": 3.29, "learning_rate": 4.336568566061365e-05, "loss": 91.1105, "step": 3894, "task_loss": 2.1763570308685303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.815256665969957, "compression/movement_sparsity/importance_threshold": -0.0011990110032153514, "compression/movement_sparsity/linear_layer_sparsity": 0.8212295638845389, "compression/movement_sparsity/model_sparsity": 0.7930178019674634, "compression_loss": 86.90792083740234, "distillation_loss": 3.5070157051086426, "epoch": 3.29, "learning_rate": 4.336255479023169e-05, "loss": 90.808, "step": 3895, "task_loss": 1.565128207206726 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8155304949777777, "compression/movement_sparsity/importance_threshold": -0.001197233813282625, "compression/movement_sparsity/linear_layer_sparsity": 0.8213955602221978, "compression/movement_sparsity/model_sparsity": 0.7931780958202592, "compression_loss": 86.93647766113281, "distillation_loss": 2.8362417221069336, "epoch": 3.29, "learning_rate": 4.335942391984972e-05, "loss": 90.4715, "step": 3896, "task_loss": 1.4448102712631226 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8158040532698443, "compression/movement_sparsity/importance_threshold": -0.0011954583803343638, "compression/movement_sparsity/linear_layer_sparsity": 0.8214760006570694, "compression/movement_sparsity/model_sparsity": 0.7932557728787292, "compression_loss": 86.96501159667969, "distillation_loss": 3.2832934856414795, "epoch": 3.29, "learning_rate": 4.335629304946776e-05, "loss": 90.3374, "step": 3897, "task_loss": 3.0713555812835693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8160773409800414, "compression/movement_sparsity/importance_threshold": -0.0011936847035016367, "compression/movement_sparsity/linear_layer_sparsity": 0.8218868120804601, "compression/movement_sparsity/model_sparsity": 0.7936524716659227, "compression_loss": 86.99351501464844, "distillation_loss": 5.042202472686768, "epoch": 3.29, "learning_rate": 4.335316217908579e-05, "loss": 91.0032, "step": 3898, "task_loss": 1.6975409984588623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8163503582422547, "compression/movement_sparsity/importance_threshold": -0.001191912781915507, "compression/movement_sparsity/linear_layer_sparsity": 0.822104511608988, "compression/movement_sparsity/model_sparsity": 0.7938626925459237, "compression_loss": 87.02205657958984, "distillation_loss": 3.8549537658691406, "epoch": 3.3, "learning_rate": 4.335003130870382e-05, "loss": 90.6587, "step": 3899, "task_loss": 2.0667450428009033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8166231051903688, "compression/movement_sparsity/importance_threshold": -0.0011901426147070421, "compression/movement_sparsity/linear_layer_sparsity": 0.8222725112068097, "compression/movement_sparsity/model_sparsity": 0.794024920840733, "compression_loss": 87.05049133300781, "distillation_loss": 3.7254252433776855, "epoch": 3.3, "learning_rate": 4.334690043832186e-05, "loss": 90.7561, "step": 3900, "task_loss": 3.731259822845459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8168955819582693, "compression/movement_sparsity/importance_threshold": -0.0011883742010073087, "compression/movement_sparsity/linear_layer_sparsity": 0.8224663147033954, "compression/movement_sparsity/model_sparsity": 0.7942120665910016, "compression_loss": 87.07894134521484, "distillation_loss": 4.506944179534912, "epoch": 3.3, "learning_rate": 4.334376956793989e-05, "loss": 90.7585, "step": 3901, "task_loss": 2.096888542175293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8171677886798413, "compression/movement_sparsity/importance_threshold": -0.0011866075399473708, "compression/movement_sparsity/linear_layer_sparsity": 0.8226658298762787, "compression/movement_sparsity/model_sparsity": 0.7944047278039158, "compression_loss": 87.10740661621094, "distillation_loss": 4.093010902404785, "epoch": 3.3, "learning_rate": 4.3340638697557924e-05, "loss": 90.8355, "step": 3902, "task_loss": 2.843141555786133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8174397254889696, "compression/movement_sparsity/importance_threshold": -0.0011848426306582966, "compression/movement_sparsity/linear_layer_sparsity": 0.8229142818331395, "compression/movement_sparsity/model_sparsity": 0.7946446446717309, "compression_loss": 87.13570404052734, "distillation_loss": 3.8430702686309814, "epoch": 3.3, "learning_rate": 4.3337507827175956e-05, "loss": 91.0189, "step": 3903, "task_loss": 2.1826038360595703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8177113925195397, "compression/movement_sparsity/importance_threshold": -0.001183079472271151, "compression/movement_sparsity/linear_layer_sparsity": 0.8230192860533408, "compression/movement_sparsity/model_sparsity": 0.7947460416739376, "compression_loss": 87.16403198242188, "distillation_loss": 3.6215457916259766, "epoch": 3.3, "learning_rate": 4.3334376956793995e-05, "loss": 90.3916, "step": 3904, "task_loss": 3.448094367980957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8179827899054365, "compression/movement_sparsity/importance_threshold": -0.001181318063917, "compression/movement_sparsity/linear_layer_sparsity": 0.823012048083586, "compression/movement_sparsity/model_sparsity": 0.7947390523507103, "compression_loss": 87.19232177734375, "distillation_loss": 4.607474327087402, "epoch": 3.3, "learning_rate": 4.3331246086412026e-05, "loss": 91.1128, "step": 3905, "task_loss": 1.77225661277771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8182539177805455, "compression/movement_sparsity/importance_threshold": -0.001179558404726909, "compression/movement_sparsity/linear_layer_sparsity": 0.8233285751134799, "compression/movement_sparsity/model_sparsity": 0.7950447057033764, "compression_loss": 87.22061157226562, "distillation_loss": 3.668797492980957, "epoch": 3.3, "learning_rate": 4.332811521603006e-05, "loss": 91.3936, "step": 3906, "task_loss": 1.6493163108825684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8185247762787512, "compression/movement_sparsity/importance_threshold": -0.0011778004938319472, "compression/movement_sparsity/linear_layer_sparsity": 0.8236101047113626, "compression/movement_sparsity/model_sparsity": 0.7953165638934855, "compression_loss": 87.24883270263672, "distillation_loss": 3.58363676071167, "epoch": 3.3, "learning_rate": 4.332498434564809e-05, "loss": 91.1557, "step": 3907, "task_loss": 2.331023693084717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8187953655339393, "compression/movement_sparsity/importance_threshold": -0.0011760443303631771, "compression/movement_sparsity/linear_layer_sparsity": 0.8238374747398433, "compression/movement_sparsity/model_sparsity": 0.7955361230620158, "compression_loss": 87.27699279785156, "distillation_loss": 3.945359230041504, "epoch": 3.3, "learning_rate": 4.332185347526613e-05, "loss": 91.0653, "step": 3908, "task_loss": 2.270663022994995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8190656856799947, "compression/movement_sparsity/importance_threshold": -0.0011742899134516652, "compression/movement_sparsity/linear_layer_sparsity": 0.8242032128095704, "compression/movement_sparsity/model_sparsity": 0.795889296903906, "compression_loss": 87.30522155761719, "distillation_loss": 4.392266750335693, "epoch": 3.3, "learning_rate": 4.331872260488416e-05, "loss": 91.1743, "step": 3909, "task_loss": 2.715712308883667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8193357368508025, "compression/movement_sparsity/importance_threshold": -0.0011725372422284807, "compression/movement_sparsity/linear_layer_sparsity": 0.8244195410588203, "compression/movement_sparsity/model_sparsity": 0.7960981936122905, "compression_loss": 87.33338928222656, "distillation_loss": 4.622209548950195, "epoch": 3.3, "learning_rate": 4.331559173450219e-05, "loss": 91.7413, "step": 3910, "task_loss": 2.7690091133117676 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8196055191802482, "compression/movement_sparsity/importance_threshold": -0.001170786315824685, "compression/movement_sparsity/linear_layer_sparsity": 0.8246359647014112, "compression/movement_sparsity/model_sparsity": 0.7963071824369614, "compression_loss": 87.36146545410156, "distillation_loss": 4.120675086975098, "epoch": 3.31, "learning_rate": 4.331246086412023e-05, "loss": 91.3939, "step": 3911, "task_loss": 1.4449084997177124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8198750328022164, "compression/movement_sparsity/importance_threshold": -0.0011690371333713482, "compression/movement_sparsity/linear_layer_sparsity": 0.8248545585425119, "compression/movement_sparsity/model_sparsity": 0.796518266907147, "compression_loss": 87.3895492553711, "distillation_loss": 3.046053409576416, "epoch": 3.31, "learning_rate": 4.330932999373826e-05, "loss": 91.2312, "step": 3912, "task_loss": 1.4225794076919556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8201442778505927, "compression/movement_sparsity/importance_threshold": -0.0011672896939995334, "compression/movement_sparsity/linear_layer_sparsity": 0.8250810342584198, "compression/movement_sparsity/model_sparsity": 0.7967369624854926, "compression_loss": 87.4175796508789, "distillation_loss": 4.011629104614258, "epoch": 3.31, "learning_rate": 4.3306199123356294e-05, "loss": 91.049, "step": 3913, "task_loss": 2.3447556495666504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8204132544592619, "compression/movement_sparsity/importance_threshold": -0.001165543996840308, "compression/movement_sparsity/linear_layer_sparsity": 0.8253392521085744, "compression/movement_sparsity/model_sparsity": 0.7969863097581235, "compression_loss": 87.44560241699219, "distillation_loss": 2.796684980392456, "epoch": 3.31, "learning_rate": 4.3303068252974326e-05, "loss": 91.1716, "step": 3914, "task_loss": 1.5780407190322876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8206819627621096, "compression/movement_sparsity/importance_threshold": -0.0011638000410247379, "compression/movement_sparsity/linear_layer_sparsity": 0.8255018619826247, "compression/movement_sparsity/model_sparsity": 0.7971433334827537, "compression_loss": 87.47361755371094, "distillation_loss": 4.926808834075928, "epoch": 3.31, "learning_rate": 4.3299937382592365e-05, "loss": 91.3083, "step": 3915, "task_loss": 2.9893360137939453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8209504028930203, "compression/movement_sparsity/importance_threshold": -0.0011620578256838895, "compression/movement_sparsity/linear_layer_sparsity": 0.8257524125929895, "compression/movement_sparsity/model_sparsity": 0.7973852769088685, "compression_loss": 87.50151824951172, "distillation_loss": 5.891541481018066, "epoch": 3.31, "learning_rate": 4.3296806512210396e-05, "loss": 92.2765, "step": 3916, "task_loss": 3.8272228240966797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8212185749858798, "compression/movement_sparsity/importance_threshold": -0.001160317349948827, "compression/movement_sparsity/linear_layer_sparsity": 0.8259065563080182, "compression/movement_sparsity/model_sparsity": 0.7975341253130845, "compression_loss": 87.52947998046875, "distillation_loss": 4.357071399688721, "epoch": 3.31, "learning_rate": 4.329367564182843e-05, "loss": 91.1771, "step": 3917, "task_loss": 1.7489439249038696 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8214864791745726, "compression/movement_sparsity/importance_threshold": -0.0011585786129506203, "compression/movement_sparsity/linear_layer_sparsity": 0.8261034362398539, "compression/movement_sparsity/model_sparsity": 0.7977242418135883, "compression_loss": 87.55738830566406, "distillation_loss": 4.7571940422058105, "epoch": 3.31, "learning_rate": 4.329054477144646e-05, "loss": 91.5513, "step": 3918, "task_loss": 1.8926148414611816 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8217541155929845, "compression/movement_sparsity/importance_threshold": -0.00115684161382033, "compression/movement_sparsity/linear_layer_sparsity": 0.8263340854144344, "compression/movement_sparsity/model_sparsity": 0.797946967479462, "compression_loss": 87.58527374267578, "distillation_loss": 6.058923721313477, "epoch": 3.31, "learning_rate": 4.32874139010645e-05, "loss": 92.7729, "step": 3919, "task_loss": 3.890106678009033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.822021484375, "compression/movement_sparsity/importance_threshold": -0.001155106351689028, "compression/movement_sparsity/linear_layer_sparsity": 0.8265405762253847, "compression/movement_sparsity/model_sparsity": 0.7981463646958161, "compression_loss": 87.61312866210938, "distillation_loss": 3.446645736694336, "epoch": 3.31, "learning_rate": 4.328428303068253e-05, "loss": 91.711, "step": 3920, "task_loss": 1.6339914798736572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8222885856545046, "compression/movement_sparsity/importance_threshold": -0.0011533728256877774, "compression/movement_sparsity/linear_layer_sparsity": 0.8266789800391342, "compression/movement_sparsity/model_sparsity": 0.7982800139127834, "compression_loss": 87.64093017578125, "distillation_loss": 4.506110191345215, "epoch": 3.31, "learning_rate": 4.328115216030056e-05, "loss": 91.9648, "step": 3921, "task_loss": 1.7449910640716553 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8225554195653835, "compression/movement_sparsity/importance_threshold": -0.001151641034947643, "compression/movement_sparsity/linear_layer_sparsity": 0.8269320704972054, "compression/movement_sparsity/model_sparsity": 0.7985244099350224, "compression_loss": 87.66871643066406, "distillation_loss": 3.914494037628174, "epoch": 3.32, "learning_rate": 4.32780212899186e-05, "loss": 91.4708, "step": 3922, "task_loss": 1.8700590133666992 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8228219862415216, "compression/movement_sparsity/importance_threshold": -0.0011499109785996922, "compression/movement_sparsity/linear_layer_sparsity": 0.8271774221704808, "compression/movement_sparsity/model_sparsity": 0.7987613330235309, "compression_loss": 87.69654083251953, "distillation_loss": 7.028143405914307, "epoch": 3.32, "learning_rate": 4.327489041953663e-05, "loss": 93.239, "step": 3923, "task_loss": 3.697472333908081 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8230882858168044, "compression/movement_sparsity/importance_threshold": -0.001148182655774991, "compression/movement_sparsity/linear_layer_sparsity": 0.8273812061953778, "compression/movement_sparsity/model_sparsity": 0.7989581164402596, "compression_loss": 87.72428131103516, "distillation_loss": 3.292959690093994, "epoch": 3.32, "learning_rate": 4.3271759549154664e-05, "loss": 91.7945, "step": 3924, "task_loss": 2.673332929611206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8233543184251164, "compression/movement_sparsity/importance_threshold": -0.0011464560656046065, "compression/movement_sparsity/linear_layer_sparsity": 0.827563204766004, "compression/movement_sparsity/model_sparsity": 0.7991338628000917, "compression_loss": 87.7520523071289, "distillation_loss": 5.022679805755615, "epoch": 3.32, "learning_rate": 4.3268628678772696e-05, "loss": 91.5458, "step": 3925, "task_loss": 2.6603200435638428 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8236200842003436, "compression/movement_sparsity/importance_threshold": -0.0011447312072196021, "compression/movement_sparsity/linear_layer_sparsity": 0.8276649656126084, "compression/movement_sparsity/model_sparsity": 0.7992321278485623, "compression_loss": 87.77980041503906, "distillation_loss": 3.9436240196228027, "epoch": 3.32, "learning_rate": 4.3265497808390734e-05, "loss": 91.7411, "step": 3926, "task_loss": 2.1330065727233887 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8238855832763703, "compression/movement_sparsity/importance_threshold": -0.0011430080797510478, "compression/movement_sparsity/linear_layer_sparsity": 0.8279279173573144, "compression/movement_sparsity/model_sparsity": 0.7994860463919035, "compression_loss": 87.8074722290039, "distillation_loss": 2.745420455932617, "epoch": 3.32, "learning_rate": 4.3262366938008766e-05, "loss": 90.9395, "step": 3927, "task_loss": 1.2260640859603882 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8241508157870824, "compression/movement_sparsity/importance_threshold": -0.0011412866823300042, "compression/movement_sparsity/linear_layer_sparsity": 0.8281837861464448, "compression/movement_sparsity/model_sparsity": 0.7997331253009827, "compression_loss": 87.83515930175781, "distillation_loss": 5.144534111022949, "epoch": 3.32, "learning_rate": 4.3259236067626805e-05, "loss": 91.8129, "step": 3928, "task_loss": 3.430816411972046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8244157818663643, "compression/movement_sparsity/importance_threshold": -0.001139567014087543, "compression/movement_sparsity/linear_layer_sparsity": 0.8284625731857713, "compression/movement_sparsity/model_sparsity": 0.8000023351478591, "compression_loss": 87.86278533935547, "distillation_loss": 4.690457344055176, "epoch": 3.32, "learning_rate": 4.3256105197244837e-05, "loss": 92.415, "step": 3929, "task_loss": 2.425628662109375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8246804816481018, "compression/movement_sparsity/importance_threshold": -0.0011378490741547264, "compression/movement_sparsity/linear_layer_sparsity": 0.8287196343916652, "compression/movement_sparsity/model_sparsity": 0.8002505655105179, "compression_loss": 87.89041900634766, "distillation_loss": 3.4347591400146484, "epoch": 3.32, "learning_rate": 4.325297432686287e-05, "loss": 91.7017, "step": 3930, "task_loss": 1.534367561340332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8249449152661795, "compression/movement_sparsity/importance_threshold": -0.001136132861662623, "compression/movement_sparsity/linear_layer_sparsity": 0.8289612299521354, "compression/movement_sparsity/model_sparsity": 0.8004838615202511, "compression_loss": 87.91796875, "distillation_loss": 3.7788782119750977, "epoch": 3.32, "learning_rate": 4.324984345648091e-05, "loss": 91.9197, "step": 3931, "task_loss": 2.6203463077545166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8252090828544829, "compression/movement_sparsity/importance_threshold": -0.0011344183757422965, "compression/movement_sparsity/linear_layer_sparsity": 0.8290971058423461, "compression/movement_sparsity/model_sparsity": 0.8006150696556298, "compression_loss": 87.94548034667969, "distillation_loss": 5.723841190338135, "epoch": 3.32, "learning_rate": 4.324671258609894e-05, "loss": 92.9557, "step": 3932, "task_loss": 1.764642357826233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8254729845468971, "compression/movement_sparsity/importance_threshold": -0.0011327056155248138, "compression/movement_sparsity/linear_layer_sparsity": 0.8293411935538522, "compression/movement_sparsity/model_sparsity": 0.8008507722033441, "compression_loss": 87.97306060791016, "distillation_loss": 4.29743766784668, "epoch": 3.32, "learning_rate": 4.324358171571697e-05, "loss": 92.1679, "step": 3933, "task_loss": 1.8665852546691895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8257366204773071, "compression/movement_sparsity/importance_threshold": -0.0011309945801412412, "compression/movement_sparsity/linear_layer_sparsity": 0.8295546480787018, "compression/movement_sparsity/model_sparsity": 0.8010568939086021, "compression_loss": 88.00049591064453, "distillation_loss": 4.0361480712890625, "epoch": 3.33, "learning_rate": 4.324045084533501e-05, "loss": 92.5446, "step": 3934, "task_loss": 2.0032567977905273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8259999907795982, "compression/movement_sparsity/importance_threshold": -0.0011292852687226447, "compression/movement_sparsity/linear_layer_sparsity": 0.8296777054887039, "compression/movement_sparsity/model_sparsity": 0.8011757239180018, "compression_loss": 88.02798461914062, "distillation_loss": 4.567870140075684, "epoch": 3.33, "learning_rate": 4.323731997495304e-05, "loss": 92.7124, "step": 3935, "task_loss": 2.7447361946105957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8262630955876553, "compression/movement_sparsity/importance_threshold": -0.0011275776804000907, "compression/movement_sparsity/linear_layer_sparsity": 0.8299584480915226, "compression/movement_sparsity/model_sparsity": 0.8014468221487485, "compression_loss": 88.0553970336914, "distillation_loss": 2.7738659381866455, "epoch": 3.33, "learning_rate": 4.323418910457107e-05, "loss": 92.084, "step": 3936, "task_loss": 2.052676200866699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8265259350353639, "compression/movement_sparsity/importance_threshold": -0.0011258718143046442, "compression/movement_sparsity/linear_layer_sparsity": 0.8301870820817727, "compression/movement_sparsity/model_sparsity": 0.801667601858073, "compression_loss": 88.08280944824219, "distillation_loss": 2.3358583450317383, "epoch": 3.33, "learning_rate": 4.323105823418911e-05, "loss": 92.1462, "step": 3937, "task_loss": 1.480911374092102 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8267885092566086, "compression/movement_sparsity/importance_threshold": -0.0011241676695673735, "compression/movement_sparsity/linear_layer_sparsity": 0.830470328759795, "compression/movement_sparsity/model_sparsity": 0.8019411181413365, "compression_loss": 88.11016845703125, "distillation_loss": 5.002298355102539, "epoch": 3.33, "learning_rate": 4.322792736380714e-05, "loss": 92.5312, "step": 3938, "task_loss": 3.2623841762542725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8270508183852753, "compression/movement_sparsity/importance_threshold": -0.0011224652453193408, "compression/movement_sparsity/linear_layer_sparsity": 0.8308985256195989, "compression/movement_sparsity/model_sparsity": 0.8023546051217185, "compression_loss": 88.13748168945312, "distillation_loss": 5.0642595291137695, "epoch": 3.33, "learning_rate": 4.3224796493425175e-05, "loss": 92.5593, "step": 3939, "task_loss": 2.407445192337036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8273128625552484, "compression/movement_sparsity/importance_threshold": -0.0011207645406916163, "compression/movement_sparsity/linear_layer_sparsity": 0.8310591203093186, "compression/movement_sparsity/model_sparsity": 0.8025096828897994, "compression_loss": 88.16487884521484, "distillation_loss": 3.0572566986083984, "epoch": 3.33, "learning_rate": 4.3221665623043207e-05, "loss": 92.0596, "step": 3940, "task_loss": 1.8451941013336182 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8275746419004133, "compression/movement_sparsity/importance_threshold": -0.0011190655548152639, "compression/movement_sparsity/linear_layer_sparsity": 0.8312685802380102, "compression/movement_sparsity/model_sparsity": 0.8027119472255663, "compression_loss": 88.1921157836914, "distillation_loss": 3.4733290672302246, "epoch": 3.33, "learning_rate": 4.3218534752661245e-05, "loss": 92.4622, "step": 3941, "task_loss": 2.242257833480835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8278361565546555, "compression/movement_sparsity/importance_threshold": -0.0011173682868213484, "compression/movement_sparsity/linear_layer_sparsity": 0.8313825872047766, "compression/movement_sparsity/model_sparsity": 0.8028220377022979, "compression_loss": 88.21944427490234, "distillation_loss": 4.3063178062438965, "epoch": 3.33, "learning_rate": 4.321540388227928e-05, "loss": 91.6087, "step": 3942, "task_loss": 2.60442852973938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8280974066518596, "compression/movement_sparsity/importance_threshold": -0.0011156727358409392, "compression/movement_sparsity/linear_layer_sparsity": 0.8314721019312191, "compression/movement_sparsity/model_sparsity": 0.8029084773225075, "compression_loss": 88.24663543701172, "distillation_loss": 3.0103609561920166, "epoch": 3.33, "learning_rate": 4.321227301189731e-05, "loss": 91.8884, "step": 3943, "task_loss": 1.3890726566314697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8283583923259111, "compression/movement_sparsity/importance_threshold": -0.0011139789010050985, "compression/movement_sparsity/linear_layer_sparsity": 0.8316877028062432, "compression/movement_sparsity/model_sparsity": 0.8031166716442086, "compression_loss": 88.27381896972656, "distillation_loss": 3.3015501499176025, "epoch": 3.33, "learning_rate": 4.320914214151534e-05, "loss": 91.5106, "step": 3944, "task_loss": 2.8351337909698486 }, { "compression/movement_sparsity/importance_regularization_factor": 0.828619113710695, "compression/movement_sparsity/importance_threshold": -0.0011122867814448954, "compression/movement_sparsity/linear_layer_sparsity": 0.8319494621341856, "compression/movement_sparsity/model_sparsity": 0.8033694387339704, "compression_loss": 88.30106353759766, "distillation_loss": 4.229666233062744, "epoch": 3.33, "learning_rate": 4.320601127113338e-05, "loss": 92.8089, "step": 3945, "task_loss": 2.118103265762329 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8288795709400965, "compression/movement_sparsity/importance_threshold": -0.001110596376291394, "compression/movement_sparsity/linear_layer_sparsity": 0.8320687515072148, "compression/movement_sparsity/model_sparsity": 0.8034846301500589, "compression_loss": 88.32813262939453, "distillation_loss": 3.309422254562378, "epoch": 3.34, "learning_rate": 4.320288040075141e-05, "loss": 91.6194, "step": 3946, "task_loss": 1.7687904834747314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8291397641480005, "compression/movement_sparsity/importance_threshold": -0.0011089076846756626, "compression/movement_sparsity/linear_layer_sparsity": 0.8322666807658025, "compression/movement_sparsity/model_sparsity": 0.8036757599297125, "compression_loss": 88.3552474975586, "distillation_loss": 5.751396656036377, "epoch": 3.34, "learning_rate": 4.319974953036944e-05, "loss": 93.3447, "step": 3947, "task_loss": 2.957625150680542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8293996934682926, "compression/movement_sparsity/importance_threshold": -0.0011072207057287644, "compression/movement_sparsity/linear_layer_sparsity": 0.8323829414002522, "compression/movement_sparsity/model_sparsity": 0.8037880266537093, "compression_loss": 88.38230895996094, "distillation_loss": 2.9780044555664062, "epoch": 3.34, "learning_rate": 4.319661865998748e-05, "loss": 92.8969, "step": 3948, "task_loss": 1.6851153373718262 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8296593590348574, "compression/movement_sparsity/importance_threshold": -0.0011055354385817676, "compression/movement_sparsity/linear_layer_sparsity": 0.8326054105958345, "compression/movement_sparsity/model_sparsity": 0.804002853348028, "compression_loss": 88.4093246459961, "distillation_loss": 4.969673156738281, "epoch": 3.34, "learning_rate": 4.319348778960551e-05, "loss": 93.0049, "step": 3949, "task_loss": 2.4572839736938477 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8299187609815805, "compression/movement_sparsity/importance_threshold": -0.001103851882365738, "compression/movement_sparsity/linear_layer_sparsity": 0.8328413541008453, "compression/movement_sparsity/model_sparsity": 0.8042306914677946, "compression_loss": 88.43630981445312, "distillation_loss": 4.390437126159668, "epoch": 3.34, "learning_rate": 4.3190356919223545e-05, "loss": 92.2123, "step": 3950, "task_loss": 1.5706473588943481 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8301778994423469, "compression/movement_sparsity/importance_threshold": -0.0011021700362117397, "compression/movement_sparsity/linear_layer_sparsity": 0.8330611642070448, "compression/movement_sparsity/model_sparsity": 0.8044429504206312, "compression_loss": 88.46329498291016, "distillation_loss": 3.570308208465576, "epoch": 3.34, "learning_rate": 4.3187226048841576e-05, "loss": 91.374, "step": 3951, "task_loss": 1.5285487174987793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8304367745510416, "compression/movement_sparsity/importance_threshold": -0.001100489899250841, "compression/movement_sparsity/linear_layer_sparsity": 0.833075532829046, "compression/movement_sparsity/model_sparsity": 0.8044568254362636, "compression_loss": 88.49021911621094, "distillation_loss": 5.756222724914551, "epoch": 3.34, "learning_rate": 4.3184095178459615e-05, "loss": 92.7654, "step": 3952, "task_loss": 2.798248529434204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8306953864415499, "compression/movement_sparsity/importance_threshold": -0.0010988114706141066, "compression/movement_sparsity/linear_layer_sparsity": 0.8332317155767405, "compression/movement_sparsity/model_sparsity": 0.8046076428261005, "compression_loss": 88.51709747314453, "distillation_loss": 4.198023796081543, "epoch": 3.34, "learning_rate": 4.318096430807765e-05, "loss": 92.179, "step": 3953, "task_loss": 1.9972014427185059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8309537352477568, "compression/movement_sparsity/importance_threshold": -0.0010971347494326041, "compression/movement_sparsity/linear_layer_sparsity": 0.8334061423009179, "compression/movement_sparsity/model_sparsity": 0.804776077455703, "compression_loss": 88.54397583007812, "distillation_loss": 4.30730676651001, "epoch": 3.34, "learning_rate": 4.317783343769568e-05, "loss": 92.246, "step": 3954, "task_loss": 1.476380467414856 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8312118211035477, "compression/movement_sparsity/importance_threshold": -0.0010954597348373967, "compression/movement_sparsity/linear_layer_sparsity": 0.8334996754718537, "compression/movement_sparsity/model_sparsity": 0.8048663974744754, "compression_loss": 88.57086181640625, "distillation_loss": 4.007308006286621, "epoch": 3.34, "learning_rate": 4.317470256731371e-05, "loss": 92.5318, "step": 3955, "task_loss": 1.9288029670715332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8314696441428073, "compression/movement_sparsity/importance_threshold": -0.0010937864259595543, "compression/movement_sparsity/linear_layer_sparsity": 0.8337766261893642, "compression/movement_sparsity/model_sparsity": 0.8051338340828395, "compression_loss": 88.59764099121094, "distillation_loss": 3.8711419105529785, "epoch": 3.34, "learning_rate": 4.317157169693175e-05, "loss": 92.638, "step": 3956, "task_loss": 2.259958505630493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8317272044994214, "compression/movement_sparsity/importance_threshold": -0.0010921148219301385, "compression/movement_sparsity/linear_layer_sparsity": 0.833945997066464, "compression/movement_sparsity/model_sparsity": 0.8052973865492651, "compression_loss": 88.62450408935547, "distillation_loss": 3.539682626724243, "epoch": 3.34, "learning_rate": 4.316844082654978e-05, "loss": 91.8669, "step": 3957, "task_loss": 1.7548984289169312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8319845023072744, "compression/movement_sparsity/importance_threshold": -0.00109044492188022, "compression/movement_sparsity/linear_layer_sparsity": 0.8341295100063799, "compression/movement_sparsity/model_sparsity": 0.8054745952551431, "compression_loss": 88.65127563476562, "distillation_loss": 4.405172348022461, "epoch": 3.35, "learning_rate": 4.316530995616781e-05, "loss": 93.1835, "step": 3958, "task_loss": 1.9390153884887695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.832241537700252, "compression/movement_sparsity/importance_threshold": -0.0010887767249408604, "compression/movement_sparsity/linear_layer_sparsity": 0.8343749332246613, "compression/movement_sparsity/model_sparsity": 0.8057115874308663, "compression_loss": 88.67803955078125, "distillation_loss": 4.411349296569824, "epoch": 3.35, "learning_rate": 4.316217908578585e-05, "loss": 93.5226, "step": 3959, "task_loss": 2.2001800537109375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8324983108122389, "compression/movement_sparsity/importance_threshold": -0.0010871102302431298, "compression/movement_sparsity/linear_layer_sparsity": 0.8344598332982286, "compression/movement_sparsity/model_sparsity": 0.8057935709257235, "compression_loss": 88.70478057861328, "distillation_loss": 3.9497475624084473, "epoch": 3.35, "learning_rate": 4.315904821540388e-05, "loss": 93.0102, "step": 3960, "task_loss": 2.1542162895202637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8327548217771207, "compression/movement_sparsity/importance_threshold": -0.001085445436918091, "compression/movement_sparsity/linear_layer_sparsity": 0.8346575717701341, "compression/movement_sparsity/model_sparsity": 0.8059845164728043, "compression_loss": 88.73152923583984, "distillation_loss": 4.730204105377197, "epoch": 3.35, "learning_rate": 4.3155917345021915e-05, "loss": 92.9902, "step": 3961, "task_loss": 2.658562421798706 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8330110707287821, "compression/movement_sparsity/importance_threshold": -0.001083782344096812, "compression/movement_sparsity/linear_layer_sparsity": 0.8348086867465834, "compression/movement_sparsity/model_sparsity": 0.8061304401849285, "compression_loss": 88.75819396972656, "distillation_loss": 3.174647092819214, "epoch": 3.35, "learning_rate": 4.315278647463995e-05, "loss": 92.7982, "step": 3962, "task_loss": 2.57258939743042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8332670578011085, "compression/movement_sparsity/importance_threshold": -0.001082120950910359, "compression/movement_sparsity/linear_layer_sparsity": 0.8350681566343398, "compression/movement_sparsity/model_sparsity": 0.8063809964838177, "compression_loss": 88.78478240966797, "distillation_loss": 4.053418159484863, "epoch": 3.35, "learning_rate": 4.3149655604257985e-05, "loss": 92.9112, "step": 3963, "task_loss": 1.9943886995315552 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8335227831279851, "compression/movement_sparsity/importance_threshold": -0.001080461256489796, "compression/movement_sparsity/linear_layer_sparsity": 0.8352793336431709, "compression/movement_sparsity/model_sparsity": 0.8065849189127391, "compression_loss": 88.81146240234375, "distillation_loss": 3.4394378662109375, "epoch": 3.35, "learning_rate": 4.3146524733876023e-05, "loss": 93.5025, "step": 3964, "task_loss": 2.5788915157318115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8337782468432968, "compression/movement_sparsity/importance_threshold": -0.0010788032599661906, "compression/movement_sparsity/linear_layer_sparsity": 0.8353816787739895, "compression/movement_sparsity/model_sparsity": 0.8066837481734636, "compression_loss": 88.83807373046875, "distillation_loss": 5.031275272369385, "epoch": 3.35, "learning_rate": 4.3143393863494055e-05, "loss": 93.1384, "step": 3965, "task_loss": 2.744259834289551 }, { "compression/movement_sparsity/importance_regularization_factor": 0.834033449080929, "compression/movement_sparsity/importance_threshold": -0.0010771469604706077, "compression/movement_sparsity/linear_layer_sparsity": 0.8355992590608411, "compression/movement_sparsity/model_sparsity": 0.8068938539081065, "compression_loss": 88.86470794677734, "distillation_loss": 3.9868226051330566, "epoch": 3.35, "learning_rate": 4.314026299311209e-05, "loss": 93.0254, "step": 3966, "task_loss": 2.4922752380371094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8342883899747666, "compression/movement_sparsity/importance_threshold": -0.0010754923571341156, "compression/movement_sparsity/linear_layer_sparsity": 0.8357160443586668, "compression/movement_sparsity/model_sparsity": 0.8070066272716783, "compression_loss": 88.8912582397461, "distillation_loss": 4.0825090408325195, "epoch": 3.35, "learning_rate": 4.3137132122730126e-05, "loss": 93.0469, "step": 3967, "task_loss": 2.4417834281921387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8345430696586951, "compression/movement_sparsity/importance_threshold": -0.0010738394490877774, "compression/movement_sparsity/linear_layer_sparsity": 0.8359916834177373, "compression/movement_sparsity/model_sparsity": 0.8072727972811049, "compression_loss": 88.91777801513672, "distillation_loss": 5.227862358093262, "epoch": 3.35, "learning_rate": 4.313400125234816e-05, "loss": 92.8923, "step": 3968, "task_loss": 2.9926702976226807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.834797488266599, "compression/movement_sparsity/importance_threshold": -0.0010721882354626623, "compression/movement_sparsity/linear_layer_sparsity": 0.8361429057116954, "compression/movement_sparsity/model_sparsity": 0.8074188246240512, "compression_loss": 88.94424438476562, "distillation_loss": 5.6753315925598145, "epoch": 3.35, "learning_rate": 4.313087038196619e-05, "loss": 93.4321, "step": 3969, "task_loss": 3.862779140472412 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8350516459323641, "compression/movement_sparsity/importance_threshold": -0.0010705387153898335, "compression/movement_sparsity/linear_layer_sparsity": 0.8364210846184724, "compression/movement_sparsity/model_sparsity": 0.8076874472296021, "compression_loss": 88.97073364257812, "distillation_loss": 4.12469482421875, "epoch": 3.36, "learning_rate": 4.312773951158422e-05, "loss": 92.9891, "step": 3970, "task_loss": 1.0273054838180542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8353055427898752, "compression/movement_sparsity/importance_threshold": -0.0010688908880003584, "compression/movement_sparsity/linear_layer_sparsity": 0.8365658320894042, "compression/movement_sparsity/model_sparsity": 0.807827222179612, "compression_loss": 88.99726867675781, "distillation_loss": 4.206568717956543, "epoch": 3.36, "learning_rate": 4.312460864120226e-05, "loss": 92.9007, "step": 3971, "task_loss": 1.7135863304138184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8355591789730175, "compression/movement_sparsity/importance_threshold": -0.0010672447524253037, "compression/movement_sparsity/linear_layer_sparsity": 0.8367113546312323, "compression/movement_sparsity/model_sparsity": 0.8079677455744486, "compression_loss": 89.02364349365234, "distillation_loss": 5.141359329223633, "epoch": 3.36, "learning_rate": 4.312147777082029e-05, "loss": 93.472, "step": 3972, "task_loss": 3.349090576171875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8358125546156762, "compression/movement_sparsity/importance_threshold": -0.001065600307795734, "compression/movement_sparsity/linear_layer_sparsity": 0.8368357236996743, "compression/movement_sparsity/model_sparsity": 0.8080878421827857, "compression_loss": 89.0500717163086, "distillation_loss": 4.97414493560791, "epoch": 3.36, "learning_rate": 4.311834690043832e-05, "loss": 93.3274, "step": 3973, "task_loss": 2.894482374191284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8360656698517362, "compression/movement_sparsity/importance_threshold": -0.0010639575532427163, "compression/movement_sparsity/linear_layer_sparsity": 0.8369733405183598, "compression/movement_sparsity/model_sparsity": 0.8082207314403905, "compression_loss": 89.07647705078125, "distillation_loss": 3.761120319366455, "epoch": 3.36, "learning_rate": 4.311521603005636e-05, "loss": 92.894, "step": 3974, "task_loss": 2.083743095397949 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8363185248150832, "compression/movement_sparsity/importance_threshold": -0.0010623164878973152, "compression/movement_sparsity/linear_layer_sparsity": 0.8370273450735826, "compression/movement_sparsity/model_sparsity": 0.808272880773004, "compression_loss": 89.10283660888672, "distillation_loss": 3.084249973297119, "epoch": 3.36, "learning_rate": 4.3112085159674393e-05, "loss": 93.0585, "step": 3975, "task_loss": 2.527886390686035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8365711196396016, "compression/movement_sparsity/importance_threshold": -0.001060677110890599, "compression/movement_sparsity/linear_layer_sparsity": 0.8374076067794903, "compression/movement_sparsity/model_sparsity": 0.8086400793194919, "compression_loss": 89.12918853759766, "distillation_loss": 4.271529197692871, "epoch": 3.36, "learning_rate": 4.3108954289292425e-05, "loss": 92.967, "step": 3976, "task_loss": 2.570305109024048 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8368234544591773, "compression/movement_sparsity/importance_threshold": -0.001059039421353632, "compression/movement_sparsity/linear_layer_sparsity": 0.837591847093632, "compression/movement_sparsity/model_sparsity": 0.8088179904120533, "compression_loss": 89.15544891357422, "distillation_loss": 3.089662551879883, "epoch": 3.36, "learning_rate": 4.310582341891046e-05, "loss": 92.6679, "step": 3977, "task_loss": 1.6243504285812378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8370755294076948, "compression/movement_sparsity/importance_threshold": -0.0010574034184174805, "compression/movement_sparsity/linear_layer_sparsity": 0.8378190501837657, "compression/movement_sparsity/model_sparsity": 0.8090373883770825, "compression_loss": 89.1817398071289, "distillation_loss": 4.379918098449707, "epoch": 3.36, "learning_rate": 4.3102692548528496e-05, "loss": 93.8865, "step": 3978, "task_loss": 2.8854451179504395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8373273446190396, "compression/movement_sparsity/importance_threshold": -0.0010557691012132112, "compression/movement_sparsity/linear_layer_sparsity": 0.8379532089938369, "compression/movement_sparsity/model_sparsity": 0.8091669384193069, "compression_loss": 89.20797729492188, "distillation_loss": 4.544582843780518, "epoch": 3.36, "learning_rate": 4.309956167814653e-05, "loss": 93.2592, "step": 3979, "task_loss": 2.216291666030884 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8375789002270966, "compression/movement_sparsity/importance_threshold": -0.0010541364688718908, "compression/movement_sparsity/linear_layer_sparsity": 0.8380486500315942, "compression/movement_sparsity/model_sparsity": 0.8092591007638064, "compression_loss": 89.23428344726562, "distillation_loss": 4.217299938201904, "epoch": 3.36, "learning_rate": 4.309643080776456e-05, "loss": 92.8231, "step": 3980, "task_loss": 1.9187227487564087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8378301963657511, "compression/movement_sparsity/importance_threshold": -0.001052505520524584, "compression/movement_sparsity/linear_layer_sparsity": 0.8382560232309496, "compression/movement_sparsity/model_sparsity": 0.8094593500558093, "compression_loss": 89.26040649414062, "distillation_loss": 5.499324798583984, "epoch": 3.36, "learning_rate": 4.309329993738259e-05, "loss": 93.9487, "step": 3981, "task_loss": 1.9122074842453003 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8380812331688884, "compression/movement_sparsity/importance_threshold": -0.0010508762553023557, "compression/movement_sparsity/linear_layer_sparsity": 0.8384370797923325, "compression/movement_sparsity/model_sparsity": 0.8096341867673136, "compression_loss": 89.28656005859375, "distillation_loss": 6.267455101013184, "epoch": 3.37, "learning_rate": 4.309016906700063e-05, "loss": 93.4223, "step": 3982, "task_loss": 3.247669219970703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8383320107703933, "compression/movement_sparsity/importance_threshold": -0.0010492486723362743, "compression/movement_sparsity/linear_layer_sparsity": 0.8385792397188873, "compression/movement_sparsity/model_sparsity": 0.809771463063056, "compression_loss": 89.31273651123047, "distillation_loss": 4.1730055809021, "epoch": 3.37, "learning_rate": 4.308703819661866e-05, "loss": 93.6238, "step": 3983, "task_loss": 1.3184705972671509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8385825293041512, "compression/movement_sparsity/importance_threshold": -0.0010476227707574047, "compression/movement_sparsity/linear_layer_sparsity": 0.8387088077244186, "compression/movement_sparsity/model_sparsity": 0.8098965800089996, "compression_loss": 89.33878326416016, "distillation_loss": 4.503450870513916, "epoch": 3.37, "learning_rate": 4.308390732623669e-05, "loss": 92.9409, "step": 3984, "task_loss": 3.323394775390625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.838832788904047, "compression/movement_sparsity/importance_threshold": -0.0010459985496968135, "compression/movement_sparsity/linear_layer_sparsity": 0.8388548310812873, "compression/movement_sparsity/model_sparsity": 0.8100375870143395, "compression_loss": 89.36480712890625, "distillation_loss": 3.7511191368103027, "epoch": 3.37, "learning_rate": 4.308077645585473e-05, "loss": 93.2051, "step": 3985, "task_loss": 2.819544553756714 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8390827897039662, "compression/movement_sparsity/importance_threshold": -0.0010443760082855654, "compression/movement_sparsity/linear_layer_sparsity": 0.839091501960524, "compression/movement_sparsity/model_sparsity": 0.8102661275207895, "compression_loss": 89.39079284667969, "distillation_loss": 4.953083038330078, "epoch": 3.37, "learning_rate": 4.307764558547276e-05, "loss": 93.7165, "step": 3986, "task_loss": 3.5566282272338867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8393325318377935, "compression/movement_sparsity/importance_threshold": -0.001042755145654729, "compression/movement_sparsity/linear_layer_sparsity": 0.8392046980838912, "compression/movement_sparsity/model_sparsity": 0.8103754350090872, "compression_loss": 89.41676330566406, "distillation_loss": 4.384110450744629, "epoch": 3.37, "learning_rate": 4.3074514715090795e-05, "loss": 94.1783, "step": 3987, "task_loss": 2.15273380279541 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8395820154394146, "compression/movement_sparsity/importance_threshold": -0.0010411359609353672, "compression/movement_sparsity/linear_layer_sparsity": 0.8392807265767375, "compression/movement_sparsity/model_sparsity": 0.8104488516893131, "compression_loss": 89.44276428222656, "distillation_loss": 4.753912925720215, "epoch": 3.37, "learning_rate": 4.307138384470883e-05, "loss": 93.0982, "step": 3988, "task_loss": 1.9540103673934937 }, { "compression/movement_sparsity/importance_regularization_factor": 0.839831240642714, "compression/movement_sparsity/importance_threshold": -0.0010395184532585493, "compression/movement_sparsity/linear_layer_sparsity": 0.839301665415106, "compression/movement_sparsity/model_sparsity": 0.8104690712141683, "compression_loss": 89.46868896484375, "distillation_loss": 4.17183256149292, "epoch": 3.37, "learning_rate": 4.3068252974326865e-05, "loss": 93.5264, "step": 3989, "task_loss": 2.928349018096924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8400802075815771, "compression/movement_sparsity/importance_threshold": -0.0010379026217553393, "compression/movement_sparsity/linear_layer_sparsity": 0.839417556400359, "compression/movement_sparsity/model_sparsity": 0.8105809809875555, "compression_loss": 89.49464416503906, "distillation_loss": 3.3957772254943848, "epoch": 3.37, "learning_rate": 4.30651221039449e-05, "loss": 93.5864, "step": 3990, "task_loss": 1.577222228050232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8403289163898893, "compression/movement_sparsity/importance_threshold": -0.001036288465556803, "compression/movement_sparsity/linear_layer_sparsity": 0.8395357725983009, "compression/movement_sparsity/model_sparsity": 0.8106951360954225, "compression_loss": 89.5205078125, "distillation_loss": 5.7244439125061035, "epoch": 3.37, "learning_rate": 4.306199123356293e-05, "loss": 93.8281, "step": 3991, "task_loss": 2.531587839126587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8405773672015353, "compression/movement_sparsity/importance_threshold": -0.0010346759837940086, "compression/movement_sparsity/linear_layer_sparsity": 0.8396534879812022, "compression/movement_sparsity/model_sparsity": 0.8108088075927863, "compression_loss": 89.54634857177734, "distillation_loss": 2.6583986282348633, "epoch": 3.37, "learning_rate": 4.305886036318096e-05, "loss": 93.5386, "step": 3992, "task_loss": 1.6876633167266846 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8408255601504008, "compression/movement_sparsity/importance_threshold": -0.0010330651755980176, "compression/movement_sparsity/linear_layer_sparsity": 0.8397786559688758, "compression/movement_sparsity/model_sparsity": 0.8109296756750215, "compression_loss": 89.5721435546875, "distillation_loss": 5.816449165344238, "epoch": 3.38, "learning_rate": 4.3055729492799e-05, "loss": 93.5349, "step": 3993, "task_loss": 3.170114517211914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8410734953703702, "compression/movement_sparsity/importance_threshold": -0.0010314560400999026, "compression/movement_sparsity/linear_layer_sparsity": 0.8399747489056476, "compression/movement_sparsity/model_sparsity": 0.8111190322161628, "compression_loss": 89.59793090820312, "distillation_loss": 4.5987443923950195, "epoch": 3.38, "learning_rate": 4.305259862241703e-05, "loss": 93.5671, "step": 3994, "task_loss": 1.6401904821395874 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8413211729953295, "compression/movement_sparsity/importance_threshold": -0.0010298485764307225, "compression/movement_sparsity/linear_layer_sparsity": 0.8402276962737072, "compression/movement_sparsity/model_sparsity": 0.8113632900639723, "compression_loss": 89.62367248535156, "distillation_loss": 4.124173164367676, "epoch": 3.38, "learning_rate": 4.304946775203507e-05, "loss": 94.1639, "step": 3995, "task_loss": 1.9597049951553345 }, { "compression/movement_sparsity/importance_regularization_factor": 0.841568593159163, "compression/movement_sparsity/importance_threshold": -0.0010282427837215499, "compression/movement_sparsity/linear_layer_sparsity": 0.840392380952926, "compression/movement_sparsity/model_sparsity": 0.8115223173178308, "compression_loss": 89.64933776855469, "distillation_loss": 3.400998830795288, "epoch": 3.38, "learning_rate": 4.30463368816531e-05, "loss": 93.6989, "step": 3996, "task_loss": 1.9402583837509155 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8418157559957565, "compression/movement_sparsity/importance_threshold": -0.0010266386611034453, "compression/movement_sparsity/linear_layer_sparsity": 0.8406228631891597, "compression/movement_sparsity/model_sparsity": 0.8117448817802034, "compression_loss": 89.67512512207031, "distillation_loss": 4.525179862976074, "epoch": 3.38, "learning_rate": 4.304320601127113e-05, "loss": 93.5663, "step": 3997, "task_loss": 3.0269675254821777 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8420626616389947, "compression/movement_sparsity/importance_threshold": -0.0010250362077074797, "compression/movement_sparsity/linear_layer_sparsity": 0.8407626502063549, "compression/movement_sparsity/model_sparsity": 0.8118798666833228, "compression_loss": 89.70074462890625, "distillation_loss": 3.6089956760406494, "epoch": 3.38, "learning_rate": 4.304007514088917e-05, "loss": 93.3839, "step": 3998, "task_loss": 2.1528193950653076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.842309310222763, "compression/movement_sparsity/importance_threshold": -0.0010234354226647153, "compression/movement_sparsity/linear_layer_sparsity": 0.8408644349012946, "compression/movement_sparsity/model_sparsity": 0.811978154760865, "compression_loss": 89.72640991210938, "distillation_loss": 4.523263931274414, "epoch": 3.38, "learning_rate": 4.3036944270507204e-05, "loss": 92.8959, "step": 3999, "task_loss": 1.73624849319458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8425557018809464, "compression/movement_sparsity/importance_threshold": -0.0010218363051062206, "compression/movement_sparsity/linear_layer_sparsity": 0.8409990468297358, "compression/movement_sparsity/model_sparsity": 0.8121081423554496, "compression_loss": 89.75201416015625, "distillation_loss": 4.728519439697266, "epoch": 3.38, "learning_rate": 4.303381340012524e-05, "loss": 93.6153, "step": 4000, "task_loss": 2.935675859451294 }, { "epoch": 3.38, "eval_accuracy": 0.5928712871287128, "eval_loss": 93.30909729003906, "eval_runtime": 210.3198, "eval_samples_per_second": 120.055, "eval_steps_per_second": 0.941, "step": 4000 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8428018367474303, "compression/movement_sparsity/importance_threshold": -0.0010202388541630586, "compression/movement_sparsity/linear_layer_sparsity": 0.8411793640927254, "compression/movement_sparsity/model_sparsity": 0.8122822651657347, "compression_loss": 89.77755737304688, "distillation_loss": 4.986965179443359, "epoch": 3.38, "learning_rate": 4.3030682529743274e-05, "loss": 93.3432, "step": 4001, "task_loss": 2.253070592880249 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8430477149560993, "compression/movement_sparsity/importance_threshold": -0.0010186430689662993, "compression/movement_sparsity/linear_layer_sparsity": 0.8414219493591094, "compression/movement_sparsity/model_sparsity": 0.8125165168819388, "compression_loss": 89.80308532714844, "distillation_loss": 4.920845031738281, "epoch": 3.38, "learning_rate": 4.3027551659361306e-05, "loss": 93.635, "step": 4002, "task_loss": 3.5277578830718994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8432933366408389, "compression/movement_sparsity/importance_threshold": -0.0010170489486470077, "compression/movement_sparsity/linear_layer_sparsity": 0.841727255747258, "compression/movement_sparsity/model_sparsity": 0.8128113350564222, "compression_loss": 89.82856750488281, "distillation_loss": 3.9091882705688477, "epoch": 3.38, "learning_rate": 4.302442078897934e-05, "loss": 93.6534, "step": 4003, "task_loss": 2.8470184803009033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8435387019355345, "compression/movement_sparsity/importance_threshold": -0.001015456492336247, "compression/movement_sparsity/linear_layer_sparsity": 0.8418631554858039, "compression/movement_sparsity/model_sparsity": 0.8129425662208726, "compression_loss": 89.85409545898438, "distillation_loss": 4.672083377838135, "epoch": 3.38, "learning_rate": 4.3021289918597376e-05, "loss": 93.9763, "step": 4004, "task_loss": 2.1580212116241455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8437838109740706, "compression/movement_sparsity/importance_threshold": -0.0010138656991650861, "compression/movement_sparsity/linear_layer_sparsity": 0.8420543356655097, "compression/movement_sparsity/model_sparsity": 0.8131271787732665, "compression_loss": 89.8795166015625, "distillation_loss": 4.222705841064453, "epoch": 3.39, "learning_rate": 4.301815904821541e-05, "loss": 93.7158, "step": 4005, "task_loss": 2.491119623184204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.844028663890333, "compression/movement_sparsity/importance_threshold": -0.0010122765682645894, "compression/movement_sparsity/linear_layer_sparsity": 0.8422475429537136, "compression/movement_sparsity/model_sparsity": 0.8133137487967455, "compression_loss": 89.90495300292969, "distillation_loss": 4.957218170166016, "epoch": 3.39, "learning_rate": 4.301502817783344e-05, "loss": 93.4513, "step": 4006, "task_loss": 1.9014261960983276 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8442732608182064, "compression/movement_sparsity/importance_threshold": -0.001010689098765824, "compression/movement_sparsity/linear_layer_sparsity": 0.842403367976379, "compression/movement_sparsity/model_sparsity": 0.8134642207505085, "compression_loss": 89.93043518066406, "distillation_loss": 3.229279041290283, "epoch": 3.39, "learning_rate": 4.301189730745147e-05, "loss": 93.7757, "step": 4007, "task_loss": 1.3677599430084229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8445176018915762, "compression/movement_sparsity/importance_threshold": -0.001009103289799855, "compression/movement_sparsity/linear_layer_sparsity": 0.8426212225190863, "compression/movement_sparsity/model_sparsity": 0.8136745913194747, "compression_loss": 89.9557876586914, "distillation_loss": 3.7227132320404053, "epoch": 3.39, "learning_rate": 4.300876643706951e-05, "loss": 94.607, "step": 4008, "task_loss": 2.137270212173462 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8447616872443272, "compression/movement_sparsity/importance_threshold": -0.0010075191404977507, "compression/movement_sparsity/linear_layer_sparsity": 0.8427353964241997, "compression/movement_sparsity/model_sparsity": 0.8137848429997074, "compression_loss": 89.98114776611328, "distillation_loss": 4.192074298858643, "epoch": 3.39, "learning_rate": 4.300563556668754e-05, "loss": 93.7879, "step": 4009, "task_loss": 2.215996742248535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.845005517010345, "compression/movement_sparsity/importance_threshold": -0.0010059366499905742, "compression/movement_sparsity/linear_layer_sparsity": 0.842839947526031, "compression/movement_sparsity/model_sparsity": 0.813885802449554, "compression_loss": 90.00648498535156, "distillation_loss": 3.841829299926758, "epoch": 3.39, "learning_rate": 4.3002504696305574e-05, "loss": 93.8236, "step": 4010, "task_loss": 1.8847570419311523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8452490913235142, "compression/movement_sparsity/importance_threshold": -0.0010043558174093938, "compression/movement_sparsity/linear_layer_sparsity": 0.842946501888025, "compression/movement_sparsity/model_sparsity": 0.813988696341414, "compression_loss": 90.03179168701172, "distillation_loss": 4.249720096588135, "epoch": 3.39, "learning_rate": 4.299937382592361e-05, "loss": 94.502, "step": 4011, "task_loss": 3.8541810512542725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8454924103177204, "compression/movement_sparsity/importance_threshold": -0.0010027766418852745, "compression/movement_sparsity/linear_layer_sparsity": 0.8431644518240733, "compression/movement_sparsity/model_sparsity": 0.8141991590266666, "compression_loss": 90.05709075927734, "distillation_loss": 2.818401336669922, "epoch": 3.39, "learning_rate": 4.2996242955541644e-05, "loss": 93.7466, "step": 4012, "task_loss": 1.7083042860031128 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8457354741268487, "compression/movement_sparsity/importance_threshold": -0.001001199122549281, "compression/movement_sparsity/linear_layer_sparsity": 0.8433225782110925, "compression/movement_sparsity/model_sparsity": 0.814351853285838, "compression_loss": 90.08236694335938, "distillation_loss": 5.712060928344727, "epoch": 3.39, "learning_rate": 4.2993112085159676e-05, "loss": 93.7728, "step": 4013, "task_loss": 3.1258952617645264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8459782828847839, "compression/movement_sparsity/importance_threshold": -0.0009996232585324817, "compression/movement_sparsity/linear_layer_sparsity": 0.8434793213946659, "compression/movement_sparsity/model_sparsity": 0.8145032118588572, "compression_loss": 90.10765075683594, "distillation_loss": 3.721263885498047, "epoch": 3.39, "learning_rate": 4.298998121477771e-05, "loss": 94.4996, "step": 4014, "task_loss": 2.487410545349121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8462208367254116, "compression/movement_sparsity/importance_threshold": -0.0009980490489659415, "compression/movement_sparsity/linear_layer_sparsity": 0.8436448765381223, "compression/movement_sparsity/model_sparsity": 0.8146630796738287, "compression_loss": 90.13292694091797, "distillation_loss": 2.765695095062256, "epoch": 3.39, "learning_rate": 4.2986850344395746e-05, "loss": 93.7939, "step": 4015, "task_loss": 2.0718936920166016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8464631357826164, "compression/movement_sparsity/importance_threshold": -0.000996476492980727, "compression/movement_sparsity/linear_layer_sparsity": 0.8437606363575312, "compression/movement_sparsity/model_sparsity": 0.814774862787322, "compression_loss": 90.15812683105469, "distillation_loss": 6.039024353027344, "epoch": 3.39, "learning_rate": 4.298371947401378e-05, "loss": 95.1083, "step": 4016, "task_loss": 3.6859703063964844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8467051801902841, "compression/movement_sparsity/importance_threshold": -0.000994905589707903, "compression/movement_sparsity/linear_layer_sparsity": 0.8441341728328923, "compression/movement_sparsity/model_sparsity": 0.8151355671356219, "compression_loss": 90.1833267211914, "distillation_loss": 3.7458527088165283, "epoch": 3.4, "learning_rate": 4.298058860363181e-05, "loss": 94.8751, "step": 4017, "task_loss": 2.8610918521881104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8469469700822992, "compression/movement_sparsity/importance_threshold": -0.0009933363382785372, "compression/movement_sparsity/linear_layer_sparsity": 0.8443565704834687, "compression/movement_sparsity/model_sparsity": 0.8153503247427258, "compression_loss": 90.20851135253906, "distillation_loss": 5.44598388671875, "epoch": 3.4, "learning_rate": 4.297745773324984e-05, "loss": 95.0307, "step": 4018, "task_loss": 3.0679454803466797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8471885055925473, "compression/movement_sparsity/importance_threshold": -0.000991768737823695, "compression/movement_sparsity/linear_layer_sparsity": 0.8445159966047602, "compression/movement_sparsity/model_sparsity": 0.8155042740862989, "compression_loss": 90.23368835449219, "distillation_loss": 5.371232509613037, "epoch": 3.4, "learning_rate": 4.297432686286788e-05, "loss": 94.8268, "step": 4019, "task_loss": 3.1800289154052734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8474297868549132, "compression/movement_sparsity/importance_threshold": -0.0009902027874744423, "compression/movement_sparsity/linear_layer_sparsity": 0.8446343439685462, "compression/movement_sparsity/model_sparsity": 0.8156185558540596, "compression_loss": 90.25880432128906, "distillation_loss": 4.279092788696289, "epoch": 3.4, "learning_rate": 4.297119599248591e-05, "loss": 94.1245, "step": 4020, "task_loss": 1.9926083087921143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8476708140032823, "compression/movement_sparsity/importance_threshold": -0.000988638486361845, "compression/movement_sparsity/linear_layer_sparsity": 0.8447962860892088, "compression/movement_sparsity/model_sparsity": 0.8157749347646853, "compression_loss": 90.28387451171875, "distillation_loss": 4.150299072265625, "epoch": 3.4, "learning_rate": 4.2968065122103943e-05, "loss": 95.0927, "step": 4021, "task_loss": 2.14449405670166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8479115871715397, "compression/movement_sparsity/importance_threshold": -0.0009870758336169684, "compression/movement_sparsity/linear_layer_sparsity": 0.8449686618565528, "compression/movement_sparsity/model_sparsity": 0.8159413888941311, "compression_loss": 90.3089370727539, "distillation_loss": 2.928891658782959, "epoch": 3.4, "learning_rate": 4.296493425172198e-05, "loss": 94.3167, "step": 4022, "task_loss": 1.4815245866775513 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8481521064935703, "compression/movement_sparsity/importance_threshold": -0.0009855148283708813, "compression/movement_sparsity/linear_layer_sparsity": 0.8451017236432015, "compression/movement_sparsity/model_sparsity": 0.8160698795990625, "compression_loss": 90.33399200439453, "distillation_loss": 3.4648311138153076, "epoch": 3.4, "learning_rate": 4.2961803381340014e-05, "loss": 94.0599, "step": 4023, "task_loss": 2.169811725616455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8483923721032597, "compression/movement_sparsity/importance_threshold": -0.0009839554697546457, "compression/movement_sparsity/linear_layer_sparsity": 0.8452556884957156, "compression/movement_sparsity/model_sparsity": 0.8162185552852416, "compression_loss": 90.3588638305664, "distillation_loss": 4.185301780700684, "epoch": 3.4, "learning_rate": 4.2958672510958046e-05, "loss": 94.5921, "step": 4024, "task_loss": 3.1988108158111572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8486323841344925, "compression/movement_sparsity/importance_threshold": -0.0009823977568993317, "compression/movement_sparsity/linear_layer_sparsity": 0.8453978364981029, "compression/movement_sparsity/model_sparsity": 0.8163558200664484, "compression_loss": 90.38377380371094, "distillation_loss": 4.387216567993164, "epoch": 3.4, "learning_rate": 4.295554164057608e-05, "loss": 94.8624, "step": 4025, "task_loss": 1.859614372253418 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8488721427211543, "compression/movement_sparsity/importance_threshold": -0.0009808416889360017, "compression/movement_sparsity/linear_layer_sparsity": 0.8456523936287932, "compression/movement_sparsity/model_sparsity": 0.8166016323765902, "compression_loss": 90.4086685180664, "distillation_loss": 3.193922519683838, "epoch": 3.4, "learning_rate": 4.2952410770194116e-05, "loss": 94.5225, "step": 4026, "task_loss": 1.8470412492752075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8491116479971299, "compression/movement_sparsity/importance_threshold": -0.0009792872649957247, "compression/movement_sparsity/linear_layer_sparsity": 0.845821991065078, "compression/movement_sparsity/model_sparsity": 0.8167654036191959, "compression_loss": 90.43353271484375, "distillation_loss": 4.050387859344482, "epoch": 3.4, "learning_rate": 4.294927989981215e-05, "loss": 95.1902, "step": 4027, "task_loss": 3.9207634925842285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8493509000963049, "compression/movement_sparsity/importance_threshold": -0.000977734484209564, "compression/movement_sparsity/linear_layer_sparsity": 0.8460696083302044, "compression/movement_sparsity/model_sparsity": 0.8170045144695054, "compression_loss": 90.45832824707031, "distillation_loss": 2.7942419052124023, "epoch": 3.4, "learning_rate": 4.294614902943018e-05, "loss": 94.1397, "step": 4028, "task_loss": 1.8711886405944824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8495898991525638, "compression/movement_sparsity/importance_threshold": -0.0009761833457085888, "compression/movement_sparsity/linear_layer_sparsity": 0.8462188035156644, "compression/movement_sparsity/model_sparsity": 0.8171485843413666, "compression_loss": 90.48314666748047, "distillation_loss": 3.998152017593384, "epoch": 3.41, "learning_rate": 4.294301815904822e-05, "loss": 94.3551, "step": 4029, "task_loss": 2.2982730865478516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8498286452997922, "compression/movement_sparsity/importance_threshold": -0.0009746338486238629, "compression/movement_sparsity/linear_layer_sparsity": 0.8464036161898526, "compression/movement_sparsity/model_sparsity": 0.8173270481316462, "compression_loss": 90.5079116821289, "distillation_loss": 3.894069194793701, "epoch": 3.41, "learning_rate": 4.293988728866625e-05, "loss": 94.8667, "step": 4030, "task_loss": 2.898932933807373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.850067138671875, "compression/movement_sparsity/importance_threshold": -0.0009730859920864532, "compression/movement_sparsity/linear_layer_sparsity": 0.8464863222165749, "compression/movement_sparsity/model_sparsity": 0.8174069129519171, "compression_loss": 90.5326156616211, "distillation_loss": 5.558492183685303, "epoch": 3.41, "learning_rate": 4.293675641828429e-05, "loss": 95.5326, "step": 4031, "task_loss": 2.984650135040283 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8503053794026975, "compression/movement_sparsity/importance_threshold": -0.0009715397752274261, "compression/movement_sparsity/linear_layer_sparsity": 0.8465758727155203, "compression/movement_sparsity/model_sparsity": 0.8174933871157342, "compression_loss": 90.55735778808594, "distillation_loss": 4.638569355010986, "epoch": 3.41, "learning_rate": 4.293362554790232e-05, "loss": 94.7029, "step": 4032, "task_loss": 3.7693967819213867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8505433676261449, "compression/movement_sparsity/importance_threshold": -0.0009699951971778448, "compression/movement_sparsity/linear_layer_sparsity": 0.8467593141104304, "compression/movement_sparsity/model_sparsity": 0.8176705267343974, "compression_loss": 90.58208465576172, "distillation_loss": 4.913120269775391, "epoch": 3.41, "learning_rate": 4.293049467752036e-05, "loss": 95.0023, "step": 4033, "task_loss": 2.5414211750030518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.850781103476102, "compression/movement_sparsity/importance_threshold": -0.0009684522570687802, "compression/movement_sparsity/linear_layer_sparsity": 0.8469342654979838, "compression/movement_sparsity/model_sparsity": 0.8178394680035749, "compression_loss": 90.60673522949219, "distillation_loss": 3.3780322074890137, "epoch": 3.41, "learning_rate": 4.292736380713839e-05, "loss": 94.6121, "step": 4034, "task_loss": 3.04644775390625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8510185870864545, "compression/movement_sparsity/importance_threshold": -0.0009669109540312937, "compression/movement_sparsity/linear_layer_sparsity": 0.8470916168141066, "compression/movement_sparsity/model_sparsity": 0.8179914138179196, "compression_loss": 90.63137817382812, "distillation_loss": 2.767648220062256, "epoch": 3.41, "learning_rate": 4.292423293675642e-05, "loss": 94.168, "step": 4035, "task_loss": 1.9535332918167114 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8512558185910869, "compression/movement_sparsity/importance_threshold": -0.0009653712871964546, "compression/movement_sparsity/linear_layer_sparsity": 0.8473104372143924, "compression/movement_sparsity/model_sparsity": 0.8182027170642853, "compression_loss": 90.65602111816406, "distillation_loss": 4.993373870849609, "epoch": 3.41, "learning_rate": 4.2921102066374454e-05, "loss": 94.4586, "step": 4036, "task_loss": 3.181305170059204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8514927981238849, "compression/movement_sparsity/importance_threshold": -0.0009638332556953268, "compression/movement_sparsity/linear_layer_sparsity": 0.8474842558060203, "compression/movement_sparsity/model_sparsity": 0.8183705644525623, "compression_loss": 90.68060302734375, "distillation_loss": 3.5382003784179688, "epoch": 3.41, "learning_rate": 4.291797119599249e-05, "loss": 94.6015, "step": 4037, "task_loss": 1.6565707921981812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8517295258187332, "compression/movement_sparsity/importance_threshold": -0.0009622968586589778, "compression/movement_sparsity/linear_layer_sparsity": 0.8475208391523271, "compression/movement_sparsity/model_sparsity": 0.8184058910483799, "compression_loss": 90.70510864257812, "distillation_loss": 3.6861681938171387, "epoch": 3.41, "learning_rate": 4.2914840325610524e-05, "loss": 94.7333, "step": 4038, "task_loss": 2.6273717880249023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8519660018095174, "compression/movement_sparsity/importance_threshold": -0.0009607620952184715, "compression/movement_sparsity/linear_layer_sparsity": 0.847657394720093, "compression/movement_sparsity/model_sparsity": 0.818537755512299, "compression_loss": 90.72967529296875, "distillation_loss": 3.8877859115600586, "epoch": 3.41, "learning_rate": 4.2911709455228556e-05, "loss": 94.3496, "step": 4039, "task_loss": 2.1785426139831543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8522022262301221, "compression/movement_sparsity/importance_threshold": -0.0009592289645048782, "compression/movement_sparsity/linear_layer_sparsity": 0.8478526171926274, "compression/movement_sparsity/model_sparsity": 0.8187262714923272, "compression_loss": 90.75416564941406, "distillation_loss": 4.350160598754883, "epoch": 3.41, "learning_rate": 4.290857858484659e-05, "loss": 94.6416, "step": 4040, "task_loss": 1.8413366079330444 }, { "compression/movement_sparsity/importance_regularization_factor": 0.852438199214433, "compression/movement_sparsity/importance_threshold": -0.0009576974656492583, "compression/movement_sparsity/linear_layer_sparsity": 0.8480965021932837, "compression/movement_sparsity/model_sparsity": 0.818961778292933, "compression_loss": 90.77861022949219, "distillation_loss": 4.024568557739258, "epoch": 3.42, "learning_rate": 4.2905447714464626e-05, "loss": 94.4425, "step": 4041, "task_loss": 2.827874183654785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8526739208963346, "compression/movement_sparsity/importance_threshold": -0.0009561675977826836, "compression/movement_sparsity/linear_layer_sparsity": 0.8481609403951879, "compression/movement_sparsity/model_sparsity": 0.8190240028443667, "compression_loss": 90.80303192138672, "distillation_loss": 3.9796695709228516, "epoch": 3.42, "learning_rate": 4.290231684408266e-05, "loss": 94.9995, "step": 4042, "task_loss": 1.5393496751785278 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8529093914097126, "compression/movement_sparsity/importance_threshold": -0.0009546393600362163, "compression/movement_sparsity/linear_layer_sparsity": 0.8483100044148039, "compression/movement_sparsity/model_sparsity": 0.8191679460563343, "compression_loss": 90.82738494873047, "distillation_loss": 3.624267578125, "epoch": 3.42, "learning_rate": 4.289918597370069e-05, "loss": 94.9367, "step": 4043, "task_loss": 2.0049667358398438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.853144610888452, "compression/movement_sparsity/importance_threshold": -0.0009531127515409231, "compression/movement_sparsity/linear_layer_sparsity": 0.848608549799903, "compression/movement_sparsity/model_sparsity": 0.8194562354890221, "compression_loss": 90.85172271728516, "distillation_loss": 5.643865585327148, "epoch": 3.42, "learning_rate": 4.289605510331872e-05, "loss": 96.3429, "step": 4044, "task_loss": 2.866238832473755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8533795794664376, "compression/movement_sparsity/importance_threshold": -0.0009515877714278715, "compression/movement_sparsity/linear_layer_sparsity": 0.8487938752132995, "compression/movement_sparsity/model_sparsity": 0.819635194404341, "compression_loss": 90.8760986328125, "distillation_loss": 5.27628755569458, "epoch": 3.42, "learning_rate": 4.289292423293676e-05, "loss": 95.8004, "step": 4045, "task_loss": 2.9614202976226807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8536142972775551, "compression/movement_sparsity/importance_threshold": -0.0009500644188281246, "compression/movement_sparsity/linear_layer_sparsity": 0.8489070832608344, "compression/movement_sparsity/model_sparsity": 0.8197445134071742, "compression_loss": 90.90042877197266, "distillation_loss": 4.68117618560791, "epoch": 3.42, "learning_rate": 4.288979336255479e-05, "loss": 95.3196, "step": 4046, "task_loss": 3.81087589263916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8538487644556892, "compression/movement_sparsity/importance_threshold": -0.0009485426928727515, "compression/movement_sparsity/linear_layer_sparsity": 0.8491997738796243, "compression/movement_sparsity/model_sparsity": 0.8200271492027871, "compression_loss": 90.92478942871094, "distillation_loss": 4.683239459991455, "epoch": 3.42, "learning_rate": 4.2886662492172824e-05, "loss": 95.3726, "step": 4047, "task_loss": 1.8897018432617188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8540829811347255, "compression/movement_sparsity/importance_threshold": -0.0009470225926928163, "compression/movement_sparsity/linear_layer_sparsity": 0.8493199217927232, "compression/movement_sparsity/model_sparsity": 0.8201431696654529, "compression_loss": 90.94906616210938, "distillation_loss": 4.32224178314209, "epoch": 3.42, "learning_rate": 4.288353162179086e-05, "loss": 95.8875, "step": 4048, "task_loss": 2.6872901916503906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8543169474485485, "compression/movement_sparsity/importance_threshold": -0.0009455041174193865, "compression/movement_sparsity/linear_layer_sparsity": 0.8494661120879389, "compression/movement_sparsity/model_sparsity": 0.820284337874294, "compression_loss": 90.97335815429688, "distillation_loss": 6.068289756774902, "epoch": 3.42, "learning_rate": 4.2880400751408894e-05, "loss": 95.3966, "step": 4049, "task_loss": 3.387033224105835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8545506635310439, "compression/movement_sparsity/importance_threshold": -0.000943987266183526, "compression/movement_sparsity/linear_layer_sparsity": 0.849695759632438, "compression/movement_sparsity/model_sparsity": 0.820506096319161, "compression_loss": 90.99754333496094, "distillation_loss": 3.360217571258545, "epoch": 3.42, "learning_rate": 4.2877269881026926e-05, "loss": 95.4113, "step": 4050, "task_loss": 2.389498472213745 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8547841295160965, "compression/movement_sparsity/importance_threshold": -0.0009424720381163041, "compression/movement_sparsity/linear_layer_sparsity": 0.8497611517677531, "compression/movement_sparsity/model_sparsity": 0.8205692420334583, "compression_loss": 91.02181243896484, "distillation_loss": 4.718728065490723, "epoch": 3.42, "learning_rate": 4.287413901064496e-05, "loss": 95.8023, "step": 4051, "task_loss": 3.8667855262756348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8550173455375915, "compression/movement_sparsity/importance_threshold": -0.0009409584323487848, "compression/movement_sparsity/linear_layer_sparsity": 0.8499838475225204, "compression/movement_sparsity/model_sparsity": 0.8207842875039572, "compression_loss": 91.04597473144531, "distillation_loss": 4.793120384216309, "epoch": 3.42, "learning_rate": 4.2871008140262996e-05, "loss": 95.0319, "step": 4052, "task_loss": 2.795159101486206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8552503117294142, "compression/movement_sparsity/importance_threshold": -0.0009394464480120338, "compression/movement_sparsity/linear_layer_sparsity": 0.8502007362076491, "compression/movement_sparsity/model_sparsity": 0.8209937253955241, "compression_loss": 91.07012176513672, "distillation_loss": 4.575957298278809, "epoch": 3.43, "learning_rate": 4.286787726988103e-05, "loss": 94.7697, "step": 4053, "task_loss": 2.311068296432495 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8554830282254496, "compression/movement_sparsity/importance_threshold": -0.0009379360842371176, "compression/movement_sparsity/linear_layer_sparsity": 0.8503474392420732, "compression/movement_sparsity/model_sparsity": 0.8211353887294043, "compression_loss": 91.09427642822266, "distillation_loss": 5.741728782653809, "epoch": 3.43, "learning_rate": 4.286474639949906e-05, "loss": 95.5923, "step": 4054, "task_loss": 3.1386349201202393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8557154951595831, "compression/movement_sparsity/importance_threshold": -0.0009364273401551012, "compression/movement_sparsity/linear_layer_sparsity": 0.8504821465638556, "compression/movement_sparsity/model_sparsity": 0.8212654684402751, "compression_loss": 91.11835479736328, "distillation_loss": 4.642451286315918, "epoch": 3.43, "learning_rate": 4.286161552911709e-05, "loss": 96.0672, "step": 4055, "task_loss": 2.6202549934387207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8559477126656994, "compression/movement_sparsity/importance_threshold": -0.0009349202148970538, "compression/movement_sparsity/linear_layer_sparsity": 0.8505698368926496, "compression/movement_sparsity/model_sparsity": 0.8213501463365083, "compression_loss": 91.14237976074219, "distillation_loss": 3.38331937789917, "epoch": 3.43, "learning_rate": 4.285848465873513e-05, "loss": 95.5212, "step": 4056, "task_loss": 1.834960699081421 }, { "compression/movement_sparsity/importance_regularization_factor": 0.856179680877684, "compression/movement_sparsity/importance_threshold": -0.0009334147075940368, "compression/movement_sparsity/linear_layer_sparsity": 0.8505114561679045, "compression/movement_sparsity/model_sparsity": 0.8212937711692582, "compression_loss": 91.1664047241211, "distillation_loss": 4.22911262512207, "epoch": 3.43, "learning_rate": 4.285535378835316e-05, "loss": 95.1376, "step": 4057, "task_loss": 2.0311012268066406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8564113999294217, "compression/movement_sparsity/importance_threshold": -0.0009319108173771219, "compression/movement_sparsity/linear_layer_sparsity": 0.8506258685563706, "compression/movement_sparsity/model_sparsity": 0.8214042531402068, "compression_loss": 91.19041442871094, "distillation_loss": 4.545516014099121, "epoch": 3.43, "learning_rate": 4.2852222917971194e-05, "loss": 95.0113, "step": 4058, "task_loss": 2.6082897186279297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8566428699547981, "compression/movement_sparsity/importance_threshold": -0.0009304085433773698, "compression/movement_sparsity/linear_layer_sparsity": 0.8507558419836015, "compression/movement_sparsity/model_sparsity": 0.8215297615803673, "compression_loss": 91.21437072753906, "distillation_loss": 4.775699615478516, "epoch": 3.43, "learning_rate": 4.284909204758923e-05, "loss": 95.4207, "step": 4059, "task_loss": 2.5048844814300537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8568740910876979, "compression/movement_sparsity/importance_threshold": -0.0009289078847258504, "compression/movement_sparsity/linear_layer_sparsity": 0.8508874967184691, "compression/movement_sparsity/model_sparsity": 0.8216568935700748, "compression_loss": 91.23835754394531, "distillation_loss": 3.838752031326294, "epoch": 3.43, "learning_rate": 4.2845961177207264e-05, "loss": 95.1503, "step": 4060, "task_loss": 2.093212842941284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8571050634620064, "compression/movement_sparsity/importance_threshold": -0.0009274088405536278, "compression/movement_sparsity/linear_layer_sparsity": 0.850993597962093, "compression/movement_sparsity/model_sparsity": 0.8217593499095747, "compression_loss": 91.2623062133789, "distillation_loss": 7.582210540771484, "epoch": 3.43, "learning_rate": 4.2842830306825296e-05, "loss": 96.3486, "step": 4061, "task_loss": 2.947300434112549 }, { "compression/movement_sparsity/importance_regularization_factor": 0.857335787211609, "compression/movement_sparsity/importance_threshold": -0.0009259114099917677, "compression/movement_sparsity/linear_layer_sparsity": 0.8511409329774017, "compression/movement_sparsity/model_sparsity": 0.821901623513852, "compression_loss": 91.28621673583984, "distillation_loss": 5.406777858734131, "epoch": 3.43, "learning_rate": 4.2839699436443335e-05, "loss": 95.6669, "step": 4062, "task_loss": 2.5441339015960693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8575662624703904, "compression/movement_sparsity/importance_threshold": -0.0009244155921713376, "compression/movement_sparsity/linear_layer_sparsity": 0.8512029982699464, "compression/movement_sparsity/model_sparsity": 0.8219615566726626, "compression_loss": 91.31007385253906, "distillation_loss": 4.05820894241333, "epoch": 3.43, "learning_rate": 4.2836568566061366e-05, "loss": 95.6002, "step": 4063, "task_loss": 3.196995735168457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8577964893722361, "compression/movement_sparsity/importance_threshold": -0.0009229213862234015, "compression/movement_sparsity/linear_layer_sparsity": 0.8513301218211123, "compression/movement_sparsity/model_sparsity": 0.8220843131387682, "compression_loss": 91.33393859863281, "distillation_loss": 3.867387294769287, "epoch": 3.44, "learning_rate": 4.28334376956794e-05, "loss": 95.8362, "step": 4064, "task_loss": 1.9553067684173584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.858026468051031, "compression/movement_sparsity/importance_threshold": -0.0009214287912790277, "compression/movement_sparsity/linear_layer_sparsity": 0.8514805213475034, "compression/movement_sparsity/model_sparsity": 0.8222295459787446, "compression_loss": 91.35774993896484, "distillation_loss": 3.1497139930725098, "epoch": 3.44, "learning_rate": 4.283030682529744e-05, "loss": 95.2627, "step": 4065, "task_loss": 2.134451150894165 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8582561986406604, "compression/movement_sparsity/importance_threshold": -0.0009199378064692803, "compression/movement_sparsity/linear_layer_sparsity": 0.8516576190852313, "compression/movement_sparsity/model_sparsity": 0.8224005598643652, "compression_loss": 91.38159942626953, "distillation_loss": 5.669510364532471, "epoch": 3.44, "learning_rate": 4.282717595491547e-05, "loss": 95.7254, "step": 4066, "task_loss": 3.405003070831299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8584856812750092, "compression/movement_sparsity/importance_threshold": -0.0009184484309252275, "compression/movement_sparsity/linear_layer_sparsity": 0.8517738677955133, "compression/movement_sparsity/model_sparsity": 0.8225128150738262, "compression_loss": 91.4052963256836, "distillation_loss": 3.8016552925109863, "epoch": 3.44, "learning_rate": 4.282404508453351e-05, "loss": 95.0453, "step": 4067, "task_loss": 3.0243308544158936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.858714916087963, "compression/movement_sparsity/importance_threshold": -0.0009169606637779333, "compression/movement_sparsity/linear_layer_sparsity": 0.8517783155100415, "compression/movement_sparsity/model_sparsity": 0.8225171099956775, "compression_loss": 91.42909240722656, "distillation_loss": 3.2965478897094727, "epoch": 3.44, "learning_rate": 4.282091421415154e-05, "loss": 95.3276, "step": 4068, "task_loss": 1.5958503484725952 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8589439032134064, "compression/movement_sparsity/importance_threshold": -0.0009154745041584653, "compression/movement_sparsity/linear_layer_sparsity": 0.8519607075781996, "compression/movement_sparsity/model_sparsity": 0.8226932363351909, "compression_loss": 91.4527816772461, "distillation_loss": 5.953710079193115, "epoch": 3.44, "learning_rate": 4.281778334376957e-05, "loss": 95.2799, "step": 4069, "task_loss": 2.1279356479644775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8591726427852251, "compression/movement_sparsity/importance_threshold": -0.0009139899511978874, "compression/movement_sparsity/linear_layer_sparsity": 0.8520507588921857, "compression/movement_sparsity/model_sparsity": 0.8227801941095112, "compression_loss": 91.47644805908203, "distillation_loss": 5.350955009460449, "epoch": 3.44, "learning_rate": 4.281465247338761e-05, "loss": 95.7455, "step": 4070, "task_loss": 3.04465913772583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8594011349373036, "compression/movement_sparsity/importance_threshold": -0.0009125070040272689, "compression/movement_sparsity/linear_layer_sparsity": 0.8521991193859111, "compression/movement_sparsity/model_sparsity": 0.8229234579638669, "compression_loss": 91.50011444091797, "distillation_loss": 6.037020683288574, "epoch": 3.44, "learning_rate": 4.281152160300564e-05, "loss": 96.0376, "step": 4071, "task_loss": 3.7693352699279785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8596293798035276, "compression/movement_sparsity/importance_threshold": -0.0009110256617776728, "compression/movement_sparsity/linear_layer_sparsity": 0.8522600161100274, "compression/movement_sparsity/model_sparsity": 0.8229822626981697, "compression_loss": 91.52375793457031, "distillation_loss": 4.624373435974121, "epoch": 3.44, "learning_rate": 4.280839073262367e-05, "loss": 95.8345, "step": 4072, "task_loss": 3.052806854248047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8598573775177818, "compression/movement_sparsity/importance_threshold": -0.0009095459235801675, "compression/movement_sparsity/linear_layer_sparsity": 0.8523854225810538, "compression/movement_sparsity/model_sparsity": 0.8231033610711208, "compression_loss": 91.5472412109375, "distillation_loss": 5.122166633605957, "epoch": 3.44, "learning_rate": 4.2805259862241704e-05, "loss": 96.4756, "step": 4073, "task_loss": 2.501249074935913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8600851282139517, "compression/movement_sparsity/importance_threshold": -0.0009080677885658171, "compression/movement_sparsity/linear_layer_sparsity": 0.8525089331094261, "compression/movement_sparsity/model_sparsity": 0.8232226286328806, "compression_loss": 91.57084655761719, "distillation_loss": 3.7264931201934814, "epoch": 3.44, "learning_rate": 4.280212899185974e-05, "loss": 96.0387, "step": 4074, "task_loss": 2.0143940448760986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8603126320259223, "compression/movement_sparsity/importance_threshold": -0.0009065912558656871, "compression/movement_sparsity/linear_layer_sparsity": 0.8526852915487605, "compression/movement_sparsity/model_sparsity": 0.823392928617282, "compression_loss": 91.59439086914062, "distillation_loss": 4.197587966918945, "epoch": 3.44, "learning_rate": 4.2798998121477775e-05, "loss": 96.0125, "step": 4075, "task_loss": 2.398454189300537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8605398890875786, "compression/movement_sparsity/importance_threshold": -0.0009051163246108478, "compression/movement_sparsity/linear_layer_sparsity": 0.8527759986919665, "compression/movement_sparsity/model_sparsity": 0.823480519691071, "compression_loss": 91.61791229248047, "distillation_loss": 3.3390684127807617, "epoch": 3.45, "learning_rate": 4.2795867251095807e-05, "loss": 95.9049, "step": 4076, "task_loss": 2.356628894805908 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8607668995328062, "compression/movement_sparsity/importance_threshold": -0.0009036429939323597, "compression/movement_sparsity/linear_layer_sparsity": 0.8528186633637677, "compression/movement_sparsity/model_sparsity": 0.8235217187001439, "compression_loss": 91.641357421875, "distillation_loss": 5.169465065002441, "epoch": 3.45, "learning_rate": 4.279273638071384e-05, "loss": 95.9708, "step": 4077, "task_loss": 3.2168400287628174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8609936634954896, "compression/movement_sparsity/importance_threshold": -0.0009021712629612927, "compression/movement_sparsity/linear_layer_sparsity": 0.8528899341137272, "compression/movement_sparsity/model_sparsity": 0.8235905410805878, "compression_loss": 91.66477966308594, "distillation_loss": 3.532865524291992, "epoch": 3.45, "learning_rate": 4.278960551033188e-05, "loss": 95.3649, "step": 4078, "task_loss": 1.7807132005691528 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8612201811095145, "compression/movement_sparsity/importance_threshold": -0.0009007011308287101, "compression/movement_sparsity/linear_layer_sparsity": 0.8531127252618357, "compression/movement_sparsity/model_sparsity": 0.823805678667373, "compression_loss": 91.68819427490234, "distillation_loss": 5.101552963256836, "epoch": 3.45, "learning_rate": 4.278647463994991e-05, "loss": 96.6948, "step": 4079, "task_loss": 2.1609959602355957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8614464525087656, "compression/movement_sparsity/importance_threshold": -0.0008992325966656811, "compression/movement_sparsity/linear_layer_sparsity": 0.8532247289684393, "compression/movement_sparsity/model_sparsity": 0.8239138347020911, "compression_loss": 91.71165466308594, "distillation_loss": 4.482922077178955, "epoch": 3.45, "learning_rate": 4.278334376956794e-05, "loss": 96.7284, "step": 4080, "task_loss": 2.132432460784912 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8616724778271284, "compression/movement_sparsity/importance_threshold": -0.0008977656596032687, "compression/movement_sparsity/linear_layer_sparsity": 0.853476817796429, "compression/movement_sparsity/model_sparsity": 0.8241572635033235, "compression_loss": 91.73504638671875, "distillation_loss": 4.285619258880615, "epoch": 3.45, "learning_rate": 4.278021289918598e-05, "loss": 95.9288, "step": 4081, "task_loss": 2.952066659927368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8618982571984878, "compression/movement_sparsity/importance_threshold": -0.0008963003187725414, "compression/movement_sparsity/linear_layer_sparsity": 0.8535310131383341, "compression/movement_sparsity/model_sparsity": 0.8242095970685097, "compression_loss": 91.75840759277344, "distillation_loss": 6.511777400970459, "epoch": 3.45, "learning_rate": 4.277708202880401e-05, "loss": 96.7119, "step": 4082, "task_loss": 5.260255336761475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.862123790756729, "compression/movement_sparsity/importance_threshold": -0.000894836573304564, "compression/movement_sparsity/linear_layer_sparsity": 0.8537119266097054, "compression/movement_sparsity/model_sparsity": 0.8243842956055845, "compression_loss": 91.78173828125, "distillation_loss": 3.144404411315918, "epoch": 3.45, "learning_rate": 4.277395115842204e-05, "loss": 95.1451, "step": 4083, "task_loss": 1.4554139375686646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8623490786357372, "compression/movement_sparsity/importance_threshold": -0.000893374422330403, "compression/movement_sparsity/linear_layer_sparsity": 0.8538991360415884, "compression/movement_sparsity/model_sparsity": 0.8245650738175588, "compression_loss": 91.80508422851562, "distillation_loss": 3.747413396835327, "epoch": 3.45, "learning_rate": 4.2770820288040074e-05, "loss": 96.0429, "step": 4084, "task_loss": 0.7696890234947205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8625741209693975, "compression/movement_sparsity/importance_threshold": -0.0008919138649811234, "compression/movement_sparsity/linear_layer_sparsity": 0.8539756415011401, "compression/movement_sparsity/model_sparsity": 0.8246389510792166, "compression_loss": 91.82843780517578, "distillation_loss": 5.231804847717285, "epoch": 3.45, "learning_rate": 4.276768941765811e-05, "loss": 96.3289, "step": 4085, "task_loss": 2.050168991088867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8627989178915951, "compression/movement_sparsity/importance_threshold": -0.000890454900387791, "compression/movement_sparsity/linear_layer_sparsity": 0.8540985558211306, "compression/movement_sparsity/model_sparsity": 0.8247576429141867, "compression_loss": 91.85169982910156, "distillation_loss": 4.534160614013672, "epoch": 3.45, "learning_rate": 4.2764558547276145e-05, "loss": 96.3221, "step": 4086, "task_loss": 2.1881096363067627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8630234695362149, "compression/movement_sparsity/importance_threshold": -0.0008889975276814756, "compression/movement_sparsity/linear_layer_sparsity": 0.854243446382074, "compression/movement_sparsity/model_sparsity": 0.8248975560386261, "compression_loss": 91.87496185302734, "distillation_loss": 4.491280555725098, "epoch": 3.45, "learning_rate": 4.2761427676894177e-05, "loss": 96.4998, "step": 4087, "task_loss": 2.3135406970977783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8632477760371424, "compression/movement_sparsity/importance_threshold": -0.000887541745993238, "compression/movement_sparsity/linear_layer_sparsity": 0.8544419599248758, "compression/movement_sparsity/model_sparsity": 0.8250892500305337, "compression_loss": 91.89823150634766, "distillation_loss": 5.49393367767334, "epoch": 3.46, "learning_rate": 4.275829680651221e-05, "loss": 96.7082, "step": 4088, "task_loss": 2.956188201904297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8634718375282625, "compression/movement_sparsity/importance_threshold": -0.0008860875544541473, "compression/movement_sparsity/linear_layer_sparsity": 0.854577525786728, "compression/movement_sparsity/model_sparsity": 0.8252201587879818, "compression_loss": 91.92144775390625, "distillation_loss": 7.493169784545898, "epoch": 3.46, "learning_rate": 4.275516593613025e-05, "loss": 97.1446, "step": 4089, "task_loss": 4.365863800048828 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8636956541434606, "compression/movement_sparsity/importance_threshold": -0.0008846349521952684, "compression/movement_sparsity/linear_layer_sparsity": 0.8547122211843426, "compression/movement_sparsity/model_sparsity": 0.8253502269843169, "compression_loss": 91.9446792602539, "distillation_loss": 3.7829973697662354, "epoch": 3.46, "learning_rate": 4.275203506574828e-05, "loss": 97.3043, "step": 4090, "task_loss": 2.840228319168091 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8639192260166213, "compression/movement_sparsity/importance_threshold": -0.0008831839383476696, "compression/movement_sparsity/linear_layer_sparsity": 0.8547125312127012, "compression/movement_sparsity/model_sparsity": 0.8253505263622476, "compression_loss": 91.96780395507812, "distillation_loss": 6.404736518859863, "epoch": 3.46, "learning_rate": 4.274890419536631e-05, "loss": 96.9881, "step": 4091, "task_loss": 3.1974544525146484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8641425532816303, "compression/movement_sparsity/importance_threshold": -0.000881734512042414, "compression/movement_sparsity/linear_layer_sparsity": 0.8549140496457472, "compression/movement_sparsity/model_sparsity": 0.8255451220171753, "compression_loss": 91.99095916748047, "distillation_loss": 4.152040481567383, "epoch": 3.46, "learning_rate": 4.274577332498434e-05, "loss": 96.4676, "step": 4092, "task_loss": 2.2532827854156494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8643656360723726, "compression/movement_sparsity/importance_threshold": -0.0008802866724105683, "compression/movement_sparsity/linear_layer_sparsity": 0.8550394561167737, "compression/movement_sparsity/model_sparsity": 0.8256662203901265, "compression_loss": 92.01403045654297, "distillation_loss": 5.130697250366211, "epoch": 3.46, "learning_rate": 4.274264245460238e-05, "loss": 97.0667, "step": 4093, "task_loss": 2.6219992637634277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.864588474522733, "compression/movement_sparsity/importance_threshold": -0.0008788404185832007, "compression/movement_sparsity/linear_layer_sparsity": 0.8551026899777466, "compression/movement_sparsity/model_sparsity": 0.8257272819734449, "compression_loss": 92.03719329833984, "distillation_loss": 4.272568702697754, "epoch": 3.46, "learning_rate": 4.273951158422041e-05, "loss": 96.1509, "step": 4094, "task_loss": 1.5554879903793335 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8648110687665972, "compression/movement_sparsity/importance_threshold": -0.0008773957496913745, "compression/movement_sparsity/linear_layer_sparsity": 0.8551783368972286, "compression/movement_sparsity/model_sparsity": 0.8258003301885255, "compression_loss": 92.06021881103516, "distillation_loss": 3.616347551345825, "epoch": 3.46, "learning_rate": 4.2736380713838444e-05, "loss": 96.1452, "step": 4095, "task_loss": 3.4659438133239746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8650334189378499, "compression/movement_sparsity/importance_threshold": -0.0008759526648661571, "compression/movement_sparsity/linear_layer_sparsity": 0.8552283349321257, "compression/movement_sparsity/model_sparsity": 0.825848610637112, "compression_loss": 92.0832748413086, "distillation_loss": 3.5386009216308594, "epoch": 3.46, "learning_rate": 4.273324984345648e-05, "loss": 96.461, "step": 4096, "task_loss": 1.8152415752410889 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8652555251703765, "compression/movement_sparsity/importance_threshold": -0.0008745111632386133, "compression/movement_sparsity/linear_layer_sparsity": 0.8553659636749789, "compression/movement_sparsity/model_sparsity": 0.8259815114092527, "compression_loss": 92.10637664794922, "distillation_loss": 3.644550323486328, "epoch": 3.46, "learning_rate": 4.2730118973074515e-05, "loss": 96.2518, "step": 4097, "task_loss": 2.019357204437256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8654773875980619, "compression/movement_sparsity/importance_threshold": -0.0008730712439398115, "compression/movement_sparsity/linear_layer_sparsity": 0.8555034135553174, "compression/movement_sparsity/model_sparsity": 0.8261142394633563, "compression_loss": 92.12939453125, "distillation_loss": 2.752410888671875, "epoch": 3.46, "learning_rate": 4.272698810269255e-05, "loss": 96.1122, "step": 4098, "task_loss": 1.6103150844573975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8656990063547915, "compression/movement_sparsity/importance_threshold": -0.0008716329061008157, "compression/movement_sparsity/linear_layer_sparsity": 0.8556378466212442, "compression/movement_sparsity/model_sparsity": 0.826244054339904, "compression_loss": 92.15237426757812, "distillation_loss": 5.972105979919434, "epoch": 3.46, "learning_rate": 4.2723857232310585e-05, "loss": 96.7285, "step": 4099, "task_loss": 2.3267695903778076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8659203815744502, "compression/movement_sparsity/importance_threshold": -0.0008701961488526934, "compression/movement_sparsity/linear_layer_sparsity": 0.8557351716774881, "compression/movement_sparsity/model_sparsity": 0.826338035981059, "compression_loss": 92.17525482177734, "distillation_loss": 4.032285690307617, "epoch": 3.47, "learning_rate": 4.2720726361928624e-05, "loss": 96.4382, "step": 4100, "task_loss": 1.6760855913162231 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8661415133909232, "compression/movement_sparsity/importance_threshold": -0.0008687609713265093, "compression/movement_sparsity/linear_layer_sparsity": 0.855852243155337, "compression/movement_sparsity/model_sparsity": 0.8264510856934898, "compression_loss": 92.19818115234375, "distillation_loss": 4.355818271636963, "epoch": 3.47, "learning_rate": 4.2717595491546655e-05, "loss": 96.1297, "step": 4101, "task_loss": 2.057081937789917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8663624019380959, "compression/movement_sparsity/importance_threshold": -0.0008673273726533294, "compression/movement_sparsity/linear_layer_sparsity": 0.8559609677158407, "compression/movement_sparsity/model_sparsity": 0.8265560752308644, "compression_loss": 92.22102355957031, "distillation_loss": 4.509158134460449, "epoch": 3.47, "learning_rate": 4.271446462116469e-05, "loss": 96.557, "step": 4102, "task_loss": 2.020683765411377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.866583047349853, "compression/movement_sparsity/importance_threshold": -0.0008658953519642219, "compression/movement_sparsity/linear_layer_sparsity": 0.8561271190676789, "compression/movement_sparsity/model_sparsity": 0.8267165187726255, "compression_loss": 92.24385070800781, "distillation_loss": 3.8076677322387695, "epoch": 3.47, "learning_rate": 4.271133375078272e-05, "loss": 96.2854, "step": 4103, "task_loss": 2.4665651321411133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8668034497600802, "compression/movement_sparsity/importance_threshold": -0.000864464908390249, "compression/movement_sparsity/linear_layer_sparsity": 0.8563430299710615, "compression/movement_sparsity/model_sparsity": 0.8269250124722574, "compression_loss": 92.26664733886719, "distillation_loss": 6.014349460601807, "epoch": 3.47, "learning_rate": 4.270820288040076e-05, "loss": 96.9596, "step": 4104, "task_loss": 4.036872386932373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.867023609302662, "compression/movement_sparsity/importance_threshold": -0.0008630360410624809, "compression/movement_sparsity/linear_layer_sparsity": 0.8566138278180719, "compression/movement_sparsity/model_sparsity": 0.8271865075801513, "compression_loss": 92.2894287109375, "distillation_loss": 3.830193519592285, "epoch": 3.47, "learning_rate": 4.270507201001879e-05, "loss": 96.0989, "step": 4105, "task_loss": 2.3634798526763916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.867243526111484, "compression/movement_sparsity/importance_threshold": -0.0008616087491119806, "compression/movement_sparsity/linear_layer_sparsity": 0.8567274412873064, "compression/movement_sparsity/model_sparsity": 0.8272962180772017, "compression_loss": 92.3121566772461, "distillation_loss": 6.199346542358398, "epoch": 3.47, "learning_rate": 4.270194113963682e-05, "loss": 97.2922, "step": 4106, "task_loss": 2.6651480197906494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8674632003204312, "compression/movement_sparsity/importance_threshold": -0.0008601830316698157, "compression/movement_sparsity/linear_layer_sparsity": 0.8568208552165658, "compression/movement_sparsity/model_sparsity": 0.8273864229506162, "compression_loss": 92.33487701416016, "distillation_loss": 4.692887306213379, "epoch": 3.47, "learning_rate": 4.269881026925486e-05, "loss": 96.8083, "step": 4107, "task_loss": 3.11521053314209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8676826320633888, "compression/movement_sparsity/importance_threshold": -0.0008587588878670518, "compression/movement_sparsity/linear_layer_sparsity": 0.8569113954214249, "compression/movement_sparsity/model_sparsity": 0.8274738528209041, "compression_loss": 92.35761260986328, "distillation_loss": 6.040011405944824, "epoch": 3.47, "learning_rate": 4.269567939887289e-05, "loss": 97.743, "step": 4108, "task_loss": 3.1440882682800293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8679018214742417, "compression/movement_sparsity/importance_threshold": -0.0008573363168347566, "compression/movement_sparsity/linear_layer_sparsity": 0.8570661114965002, "compression/movement_sparsity/model_sparsity": 0.8276232539228383, "compression_loss": 92.38037109375, "distillation_loss": 4.844672203063965, "epoch": 3.47, "learning_rate": 4.269254852849092e-05, "loss": 96.5227, "step": 4109, "task_loss": 2.368454694747925 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8681207686868754, "compression/movement_sparsity/importance_threshold": -0.0008559153177039921, "compression/movement_sparsity/linear_layer_sparsity": 0.8572251679685949, "compression/movement_sparsity/model_sparsity": 0.8277768463158016, "compression_loss": 92.40300750732422, "distillation_loss": 5.626307487487793, "epoch": 3.47, "learning_rate": 4.2689417658108955e-05, "loss": 97.6702, "step": 4110, "task_loss": 3.3963913917541504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8683394738351747, "compression/movement_sparsity/importance_threshold": -0.0008544958896058276, "compression/movement_sparsity/linear_layer_sparsity": 0.8573907111878837, "compression/movement_sparsity/model_sparsity": 0.8279367026162373, "compression_loss": 92.4256820678711, "distillation_loss": 3.367100954055786, "epoch": 3.47, "learning_rate": 4.2686286787726993e-05, "loss": 96.8878, "step": 4111, "task_loss": 1.9786967039108276 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8685579370530251, "compression/movement_sparsity/importance_threshold": -0.0008530780316713281, "compression/movement_sparsity/linear_layer_sparsity": 0.8575610121500591, "compression/movement_sparsity/model_sparsity": 0.8281011532164548, "compression_loss": 92.44828033447266, "distillation_loss": 4.339536666870117, "epoch": 3.48, "learning_rate": 4.2683155917345025e-05, "loss": 96.6687, "step": 4112, "task_loss": 2.1698246002197266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8687761584743113, "compression/movement_sparsity/importance_threshold": -0.0008516617430315608, "compression/movement_sparsity/linear_layer_sparsity": 0.8577606227162834, "compression/movement_sparsity/model_sparsity": 0.8282939065456555, "compression_loss": 92.47090148925781, "distillation_loss": 4.436316967010498, "epoch": 3.48, "learning_rate": 4.268002504696306e-05, "loss": 97.0219, "step": 4113, "task_loss": 3.4045677185058594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8689941382329187, "compression/movement_sparsity/importance_threshold": -0.0008502470228175909, "compression/movement_sparsity/linear_layer_sparsity": 0.8578721136836788, "compression/movement_sparsity/model_sparsity": 0.8284015674553344, "compression_loss": 92.49346923828125, "distillation_loss": 6.214099884033203, "epoch": 3.48, "learning_rate": 4.267689417658109e-05, "loss": 97.3631, "step": 4114, "task_loss": 2.9369797706604004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8692118764627327, "compression/movement_sparsity/importance_threshold": -0.000848833870160483, "compression/movement_sparsity/linear_layer_sparsity": 0.8580320286958433, "compression/movement_sparsity/model_sparsity": 0.828555988894875, "compression_loss": 92.51607513427734, "distillation_loss": 4.948198318481445, "epoch": 3.48, "learning_rate": 4.267376330619913e-05, "loss": 97.0878, "step": 4115, "task_loss": 2.5488948822021484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8694293732976379, "compression/movement_sparsity/importance_threshold": -0.0008474222841913055, "compression/movement_sparsity/linear_layer_sparsity": 0.8580336861551446, "compression/movement_sparsity/model_sparsity": 0.8285575894153505, "compression_loss": 92.53862762451172, "distillation_loss": 4.380139350891113, "epoch": 3.48, "learning_rate": 4.267063243581716e-05, "loss": 96.9896, "step": 4116, "task_loss": 2.7281484603881836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8696466288715199, "compression/movement_sparsity/importance_threshold": -0.0008460122640411225, "compression/movement_sparsity/linear_layer_sparsity": 0.8581952586266106, "compression/movement_sparsity/model_sparsity": 0.8287136113753665, "compression_loss": 92.56118774414062, "distillation_loss": 4.448000907897949, "epoch": 3.48, "learning_rate": 4.266750156543519e-05, "loss": 97.3197, "step": 4117, "task_loss": 2.8121144771575928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8698636433182634, "compression/movement_sparsity/importance_threshold": -0.0008446038088410022, "compression/movement_sparsity/linear_layer_sparsity": 0.8584522125149958, "compression/movement_sparsity/model_sparsity": 0.8289617381072031, "compression_loss": 92.5837173461914, "distillation_loss": 4.381775856018066, "epoch": 3.48, "learning_rate": 4.266437069505323e-05, "loss": 96.3446, "step": 4118, "task_loss": 1.9501968622207642 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8700804167717541, "compression/movement_sparsity/importance_threshold": -0.0008431969177220078, "compression/movement_sparsity/linear_layer_sparsity": 0.8586280820634572, "compression/movement_sparsity/model_sparsity": 0.8291315659956368, "compression_loss": 92.60617065429688, "distillation_loss": 3.6623854637145996, "epoch": 3.48, "learning_rate": 4.266123982467126e-05, "loss": 96.5383, "step": 4119, "task_loss": 1.4624898433685303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8702969493658765, "compression/movement_sparsity/importance_threshold": -0.0008417915898152085, "compression/movement_sparsity/linear_layer_sparsity": 0.8587271361240083, "compression/movement_sparsity/model_sparsity": 0.829227217244482, "compression_loss": 92.628662109375, "distillation_loss": 4.480627059936523, "epoch": 3.48, "learning_rate": 4.265810895428929e-05, "loss": 96.7881, "step": 4120, "task_loss": 2.1061596870422363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8705132412345163, "compression/movement_sparsity/importance_threshold": -0.0008403878242516675, "compression/movement_sparsity/linear_layer_sparsity": 0.8588393067689588, "compression/movement_sparsity/model_sparsity": 0.8293355344827013, "compression_loss": 92.651123046875, "distillation_loss": 3.541402816772461, "epoch": 3.48, "learning_rate": 4.2654978083907325e-05, "loss": 96.6408, "step": 4121, "task_loss": 1.2738748788833618 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8707292925115583, "compression/movement_sparsity/importance_threshold": -0.0008389856201624531, "compression/movement_sparsity/linear_layer_sparsity": 0.8590455948690593, "compression/movement_sparsity/model_sparsity": 0.8295347359519468, "compression_loss": 92.67350006103516, "distillation_loss": 4.094985008239746, "epoch": 3.48, "learning_rate": 4.2651847213525363e-05, "loss": 97.0348, "step": 4122, "task_loss": 2.338268756866455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8709451033308877, "compression/movement_sparsity/importance_threshold": -0.0008375849766786301, "compression/movement_sparsity/linear_layer_sparsity": 0.8592242427485798, "compression/movement_sparsity/model_sparsity": 0.8297072467272206, "compression_loss": 92.69586944580078, "distillation_loss": 5.194477081298828, "epoch": 3.48, "learning_rate": 4.2648716343143395e-05, "loss": 97.6415, "step": 4123, "task_loss": 2.8624584674835205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8711606738263898, "compression/movement_sparsity/importance_threshold": -0.0008361858929312642, "compression/movement_sparsity/linear_layer_sparsity": 0.8593020360182362, "compression/movement_sparsity/model_sparsity": 0.8297823675587442, "compression_loss": 92.71824645996094, "distillation_loss": 3.4921305179595947, "epoch": 3.49, "learning_rate": 4.264558547276143e-05, "loss": 96.763, "step": 4124, "task_loss": 3.081404447555542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8713760041319496, "compression/movement_sparsity/importance_threshold": -0.0008347883680514222, "compression/movement_sparsity/linear_layer_sparsity": 0.8593867095326184, "compression/movement_sparsity/model_sparsity": 0.8298641322774213, "compression_loss": 92.74061584472656, "distillation_loss": 4.452943325042725, "epoch": 3.49, "learning_rate": 4.264245460237946e-05, "loss": 97.1854, "step": 4125, "task_loss": 1.9089632034301758 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8715910943814523, "compression/movement_sparsity/importance_threshold": -0.0008333924011701696, "compression/movement_sparsity/linear_layer_sparsity": 0.8594458891765953, "compression/movement_sparsity/model_sparsity": 0.8299212789185695, "compression_loss": 92.76295471191406, "distillation_loss": 4.021533966064453, "epoch": 3.49, "learning_rate": 4.26393237319975e-05, "loss": 96.8505, "step": 4126, "task_loss": 2.053182601928711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8718059447087829, "compression/movement_sparsity/importance_threshold": -0.0008319979914185741, "compression/movement_sparsity/linear_layer_sparsity": 0.8595181377083009, "compression/movement_sparsity/model_sparsity": 0.8299910454909487, "compression_loss": 92.78523254394531, "distillation_loss": 3.51828932762146, "epoch": 3.49, "learning_rate": 4.263619286161553e-05, "loss": 96.661, "step": 4127, "task_loss": 1.5712181329727173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8720205552478268, "compression/movement_sparsity/importance_threshold": -0.0008306051379276987, "compression/movement_sparsity/linear_layer_sparsity": 0.8596744277735041, "compression/movement_sparsity/model_sparsity": 0.8301419665116077, "compression_loss": 92.80746459960938, "distillation_loss": 2.7862868309020996, "epoch": 3.49, "learning_rate": 4.263306199123356e-05, "loss": 96.0541, "step": 4128, "task_loss": 1.693892002105713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8722349261324689, "compression/movement_sparsity/importance_threshold": -0.0008292138398286126, "compression/movement_sparsity/linear_layer_sparsity": 0.8597733148957084, "compression/movement_sparsity/model_sparsity": 0.8302374565569518, "compression_loss": 92.82966613769531, "distillation_loss": 4.139125347137451, "epoch": 3.49, "learning_rate": 4.26299311208516e-05, "loss": 97.1437, "step": 4129, "task_loss": 2.166367769241333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8724490574965945, "compression/movement_sparsity/importance_threshold": -0.0008278240962523798, "compression/movement_sparsity/linear_layer_sparsity": 0.8598980774616825, "compression/movement_sparsity/model_sparsity": 0.8303579331449701, "compression_loss": 92.85185241699219, "distillation_loss": 3.8545031547546387, "epoch": 3.49, "learning_rate": 4.262680025046963e-05, "loss": 96.3179, "step": 4130, "task_loss": 2.8633182048797607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8726629494740886, "compression/movement_sparsity/importance_threshold": -0.0008264359063300669, "compression/movement_sparsity/linear_layer_sparsity": 0.8599962252854932, "compression/movement_sparsity/model_sparsity": 0.830452709289095, "compression_loss": 92.87397003173828, "distillation_loss": 5.786999702453613, "epoch": 3.49, "learning_rate": 4.262366938008766e-05, "loss": 96.5771, "step": 4131, "task_loss": 4.1249823570251465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8728766021988363, "compression/movement_sparsity/importance_threshold": -0.0008250492691927415, "compression/movement_sparsity/linear_layer_sparsity": 0.8601582866478322, "compression/movement_sparsity/model_sparsity": 0.8306092033450786, "compression_loss": 92.89604949951172, "distillation_loss": 5.5101728439331055, "epoch": 3.49, "learning_rate": 4.26205385097057e-05, "loss": 96.9671, "step": 4132, "task_loss": 2.929687261581421 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8730900158047231, "compression/movement_sparsity/importance_threshold": -0.0008236641839714665, "compression/movement_sparsity/linear_layer_sparsity": 0.8602643401947856, "compression/movement_sparsity/model_sparsity": 0.8307116136264352, "compression_loss": 92.91805267333984, "distillation_loss": 4.933534622192383, "epoch": 3.49, "learning_rate": 4.261740763932373e-05, "loss": 97.1866, "step": 4133, "task_loss": 2.850085973739624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8733031904256338, "compression/movement_sparsity/importance_threshold": -0.0008222806497973105, "compression/movement_sparsity/linear_layer_sparsity": 0.8603779059673495, "compression/movement_sparsity/model_sparsity": 0.8308212780653425, "compression_loss": 92.94004821777344, "distillation_loss": 3.4735143184661865, "epoch": 3.49, "learning_rate": 4.261427676894177e-05, "loss": 96.9955, "step": 4134, "task_loss": 2.2309560775756836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8735161261954536, "compression/movement_sparsity/importance_threshold": -0.0008208986658013373, "compression/movement_sparsity/linear_layer_sparsity": 0.8604530401476231, "compression/movement_sparsity/model_sparsity": 0.8308938311553838, "compression_loss": 92.96204376220703, "distillation_loss": 4.759377479553223, "epoch": 3.5, "learning_rate": 4.2611145898559804e-05, "loss": 97.5228, "step": 4135, "task_loss": 3.2021617889404297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8737288232480677, "compression/movement_sparsity/importance_threshold": -0.0008195182311146153, "compression/movement_sparsity/linear_layer_sparsity": 0.8606337509081446, "compression/movement_sparsity/model_sparsity": 0.8310683339453501, "compression_loss": 92.98399353027344, "distillation_loss": 4.48211669921875, "epoch": 3.5, "learning_rate": 4.2608015028177835e-05, "loss": 97.51, "step": 4136, "task_loss": 2.624817132949829 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8739412817173613, "compression/movement_sparsity/importance_threshold": -0.0008181393448682085, "compression/movement_sparsity/linear_layer_sparsity": 0.8607198433984755, "compression/movement_sparsity/model_sparsity": 0.8311514688937867, "compression_loss": 93.00585174560547, "distillation_loss": 5.734959602355957, "epoch": 3.5, "learning_rate": 4.2604884157795874e-05, "loss": 97.1764, "step": 4137, "task_loss": 3.20112681388855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8741535017372193, "compression/movement_sparsity/importance_threshold": -0.0008167620061931852, "compression/movement_sparsity/linear_layer_sparsity": 0.8609623094231832, "compression/movement_sparsity/model_sparsity": 0.8313856054646329, "compression_loss": 93.0277328491211, "distillation_loss": 3.2911558151245117, "epoch": 3.5, "learning_rate": 4.2601753287413906e-05, "loss": 96.974, "step": 4138, "task_loss": 2.3615942001342773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8743654834415271, "compression/movement_sparsity/importance_threshold": -0.0008153862142206104, "compression/movement_sparsity/linear_layer_sparsity": 0.8611048389989346, "compression/movement_sparsity/model_sparsity": 0.831523238710985, "compression_loss": 93.04957580566406, "distillation_loss": 4.935049057006836, "epoch": 3.5, "learning_rate": 4.259862241703194e-05, "loss": 98.0713, "step": 4139, "task_loss": 3.062793254852295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8745772269641696, "compression/movement_sparsity/importance_threshold": -0.0008140119680815497, "compression/movement_sparsity/linear_layer_sparsity": 0.8611978117339915, "compression/movement_sparsity/model_sparsity": 0.831613017546575, "compression_loss": 93.07133483886719, "distillation_loss": 6.311256408691406, "epoch": 3.5, "learning_rate": 4.259549154664997e-05, "loss": 97.9328, "step": 4140, "task_loss": 2.999155044555664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8747887324390324, "compression/movement_sparsity/importance_threshold": -0.0008126392669070672, "compression/movement_sparsity/linear_layer_sparsity": 0.8613880141319511, "compression/movement_sparsity/model_sparsity": 0.8317966859070338, "compression_loss": 93.09317779541016, "distillation_loss": 2.9549098014831543, "epoch": 3.5, "learning_rate": 4.259236067626801e-05, "loss": 97.1398, "step": 4141, "task_loss": 2.7523841857910156 }, { "compression/movement_sparsity/importance_regularization_factor": 0.875, "compression/movement_sparsity/importance_threshold": -0.0008112681098282337, "compression/movement_sparsity/linear_layer_sparsity": 0.8615249751214167, "compression/movement_sparsity/model_sparsity": 0.8319289418651699, "compression_loss": 93.11497497558594, "distillation_loss": 4.87784481048584, "epoch": 3.5, "learning_rate": 4.258922980588604e-05, "loss": 97.4078, "step": 4142, "task_loss": 2.7173409461975098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.875211029780958, "compression/movement_sparsity/importance_threshold": -0.0008098984959761116, "compression/movement_sparsity/linear_layer_sparsity": 0.861680573584897, "compression/movement_sparsity/model_sparsity": 0.8320791950427529, "compression_loss": 93.13673400878906, "distillation_loss": 4.605889320373535, "epoch": 3.5, "learning_rate": 4.258609893550407e-05, "loss": 97.281, "step": 4143, "task_loss": 1.7068456411361694 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8754218219157913, "compression/movement_sparsity/importance_threshold": -0.0008085304244817683, "compression/movement_sparsity/linear_layer_sparsity": 0.8618098315620698, "compression/movement_sparsity/model_sparsity": 0.8322040126107657, "compression_loss": 93.15845489501953, "distillation_loss": 5.882782936096191, "epoch": 3.5, "learning_rate": 4.258296806512211e-05, "loss": 97.9349, "step": 4144, "task_loss": 2.7319395542144775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8756323765383853, "compression/movement_sparsity/importance_threshold": -0.0008071638944762687, "compression/movement_sparsity/linear_layer_sparsity": 0.8619016834253689, "compression/movement_sparsity/model_sparsity": 0.832292709079991, "compression_loss": 93.18019104003906, "distillation_loss": 4.6103596687316895, "epoch": 3.5, "learning_rate": 4.257983719474014e-05, "loss": 97.2872, "step": 4145, "task_loss": 1.8536008596420288 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8758426937826249, "compression/movement_sparsity/importance_threshold": -0.0008057989050906812, "compression/movement_sparsity/linear_layer_sparsity": 0.8620866868862392, "compression/movement_sparsity/model_sparsity": 0.8324713571028433, "compression_loss": 93.20187377929688, "distillation_loss": 4.234727382659912, "epoch": 3.5, "learning_rate": 4.2576706324358174e-05, "loss": 97.8895, "step": 4146, "task_loss": 2.467641830444336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8760527737823953, "compression/movement_sparsity/importance_threshold": -0.0008044354554560687, "compression/movement_sparsity/linear_layer_sparsity": 0.862225901543388, "compression/movement_sparsity/model_sparsity": 0.8326057893082446, "compression_loss": 93.22351837158203, "distillation_loss": 4.4289703369140625, "epoch": 3.51, "learning_rate": 4.2573575453976205e-05, "loss": 97.0559, "step": 4147, "task_loss": 2.888850450515747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8762626166715816, "compression/movement_sparsity/importance_threshold": -0.0008030735447035007, "compression/movement_sparsity/linear_layer_sparsity": 0.8624672586205054, "compression/movement_sparsity/model_sparsity": 0.8328388550272618, "compression_loss": 93.24517822265625, "distillation_loss": 4.135921478271484, "epoch": 3.51, "learning_rate": 4.2570444583594244e-05, "loss": 96.818, "step": 4148, "task_loss": 2.7057061195373535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8764722225840692, "compression/movement_sparsity/importance_threshold": -0.0008017131719640393, "compression/movement_sparsity/linear_layer_sparsity": 0.8624541897327765, "compression/movement_sparsity/model_sparsity": 0.832826235096031, "compression_loss": 93.26679229736328, "distillation_loss": 3.5068869590759277, "epoch": 3.51, "learning_rate": 4.2567313713212276e-05, "loss": 97.6825, "step": 4149, "task_loss": 1.5555847883224487 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8766815916537428, "compression/movement_sparsity/importance_threshold": -0.0008003543363687545, "compression/movement_sparsity/linear_layer_sparsity": 0.8625628665966097, "compression/movement_sparsity/model_sparsity": 0.8329311785752626, "compression_loss": 93.28838348388672, "distillation_loss": 4.378317356109619, "epoch": 3.51, "learning_rate": 4.256418284283031e-05, "loss": 97.3821, "step": 4150, "task_loss": 2.498948335647583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.876890724014488, "compression/movement_sparsity/importance_threshold": -0.0007989970370487097, "compression/movement_sparsity/linear_layer_sparsity": 0.8627770842681881, "compression/movement_sparsity/model_sparsity": 0.8331380372108115, "compression_loss": 93.3099594116211, "distillation_loss": 5.1413445472717285, "epoch": 3.51, "learning_rate": 4.256105197244834e-05, "loss": 97.715, "step": 4151, "task_loss": 2.5625040531158447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8770996198001896, "compression/movement_sparsity/importance_threshold": -0.0007976412731349721, "compression/movement_sparsity/linear_layer_sparsity": 0.8628739562060618, "compression/movement_sparsity/model_sparsity": 0.8332315812996063, "compression_loss": 93.33151245117188, "distillation_loss": 4.495284557342529, "epoch": 3.51, "learning_rate": 4.255792110206638e-05, "loss": 97.8397, "step": 4152, "task_loss": 2.5900564193725586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.877308279144733, "compression/movement_sparsity/importance_threshold": -0.0007962870437586058, "compression/movement_sparsity/linear_layer_sparsity": 0.8630358029333834, "compression/movement_sparsity/model_sparsity": 0.8333878680939456, "compression_loss": 93.35309600830078, "distillation_loss": 4.355298042297363, "epoch": 3.51, "learning_rate": 4.255479023168441e-05, "loss": 97.6605, "step": 4153, "task_loss": 3.076200246810913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8775167021820031, "compression/movement_sparsity/importance_threshold": -0.00079493434805068, "compression/movement_sparsity/linear_layer_sparsity": 0.8631580375758187, "compression/movement_sparsity/model_sparsity": 0.8335059036003755, "compression_loss": 93.37459564208984, "distillation_loss": 4.445387840270996, "epoch": 3.51, "learning_rate": 4.255165936130244e-05, "loss": 97.7314, "step": 4154, "task_loss": 2.2172040939331055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8777248890458851, "compression/movement_sparsity/importance_threshold": -0.0007935831851422588, "compression/movement_sparsity/linear_layer_sparsity": 0.8633764883269076, "compression/movement_sparsity/model_sparsity": 0.8337168498961314, "compression_loss": 93.39605712890625, "distillation_loss": 4.912913799285889, "epoch": 3.51, "learning_rate": 4.254852849092048e-05, "loss": 98.0525, "step": 4155, "task_loss": 2.9267640113830566 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8779328398702643, "compression/movement_sparsity/importance_threshold": -0.0007922335541644086, "compression/movement_sparsity/linear_layer_sparsity": 0.8635227978637997, "compression/movement_sparsity/model_sparsity": 0.8338581332503304, "compression_loss": 93.41751098632812, "distillation_loss": 4.2226152420043945, "epoch": 3.51, "learning_rate": 4.254539762053851e-05, "loss": 97.8958, "step": 4156, "task_loss": 1.7599687576293945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8781405547890256, "compression/movement_sparsity/importance_threshold": -0.0007908854542481953, "compression/movement_sparsity/linear_layer_sparsity": 0.8636756776170591, "compression/movement_sparsity/model_sparsity": 0.8340057611137522, "compression_loss": 93.43900299072266, "distillation_loss": 5.244320869445801, "epoch": 3.51, "learning_rate": 4.2542266750156544e-05, "loss": 98.1622, "step": 4157, "task_loss": 2.909557580947876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8783480339360544, "compression/movement_sparsity/importance_threshold": -0.0007895388845246846, "compression/movement_sparsity/linear_layer_sparsity": 0.8637348453368683, "compression/movement_sparsity/model_sparsity": 0.8340628962403648, "compression_loss": 93.4604263305664, "distillation_loss": 4.849429130554199, "epoch": 3.51, "learning_rate": 4.2539135879774575e-05, "loss": 98.7834, "step": 4158, "task_loss": 3.5740017890930176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8785552774452356, "compression/movement_sparsity/importance_threshold": -0.0007881938441249431, "compression/movement_sparsity/linear_layer_sparsity": 0.863919443376039, "compression/movement_sparsity/model_sparsity": 0.834241152769, "compression_loss": 93.48184967041016, "distillation_loss": 5.620876312255859, "epoch": 3.52, "learning_rate": 4.2536005009392614e-05, "loss": 97.9685, "step": 4159, "task_loss": 2.983299970626831 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8787622854504545, "compression/movement_sparsity/importance_threshold": -0.0007868503321800373, "compression/movement_sparsity/linear_layer_sparsity": 0.8640298134716766, "compression/movement_sparsity/model_sparsity": 0.8343477313123143, "compression_loss": 93.50331115722656, "distillation_loss": 4.113637924194336, "epoch": 3.52, "learning_rate": 4.2532874139010646e-05, "loss": 97.7168, "step": 4160, "task_loss": 1.3733645677566528 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8789690580855962, "compression/movement_sparsity/importance_threshold": -0.0007855083478210331, "compression/movement_sparsity/linear_layer_sparsity": 0.864110325451554, "compression/movement_sparsity/model_sparsity": 0.834425477457999, "compression_loss": 93.52473449707031, "distillation_loss": 5.526363372802734, "epoch": 3.52, "learning_rate": 4.252974326862868e-05, "loss": 97.4303, "step": 4161, "task_loss": 3.4653985500335693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8791755954845459, "compression/movement_sparsity/importance_threshold": -0.0007841678901789944, "compression/movement_sparsity/linear_layer_sparsity": 0.8642185134245141, "compression/movement_sparsity/model_sparsity": 0.8345299488412629, "compression_loss": 93.54606628417969, "distillation_loss": 5.558565616607666, "epoch": 3.52, "learning_rate": 4.252661239824671e-05, "loss": 98.6507, "step": 4162, "task_loss": 3.4733760356903076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8793818977811886, "compression/movement_sparsity/importance_threshold": -0.0007828289583849905, "compression/movement_sparsity/linear_layer_sparsity": 0.8643635232271338, "compression/movement_sparsity/model_sparsity": 0.8346699771110603, "compression_loss": 93.56746673583984, "distillation_loss": 3.870809555053711, "epoch": 3.52, "learning_rate": 4.252348152786475e-05, "loss": 98.4936, "step": 4163, "task_loss": 1.3660509586334229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8795879651094095, "compression/movement_sparsity/importance_threshold": -0.0007814915515700852, "compression/movement_sparsity/linear_layer_sparsity": 0.8644974316296846, "compression/movement_sparsity/model_sparsity": 0.834799285348033, "compression_loss": 93.58876037597656, "distillation_loss": 4.448727130889893, "epoch": 3.52, "learning_rate": 4.252035065748278e-05, "loss": 98.2958, "step": 4164, "task_loss": 2.321833610534668 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8797937976030938, "compression/movement_sparsity/importance_threshold": -0.0007801556688653453, "compression/movement_sparsity/linear_layer_sparsity": 0.8646008022389199, "compression/movement_sparsity/model_sparsity": 0.8348991048588359, "compression_loss": 93.61000061035156, "distillation_loss": 5.199078559875488, "epoch": 3.52, "learning_rate": 4.251721978710082e-05, "loss": 97.8611, "step": 4165, "task_loss": 2.8805224895477295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8799993953961265, "compression/movement_sparsity/importance_threshold": -0.0007788213094018373, "compression/movement_sparsity/linear_layer_sparsity": 0.8647163354991437, "compression/movement_sparsity/model_sparsity": 0.8350106691961492, "compression_loss": 93.63125610351562, "distillation_loss": 2.873547077178955, "epoch": 3.52, "learning_rate": 4.251408891671885e-05, "loss": 98.0483, "step": 4166, "task_loss": 1.6935880184173584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.880204758622393, "compression/movement_sparsity/importance_threshold": -0.0007774884723106262, "compression/movement_sparsity/linear_layer_sparsity": 0.864861524164278, "compression/movement_sparsity/model_sparsity": 0.8351508701839835, "compression_loss": 93.6524887084961, "distillation_loss": 3.798163414001465, "epoch": 3.52, "learning_rate": 4.251095804633689e-05, "loss": 98.2718, "step": 4167, "task_loss": 3.1979613304138184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8804098874157782, "compression/movement_sparsity/importance_threshold": -0.0007761571567227784, "compression/movement_sparsity/linear_layer_sparsity": 0.8650635672607001, "compression/movement_sparsity/model_sparsity": 0.8353459724784862, "compression_loss": 93.67366027832031, "distillation_loss": 4.858689785003662, "epoch": 3.52, "learning_rate": 4.250782717595492e-05, "loss": 98.2272, "step": 4168, "task_loss": 2.594836950302124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8806147819101673, "compression/movement_sparsity/importance_threshold": -0.0007748273617693606, "compression/movement_sparsity/linear_layer_sparsity": 0.8652160773647628, "compression/movement_sparsity/model_sparsity": 0.8354932433912984, "compression_loss": 93.69483184814453, "distillation_loss": 5.8354692459106445, "epoch": 3.52, "learning_rate": 4.250469630557295e-05, "loss": 99.1996, "step": 4169, "task_loss": 2.575535535812378 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8808194422394454, "compression/movement_sparsity/importance_threshold": -0.0007734990865814377, "compression/movement_sparsity/linear_layer_sparsity": 0.8654702171495859, "compression/movement_sparsity/model_sparsity": 0.8357386526926874, "compression_loss": 93.71590423583984, "distillation_loss": 4.481653213500977, "epoch": 3.52, "learning_rate": 4.250156543519099e-05, "loss": 97.8881, "step": 4170, "task_loss": 2.3865818977355957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8810238685374978, "compression/movement_sparsity/importance_threshold": -0.0007721723302900772, "compression/movement_sparsity/linear_layer_sparsity": 0.865538816885995, "compression/movement_sparsity/model_sparsity": 0.8358048958171135, "compression_loss": 93.73699951171875, "distillation_loss": 5.803676605224609, "epoch": 3.53, "learning_rate": 4.249843456480902e-05, "loss": 97.9774, "step": 4171, "task_loss": 3.938122510910034 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8812280609382095, "compression/movement_sparsity/importance_threshold": -0.0007708470920263439, "compression/movement_sparsity/linear_layer_sparsity": 0.8656201993301098, "compression/movement_sparsity/model_sparsity": 0.8358834825239112, "compression_loss": 93.75808715820312, "distillation_loss": 4.258127212524414, "epoch": 3.53, "learning_rate": 4.2495303694427054e-05, "loss": 98.4023, "step": 4172, "task_loss": 2.280576229095459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8814320195754657, "compression/movement_sparsity/importance_threshold": -0.0007695233709213035, "compression/movement_sparsity/linear_layer_sparsity": 0.8657329304109392, "compression/movement_sparsity/model_sparsity": 0.8359923409453127, "compression_loss": 93.77914428710938, "distillation_loss": 2.964057445526123, "epoch": 3.53, "learning_rate": 4.2492172824045086e-05, "loss": 97.7902, "step": 4173, "task_loss": 2.550966262817383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8816357445831514, "compression/movement_sparsity/importance_threshold": -0.0007682011661060244, "compression/movement_sparsity/linear_layer_sparsity": 0.8658894589594952, "compression/movement_sparsity/model_sparsity": 0.8361434922566877, "compression_loss": 93.80013275146484, "distillation_loss": 3.9454612731933594, "epoch": 3.53, "learning_rate": 4.2489041953663124e-05, "loss": 97.5605, "step": 4174, "task_loss": 3.309239149093628 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8818392360951519, "compression/movement_sparsity/importance_threshold": -0.0007668804767115706, "compression/movement_sparsity/linear_layer_sparsity": 0.8660358281172255, "compression/movement_sparsity/model_sparsity": 0.8362848331835657, "compression_loss": 93.82113647460938, "distillation_loss": 4.2325639724731445, "epoch": 3.53, "learning_rate": 4.2485911083281156e-05, "loss": 98.4594, "step": 4175, "task_loss": 1.371675729751587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8820424942453525, "compression/movement_sparsity/importance_threshold": -0.0007655613018690069, "compression/movement_sparsity/linear_layer_sparsity": 0.8661480583830142, "compression/movement_sparsity/model_sparsity": 0.8363932079944639, "compression_loss": 93.84203338623047, "distillation_loss": 5.0167365074157715, "epoch": 3.53, "learning_rate": 4.248278021289919e-05, "loss": 98.5768, "step": 4176, "task_loss": 2.1303749084472656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8822455191676379, "compression/movement_sparsity/importance_threshold": -0.0007642436407094026, "compression/movement_sparsity/linear_layer_sparsity": 0.8662644501833079, "compression/movement_sparsity/model_sparsity": 0.8365056013783544, "compression_loss": 93.86296081542969, "distillation_loss": 5.989990234375, "epoch": 3.53, "learning_rate": 4.247964934251722e-05, "loss": 99.2295, "step": 4177, "task_loss": 3.871608257293701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8824483109958936, "compression/movement_sparsity/importance_threshold": -0.0007629274923638225, "compression/movement_sparsity/linear_layer_sparsity": 0.8663851108356151, "compression/movement_sparsity/model_sparsity": 0.8366221169660594, "compression_loss": 93.88387298583984, "distillation_loss": 5.651745319366455, "epoch": 3.53, "learning_rate": 4.247651847213526e-05, "loss": 98.215, "step": 4178, "task_loss": 2.777048349380493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8826508698640047, "compression/movement_sparsity/importance_threshold": -0.0007616128559633306, "compression/movement_sparsity/linear_layer_sparsity": 0.8665617793033081, "compression/movement_sparsity/model_sparsity": 0.8367927163283913, "compression_loss": 93.90476989746094, "distillation_loss": 4.876517295837402, "epoch": 3.53, "learning_rate": 4.247338760175329e-05, "loss": 98.5171, "step": 4179, "task_loss": 2.8799307346343994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8828531959058562, "compression/movement_sparsity/importance_threshold": -0.0007602997306389954, "compression/movement_sparsity/linear_layer_sparsity": 0.8667469020058549, "compression/movement_sparsity/model_sparsity": 0.8369714794966016, "compression_loss": 93.92558288574219, "distillation_loss": 4.187796115875244, "epoch": 3.53, "learning_rate": 4.247025673137132e-05, "loss": 98.1601, "step": 4180, "task_loss": 1.7047861814498901 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8830552892553332, "compression/movement_sparsity/importance_threshold": -0.0007589881155218832, "compression/movement_sparsity/linear_layer_sparsity": 0.8669403716257468, "compression/movement_sparsity/model_sparsity": 0.8371583028398679, "compression_loss": 93.94644927978516, "distillation_loss": 3.629971981048584, "epoch": 3.53, "learning_rate": 4.246712586098936e-05, "loss": 98.4759, "step": 4181, "task_loss": 2.289292812347412 }, { "compression/movement_sparsity/importance_regularization_factor": 0.883257150046321, "compression/movement_sparsity/importance_threshold": -0.0007576780097430583, "compression/movement_sparsity/linear_layer_sparsity": 0.8670613661547478, "compression/movement_sparsity/model_sparsity": 0.8372751408345752, "compression_loss": 93.96725463867188, "distillation_loss": 3.423062801361084, "epoch": 3.53, "learning_rate": 4.246399499060739e-05, "loss": 98.2926, "step": 4182, "task_loss": 2.4275829792022705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8834587784127047, "compression/movement_sparsity/importance_threshold": -0.0007563694124335871, "compression/movement_sparsity/linear_layer_sparsity": 0.8671106606637545, "compression/movement_sparsity/model_sparsity": 0.8373227419255498, "compression_loss": 93.98806762695312, "distillation_loss": 4.624020576477051, "epoch": 3.54, "learning_rate": 4.2460864120225424e-05, "loss": 98.9935, "step": 4183, "task_loss": 2.3174972534179688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8836601744883694, "compression/movement_sparsity/importance_threshold": -0.0007550623227245363, "compression/movement_sparsity/linear_layer_sparsity": 0.867219099044235, "compression/movement_sparsity/model_sparsity": 0.8374274551140654, "compression_loss": 94.00885009765625, "distillation_loss": 5.177657604217529, "epoch": 3.54, "learning_rate": 4.2457733249843456e-05, "loss": 98.0999, "step": 4184, "task_loss": 2.3263072967529297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8838613384072003, "compression/movement_sparsity/importance_threshold": -0.0007537567397469724, "compression/movement_sparsity/linear_layer_sparsity": 0.8672766570014133, "compression/movement_sparsity/model_sparsity": 0.8374830357783456, "compression_loss": 94.02959442138672, "distillation_loss": 3.9276442527770996, "epoch": 3.54, "learning_rate": 4.2454602379461494e-05, "loss": 97.5033, "step": 4185, "task_loss": 1.3782191276550293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8840622703030825, "compression/movement_sparsity/importance_threshold": -0.0007524526626319595, "compression/movement_sparsity/linear_layer_sparsity": 0.8673598042223383, "compression/movement_sparsity/model_sparsity": 0.837563326636441, "compression_loss": 94.05027770996094, "distillation_loss": 3.985206127166748, "epoch": 3.54, "learning_rate": 4.2451471509079526e-05, "loss": 97.9413, "step": 4186, "task_loss": 2.3731460571289062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8842629703099012, "compression/movement_sparsity/importance_threshold": -0.0007511500905105651, "compression/movement_sparsity/linear_layer_sparsity": 0.8676148621680694, "compression/movement_sparsity/model_sparsity": 0.8378096225570861, "compression_loss": 94.07096099853516, "distillation_loss": 4.002029895782471, "epoch": 3.54, "learning_rate": 4.244834063869756e-05, "loss": 98.2584, "step": 4187, "task_loss": 2.047499895095825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8844634385615413, "compression/movement_sparsity/importance_threshold": -0.0007498490225138557, "compression/movement_sparsity/linear_layer_sparsity": 0.8677180777631254, "compression/movement_sparsity/model_sparsity": 0.8379092923789238, "compression_loss": 94.09156799316406, "distillation_loss": 4.830759525299072, "epoch": 3.54, "learning_rate": 4.244520976831559e-05, "loss": 97.9436, "step": 4188, "task_loss": 3.3426737785339355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8846636751918884, "compression/movement_sparsity/importance_threshold": -0.0007485494577728954, "compression/movement_sparsity/linear_layer_sparsity": 0.8677834222017698, "compression/movement_sparsity/model_sparsity": 0.8379723920350778, "compression_loss": 94.11215209960938, "distillation_loss": 4.567653656005859, "epoch": 3.54, "learning_rate": 4.244207889793363e-05, "loss": 98.3078, "step": 4189, "task_loss": 2.5932061672210693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8848636803348271, "compression/movement_sparsity/importance_threshold": -0.0007472513954187533, "compression/movement_sparsity/linear_layer_sparsity": 0.867857065861089, "compression/movement_sparsity/model_sparsity": 0.8380435058081449, "compression_loss": 94.13275146484375, "distillation_loss": 6.563697814941406, "epoch": 3.54, "learning_rate": 4.243894802755166e-05, "loss": 98.7008, "step": 4190, "task_loss": 2.69981050491333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.885063454124243, "compression/movement_sparsity/importance_threshold": -0.0007459548345824917, "compression/movement_sparsity/linear_layer_sparsity": 0.8679094248811781, "compression/movement_sparsity/model_sparsity": 0.8380940661348187, "compression_loss": 94.1533432006836, "distillation_loss": 3.8421883583068848, "epoch": 3.54, "learning_rate": 4.243581715716969e-05, "loss": 97.8713, "step": 4191, "task_loss": 2.514564037322998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.885262996694021, "compression/movement_sparsity/importance_threshold": -0.000744659774395179, "compression/movement_sparsity/linear_layer_sparsity": 0.8680501419834489, "compression/movement_sparsity/model_sparsity": 0.8382299491717301, "compression_loss": 94.17388916015625, "distillation_loss": 4.21470832824707, "epoch": 3.54, "learning_rate": 4.243268628678773e-05, "loss": 98.2616, "step": 4192, "task_loss": 2.7609574794769287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8854623081780463, "compression/movement_sparsity/importance_threshold": -0.0007433662139878809, "compression/movement_sparsity/linear_layer_sparsity": 0.8681219970176226, "compression/movement_sparsity/model_sparsity": 0.8382993357644279, "compression_loss": 94.19444274902344, "distillation_loss": 4.016214370727539, "epoch": 3.54, "learning_rate": 4.242955541640576e-05, "loss": 98.1677, "step": 4193, "task_loss": 3.0983102321624756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8856613887102041, "compression/movement_sparsity/importance_threshold": -0.0007420741524916631, "compression/movement_sparsity/linear_layer_sparsity": 0.8682213491823646, "compression/movement_sparsity/model_sparsity": 0.8383952748766681, "compression_loss": 94.21493530273438, "distillation_loss": 4.535147666931152, "epoch": 3.54, "learning_rate": 4.2426424546023794e-05, "loss": 97.7714, "step": 4194, "task_loss": 2.214716911315918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8858602384243793, "compression/movement_sparsity/importance_threshold": -0.0007407835890375922, "compression/movement_sparsity/linear_layer_sparsity": 0.8684066745957612, "compression/movement_sparsity/model_sparsity": 0.8385742337919868, "compression_loss": 94.23543548583984, "distillation_loss": 4.787393093109131, "epoch": 3.55, "learning_rate": 4.2423293675641826e-05, "loss": 98.8307, "step": 4195, "task_loss": 2.940798759460449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8860588574544573, "compression/movement_sparsity/importance_threshold": -0.000739494522756734, "compression/movement_sparsity/linear_layer_sparsity": 0.8686316001698765, "compression/movement_sparsity/model_sparsity": 0.8387914324806792, "compression_loss": 94.25587463378906, "distillation_loss": 5.206434726715088, "epoch": 3.55, "learning_rate": 4.2420162805259864e-05, "loss": 98.2159, "step": 4196, "task_loss": 2.001887559890747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8862572459343231, "compression/movement_sparsity/importance_threshold": -0.000738206952780155, "compression/movement_sparsity/linear_layer_sparsity": 0.8687661167049766, "compression/movement_sparsity/model_sparsity": 0.8389213279589774, "compression_loss": 94.2763671875, "distillation_loss": 4.382089138031006, "epoch": 3.55, "learning_rate": 4.2417031934877896e-05, "loss": 98.9836, "step": 4197, "task_loss": 2.7618191242218018 }, { "compression/movement_sparsity/importance_regularization_factor": 0.886455403997862, "compression/movement_sparsity/importance_threshold": -0.0007369208782389194, "compression/movement_sparsity/linear_layer_sparsity": 0.8689941664110125, "compression/movement_sparsity/model_sparsity": 0.8391415434560481, "compression_loss": 94.29670715332031, "distillation_loss": 4.3732147216796875, "epoch": 3.55, "learning_rate": 4.241390106449593e-05, "loss": 98.3263, "step": 4198, "task_loss": 2.3392751216888428 }, { "compression/movement_sparsity/importance_regularization_factor": 0.886653331778959, "compression/movement_sparsity/importance_threshold": -0.0007356362982640953, "compression/movement_sparsity/linear_layer_sparsity": 0.8690269101753406, "compression/movement_sparsity/model_sparsity": 0.8391731623713399, "compression_loss": 94.31715393066406, "distillation_loss": 3.3628578186035156, "epoch": 3.55, "learning_rate": 4.2410770194113966e-05, "loss": 97.5501, "step": 4199, "task_loss": 2.1115500926971436 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8868510294114992, "compression/movement_sparsity/importance_threshold": -0.0007343532119867477, "compression/movement_sparsity/linear_layer_sparsity": 0.8690968215701896, "compression/movement_sparsity/model_sparsity": 0.8392406720947033, "compression_loss": 94.3375015258789, "distillation_loss": 4.563915252685547, "epoch": 3.55, "learning_rate": 4.2407639323732e-05, "loss": 98.4079, "step": 4200, "task_loss": 2.5452628135681152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8870484970293679, "compression/movement_sparsity/importance_threshold": -0.0007330716185379423, "compression/movement_sparsity/linear_layer_sparsity": 0.8691690462535601, "compression/movement_sparsity/model_sparsity": 0.8393104156380108, "compression_loss": 94.35790252685547, "distillation_loss": 3.1495747566223145, "epoch": 3.55, "learning_rate": 4.240450845335004e-05, "loss": 97.5953, "step": 4201, "task_loss": 1.150204062461853 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8872457347664502, "compression/movement_sparsity/importance_threshold": -0.0007317915170487466, "compression/movement_sparsity/linear_layer_sparsity": 0.8692810261118284, "compression/movement_sparsity/model_sparsity": 0.8394185486436574, "compression_loss": 94.37821960449219, "distillation_loss": 4.087366580963135, "epoch": 3.55, "learning_rate": 4.240137758296807e-05, "loss": 98.6456, "step": 4202, "task_loss": 2.0196874141693115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8874427427566312, "compression/movement_sparsity/importance_threshold": -0.0007305129066502245, "compression/movement_sparsity/linear_layer_sparsity": 0.8693519510609266, "compression/movement_sparsity/model_sparsity": 0.8394870371025632, "compression_loss": 94.39851379394531, "distillation_loss": 3.3342602252960205, "epoch": 3.55, "learning_rate": 4.23982467125861e-05, "loss": 98.7869, "step": 4203, "task_loss": 1.6899839639663696 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8876395211337959, "compression/movement_sparsity/importance_threshold": -0.0007292357864734453, "compression/movement_sparsity/linear_layer_sparsity": 0.8693222241110103, "compression/movement_sparsity/model_sparsity": 0.8394583313648275, "compression_loss": 94.41873931884766, "distillation_loss": 3.2703371047973633, "epoch": 3.55, "learning_rate": 4.239511584220414e-05, "loss": 98.9999, "step": 4204, "task_loss": 1.34869384765625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8878360700318296, "compression/movement_sparsity/importance_threshold": -0.0007279601556494713, "compression/movement_sparsity/linear_layer_sparsity": 0.8694648371559353, "compression/movement_sparsity/model_sparsity": 0.8395960452129302, "compression_loss": 94.43889617919922, "distillation_loss": 4.736468315124512, "epoch": 3.55, "learning_rate": 4.239198497182217e-05, "loss": 98.7716, "step": 4205, "task_loss": 2.568250894546509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8880323895846174, "compression/movement_sparsity/importance_threshold": -0.0007266860133093717, "compression/movement_sparsity/linear_layer_sparsity": 0.8695563909150434, "compression/movement_sparsity/model_sparsity": 0.8396844538187607, "compression_loss": 94.4591064453125, "distillation_loss": 6.3802924156188965, "epoch": 3.56, "learning_rate": 4.23888541014402e-05, "loss": 99.0884, "step": 4206, "task_loss": 3.1927547454833984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8882284799260447, "compression/movement_sparsity/importance_threshold": -0.0007254133585842095, "compression/movement_sparsity/linear_layer_sparsity": 0.8696111705411627, "compression/movement_sparsity/model_sparsity": 0.8397373515962008, "compression_loss": 94.47926330566406, "distillation_loss": 3.7260985374450684, "epoch": 3.56, "learning_rate": 4.238572323105824e-05, "loss": 98.5537, "step": 4207, "task_loss": 2.740185022354126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8884243411899961, "compression/movement_sparsity/importance_threshold": -0.000724142190605054, "compression/movement_sparsity/linear_layer_sparsity": 0.8698736572433309, "compression/movement_sparsity/model_sparsity": 0.839990821072646, "compression_loss": 94.4993667602539, "distillation_loss": 4.156186103820801, "epoch": 3.56, "learning_rate": 4.238259236067627e-05, "loss": 99.0174, "step": 4208, "task_loss": 1.7430082559585571 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8886199735103572, "compression/movement_sparsity/importance_threshold": -0.0007228725085029692, "compression/movement_sparsity/linear_layer_sparsity": 0.8698931651815831, "compression/movement_sparsity/model_sparsity": 0.8400096588532059, "compression_loss": 94.51949310302734, "distillation_loss": 4.2899932861328125, "epoch": 3.56, "learning_rate": 4.2379461490294305e-05, "loss": 98.7711, "step": 4209, "task_loss": 2.7840726375579834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.888815377021013, "compression/movement_sparsity/importance_threshold": -0.0007216043114090208, "compression/movement_sparsity/linear_layer_sparsity": 0.8701024700960954, "compression/movement_sparsity/model_sparsity": 0.8402117735000074, "compression_loss": 94.53947448730469, "distillation_loss": 5.421545028686523, "epoch": 3.56, "learning_rate": 4.2376330619912336e-05, "loss": 99.2321, "step": 4210, "task_loss": 2.176476001739502 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8890105518558485, "compression/movement_sparsity/importance_threshold": -0.0007203375984542763, "compression/movement_sparsity/linear_layer_sparsity": 0.8701982450105467, "compression/movement_sparsity/model_sparsity": 0.8403042582515092, "compression_loss": 94.55961608886719, "distillation_loss": 4.2160325050354, "epoch": 3.56, "learning_rate": 4.2373199749530375e-05, "loss": 99.0586, "step": 4211, "task_loss": 2.75895094871521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.889205498148749, "compression/movement_sparsity/importance_threshold": -0.0007190723687698006, "compression/movement_sparsity/linear_layer_sparsity": 0.8702901445705162, "compression/movement_sparsity/model_sparsity": 0.8403930007788778, "compression_loss": 94.57967376708984, "distillation_loss": 5.219080448150635, "epoch": 3.56, "learning_rate": 4.237006887914841e-05, "loss": 99.2762, "step": 4212, "task_loss": 2.924978733062744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8894002160335998, "compression/movement_sparsity/importance_threshold": -0.0007178086214866602, "compression/movement_sparsity/linear_layer_sparsity": 0.870343397903178, "compression/movement_sparsity/model_sparsity": 0.8404444246957362, "compression_loss": 94.59969329833984, "distillation_loss": 4.988447189331055, "epoch": 3.56, "learning_rate": 4.236693800876644e-05, "loss": 98.4466, "step": 4213, "task_loss": 2.0805063247680664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8895947056442857, "compression/movement_sparsity/importance_threshold": -0.000716546355735921, "compression/movement_sparsity/linear_layer_sparsity": 0.870486010948103, "compression/movement_sparsity/model_sparsity": 0.8405821385438389, "compression_loss": 94.6197280883789, "distillation_loss": 3.6388604640960693, "epoch": 3.56, "learning_rate": 4.236380713838447e-05, "loss": 98.3272, "step": 4214, "task_loss": 1.9653489589691162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8897889671146918, "compression/movement_sparsity/importance_threshold": -0.0007152855706486503, "compression/movement_sparsity/linear_layer_sparsity": 0.8705604654508213, "compression/movement_sparsity/model_sparsity": 0.84065403530534, "compression_loss": 94.63971710205078, "distillation_loss": 5.854382514953613, "epoch": 3.56, "learning_rate": 4.236067626800251e-05, "loss": 100.2742, "step": 4215, "task_loss": 3.101517677307129 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8899830005787037, "compression/movement_sparsity/importance_threshold": -0.0007140262653559113, "compression/movement_sparsity/linear_layer_sparsity": 0.8706451747377064, "compression/movement_sparsity/model_sparsity": 0.8407358345676244, "compression_loss": 94.6596450805664, "distillation_loss": 4.377721786499023, "epoch": 3.56, "learning_rate": 4.235754539762054e-05, "loss": 98.9067, "step": 4216, "task_loss": 1.837020754814148 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8901768061702062, "compression/movement_sparsity/importance_threshold": -0.0007127684389887723, "compression/movement_sparsity/linear_layer_sparsity": 0.8707346775399812, "compression/movement_sparsity/model_sparsity": 0.8408222626732982, "compression_loss": 94.67961883544922, "distillation_loss": 3.4774153232574463, "epoch": 3.56, "learning_rate": 4.235441452723857e-05, "loss": 99.0772, "step": 4217, "task_loss": 1.8395946025848389 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8903703840230844, "compression/movement_sparsity/importance_threshold": -0.0007115120906782992, "compression/movement_sparsity/linear_layer_sparsity": 0.8708690867575727, "compression/movement_sparsity/model_sparsity": 0.8409520545207743, "compression_loss": 94.6995849609375, "distillation_loss": 4.786406517028809, "epoch": 3.57, "learning_rate": 4.235128365685661e-05, "loss": 98.8186, "step": 4218, "task_loss": 2.3761820793151855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8905637342712236, "compression/movement_sparsity/importance_threshold": -0.0007102572195555584, "compression/movement_sparsity/linear_layer_sparsity": 0.8709123476377556, "compression/movement_sparsity/model_sparsity": 0.8409938292566369, "compression_loss": 94.71945190429688, "distillation_loss": 4.231045246124268, "epoch": 3.57, "learning_rate": 4.234815278647464e-05, "loss": 98.6858, "step": 4219, "task_loss": 2.6288633346557617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.890756857048509, "compression/movement_sparsity/importance_threshold": -0.0007090038247516131, "compression/movement_sparsity/linear_layer_sparsity": 0.871037849502123, "compression/movement_sparsity/model_sparsity": 0.8411150197458744, "compression_loss": 94.73941040039062, "distillation_loss": 4.326460838317871, "epoch": 3.57, "learning_rate": 4.2345021916092674e-05, "loss": 98.9023, "step": 4220, "task_loss": 3.789217472076416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8909497524888255, "compression/movement_sparsity/importance_threshold": -0.0007077519053975325, "compression/movement_sparsity/linear_layer_sparsity": 0.8711209132538745, "compression/movement_sparsity/model_sparsity": 0.8411952300022192, "compression_loss": 94.75926971435547, "distillation_loss": 4.314831733703613, "epoch": 3.57, "learning_rate": 4.2341891045710706e-05, "loss": 98.8602, "step": 4221, "task_loss": 1.4241604804992676 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8911424207260584, "compression/movement_sparsity/importance_threshold": -0.0007065014606243815, "compression/movement_sparsity/linear_layer_sparsity": 0.8713319948693645, "compression/movement_sparsity/model_sparsity": 0.8413990603148541, "compression_loss": 94.77913665771484, "distillation_loss": 5.600088596343994, "epoch": 3.57, "learning_rate": 4.2338760175328745e-05, "loss": 99.2629, "step": 4222, "task_loss": 4.008947372436523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8913348618940928, "compression/movement_sparsity/importance_threshold": -0.0007052524895632259, "compression/movement_sparsity/linear_layer_sparsity": 0.8715167836952176, "compression/movement_sparsity/model_sparsity": 0.8415775010760621, "compression_loss": 94.79889678955078, "distillation_loss": 4.9017720222473145, "epoch": 3.57, "learning_rate": 4.2335629304946777e-05, "loss": 98.8645, "step": 4223, "task_loss": 2.2575430870056152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8915270761268139, "compression/movement_sparsity/importance_threshold": -0.0007040049913451321, "compression/movement_sparsity/linear_layer_sparsity": 0.8716331874196788, "compression/movement_sparsity/model_sparsity": 0.8416899059744885, "compression_loss": 94.81874084472656, "distillation_loss": 3.5778300762176514, "epoch": 3.57, "learning_rate": 4.233249843456481e-05, "loss": 98.9852, "step": 4224, "task_loss": 3.6111669540405273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8917190635581068, "compression/movement_sparsity/importance_threshold": -0.000702758965101166, "compression/movement_sparsity/linear_layer_sparsity": 0.8716816591611186, "compression/movement_sparsity/model_sparsity": 0.8417367125624933, "compression_loss": 94.83851623535156, "distillation_loss": 4.181059837341309, "epoch": 3.57, "learning_rate": 4.232936756418284e-05, "loss": 98.9759, "step": 4225, "task_loss": 1.8065448999404907 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8919108243218565, "compression/movement_sparsity/importance_threshold": -0.0007015144099623951, "compression/movement_sparsity/linear_layer_sparsity": 0.871875176477681, "compression/movement_sparsity/model_sparsity": 0.8419235819639028, "compression_loss": 94.8582763671875, "distillation_loss": 3.250277519226074, "epoch": 3.57, "learning_rate": 4.232623669380088e-05, "loss": 99.1162, "step": 4226, "task_loss": 1.7003093957901 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8921023585519484, "compression/movement_sparsity/importance_threshold": -0.0007002713250598824, "compression/movement_sparsity/linear_layer_sparsity": 0.8721208381793151, "compression/movement_sparsity/model_sparsity": 0.842160804430342, "compression_loss": 94.87804412841797, "distillation_loss": 2.4833147525787354, "epoch": 3.57, "learning_rate": 4.232310582341891e-05, "loss": 98.5236, "step": 4227, "task_loss": 2.234248161315918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8922936663822675, "compression/movement_sparsity/importance_threshold": -0.0006990297095246963, "compression/movement_sparsity/linear_layer_sparsity": 0.8721170462940069, "compression/movement_sparsity/model_sparsity": 0.8421571428079594, "compression_loss": 94.89771270751953, "distillation_loss": 3.9419474601745605, "epoch": 3.57, "learning_rate": 4.231997495303694e-05, "loss": 99.0077, "step": 4228, "task_loss": 2.0434229373931885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.892484747946699, "compression/movement_sparsity/importance_threshold": -0.0006977895624879016, "compression/movement_sparsity/linear_layer_sparsity": 0.8722984844287541, "compression/movement_sparsity/model_sparsity": 0.8423323479846091, "compression_loss": 94.91739654541016, "distillation_loss": 4.4911298751831055, "epoch": 3.57, "learning_rate": 4.231684408265498e-05, "loss": 100.2808, "step": 4229, "task_loss": 3.301196813583374 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8926756033791279, "compression/movement_sparsity/importance_threshold": -0.000696550883080565, "compression/movement_sparsity/linear_layer_sparsity": 0.872391457163811, "compression/movement_sparsity/model_sparsity": 0.8424221268201991, "compression_loss": 94.93704223632812, "distillation_loss": 7.502285480499268, "epoch": 3.58, "learning_rate": 4.231371321227301e-05, "loss": 100.6315, "step": 4230, "task_loss": 3.50606107711792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8928662328134394, "compression/movement_sparsity/importance_threshold": -0.0006953136704337531, "compression/movement_sparsity/linear_layer_sparsity": 0.8725465905847536, "compression/movement_sparsity/model_sparsity": 0.8425719309308861, "compression_loss": 94.95665740966797, "distillation_loss": 5.347997188568115, "epoch": 3.58, "learning_rate": 4.2310582341891044e-05, "loss": 99.0435, "step": 4231, "task_loss": 2.7755720615386963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8930566363835187, "compression/movement_sparsity/importance_threshold": -0.0006940779236785299, "compression/movement_sparsity/linear_layer_sparsity": 0.8725784281123413, "compression/movement_sparsity/model_sparsity": 0.8426026747414576, "compression_loss": 94.97626495361328, "distillation_loss": 3.824765205383301, "epoch": 3.58, "learning_rate": 4.230745147150908e-05, "loss": 98.7943, "step": 4232, "task_loss": 2.3735692501068115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.893246814223251, "compression/movement_sparsity/importance_threshold": -0.0006928436419459637, "compression/movement_sparsity/linear_layer_sparsity": 0.8728408432695037, "compression/movement_sparsity/model_sparsity": 0.8428560751306879, "compression_loss": 94.99581909179688, "distillation_loss": 4.5444655418396, "epoch": 3.58, "learning_rate": 4.2304320601127115e-05, "loss": 98.6556, "step": 4233, "task_loss": 1.9033831357955933 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8934367664665213, "compression/movement_sparsity/importance_threshold": -0.0006916108243671193, "compression/movement_sparsity/linear_layer_sparsity": 0.8729487212141054, "compression/movement_sparsity/model_sparsity": 0.8429602471360211, "compression_loss": 95.01527404785156, "distillation_loss": 3.109889030456543, "epoch": 3.58, "learning_rate": 4.2301189730745147e-05, "loss": 98.5796, "step": 4234, "task_loss": 2.2830705642700195 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8936264932472149, "compression/movement_sparsity/importance_threshold": -0.0006903794700730626, "compression/movement_sparsity/linear_layer_sparsity": 0.8730449611710944, "compression/movement_sparsity/model_sparsity": 0.843053180954419, "compression_loss": 95.0347671508789, "distillation_loss": 4.337301731109619, "epoch": 3.58, "learning_rate": 4.2298058860363185e-05, "loss": 99.22, "step": 4235, "task_loss": 1.7756757736206055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8938159946992167, "compression/movement_sparsity/importance_threshold": -0.0006891495781948609, "compression/movement_sparsity/linear_layer_sparsity": 0.8730919543157474, "compression/movement_sparsity/model_sparsity": 0.8430985597399852, "compression_loss": 95.05423736572266, "distillation_loss": 4.002547264099121, "epoch": 3.58, "learning_rate": 4.229492798998122e-05, "loss": 98.3891, "step": 4236, "task_loss": 2.238006591796875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8940052709564119, "compression/movement_sparsity/importance_threshold": -0.0006879211478635801, "compression/movement_sparsity/linear_layer_sparsity": 0.8732017997480089, "compression/movement_sparsity/model_sparsity": 0.8432046316437245, "compression_loss": 95.07366180419922, "distillation_loss": 3.947044849395752, "epoch": 3.58, "learning_rate": 4.2291797119599255e-05, "loss": 99.6132, "step": 4237, "task_loss": 2.079326629638672 }, { "compression/movement_sparsity/importance_regularization_factor": 0.894194322152686, "compression/movement_sparsity/importance_threshold": -0.0006866941782102832, "compression/movement_sparsity/linear_layer_sparsity": 0.8732818347611808, "compression/movement_sparsity/model_sparsity": 0.8432819172079775, "compression_loss": 95.0931167602539, "distillation_loss": 4.516515731811523, "epoch": 3.58, "learning_rate": 4.228866624921729e-05, "loss": 100.0127, "step": 4238, "task_loss": 2.4280455112457275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8943831484219236, "compression/movement_sparsity/importance_threshold": -0.0006854686683660412, "compression/movement_sparsity/linear_layer_sparsity": 0.8734966844136438, "compression/movement_sparsity/model_sparsity": 0.8434893861139235, "compression_loss": 95.11249542236328, "distillation_loss": 5.244489669799805, "epoch": 3.58, "learning_rate": 4.228553537883532e-05, "loss": 99.8337, "step": 4239, "task_loss": 3.298401355743408 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8945717498980102, "compression/movement_sparsity/importance_threshold": -0.0006842446174619163, "compression/movement_sparsity/linear_layer_sparsity": 0.8736769897524658, "compression/movement_sparsity/model_sparsity": 0.8436634974096727, "compression_loss": 95.13185119628906, "distillation_loss": 5.794568061828613, "epoch": 3.58, "learning_rate": 4.228240450845335e-05, "loss": 99.9091, "step": 4240, "task_loss": 2.981325149536133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8947601267148309, "compression/movement_sparsity/importance_threshold": -0.0006830220246289761, "compression/movement_sparsity/linear_layer_sparsity": 0.8737469249956501, "compression/movement_sparsity/model_sparsity": 0.8437310301621077, "compression_loss": 95.15123748779297, "distillation_loss": 6.471809387207031, "epoch": 3.58, "learning_rate": 4.227927363807139e-05, "loss": 100.0468, "step": 4241, "task_loss": 2.978362798690796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8949482790062707, "compression/movement_sparsity/importance_threshold": -0.0006818008889982862, "compression/movement_sparsity/linear_layer_sparsity": 0.8738727965092142, "compression/movement_sparsity/model_sparsity": 0.8438525776019549, "compression_loss": 95.17060089111328, "distillation_loss": 2.8722105026245117, "epoch": 3.59, "learning_rate": 4.227614276768942e-05, "loss": 98.906, "step": 4242, "task_loss": 2.188713312149048 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8951362069062149, "compression/movement_sparsity/importance_threshold": -0.0006805812097009124, "compression/movement_sparsity/linear_layer_sparsity": 0.8740024122114161, "compression/movement_sparsity/model_sparsity": 0.8439777406060416, "compression_loss": 95.18988800048828, "distillation_loss": 4.292150974273682, "epoch": 3.59, "learning_rate": 4.227301189730745e-05, "loss": 98.7298, "step": 4243, "task_loss": 2.535756826400757 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8953239105485484, "compression/movement_sparsity/importance_threshold": -0.0006793629858679222, "compression/movement_sparsity/linear_layer_sparsity": 0.8741046142522231, "compression/movement_sparsity/model_sparsity": 0.8440764316923366, "compression_loss": 95.20917510986328, "distillation_loss": 5.109033107757568, "epoch": 3.59, "learning_rate": 4.226988102692549e-05, "loss": 99.7878, "step": 4244, "task_loss": 3.763249397277832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8955113900671567, "compression/movement_sparsity/importance_threshold": -0.0006781462166303787, "compression/movement_sparsity/linear_layer_sparsity": 0.8740179613260133, "compression/movement_sparsity/model_sparsity": 0.8439927555607176, "compression_loss": 95.22847747802734, "distillation_loss": 4.236199378967285, "epoch": 3.59, "learning_rate": 4.226675015654352e-05, "loss": 99.6629, "step": 4245, "task_loss": 1.418630599975586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8956986455959246, "compression/movement_sparsity/importance_threshold": -0.000676930901119352, "compression/movement_sparsity/linear_layer_sparsity": 0.8741944509311916, "compression/movement_sparsity/model_sparsity": 0.8441631822050126, "compression_loss": 95.24777221679688, "distillation_loss": 3.3401546478271484, "epoch": 3.59, "learning_rate": 4.2263619286161555e-05, "loss": 99.2683, "step": 4246, "task_loss": 1.8901971578598022 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8958856772687375, "compression/movement_sparsity/importance_threshold": -0.0006757170384659042, "compression/movement_sparsity/linear_layer_sparsity": 0.8742749509869013, "compression/movement_sparsity/model_sparsity": 0.8442409168361616, "compression_loss": 95.26702117919922, "distillation_loss": 5.532265663146973, "epoch": 3.59, "learning_rate": 4.226048841577959e-05, "loss": 99.3145, "step": 4247, "task_loss": 2.4358460903167725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8960724852194804, "compression/movement_sparsity/importance_threshold": -0.0006745046278011039, "compression/movement_sparsity/linear_layer_sparsity": 0.8743380417578627, "compression/movement_sparsity/model_sparsity": 0.8443018402450505, "compression_loss": 95.28620910644531, "distillation_loss": 3.498074531555176, "epoch": 3.59, "learning_rate": 4.2257357545397625e-05, "loss": 99.1673, "step": 4248, "task_loss": 2.3442530632019043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8962590695820384, "compression/movement_sparsity/importance_threshold": -0.0006732936682560167, "compression/movement_sparsity/linear_layer_sparsity": 0.8744873442608313, "compression/movement_sparsity/model_sparsity": 0.844446013747734, "compression_loss": 95.305419921875, "distillation_loss": 4.48541259765625, "epoch": 3.59, "learning_rate": 4.225422667501566e-05, "loss": 99.6713, "step": 4249, "task_loss": 2.7768728733062744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8964454304902967, "compression/movement_sparsity/importance_threshold": -0.0006720841589617074, "compression/movement_sparsity/linear_layer_sparsity": 0.8745668665347949, "compression/movement_sparsity/model_sparsity": 0.8445228041869477, "compression_loss": 95.3246078491211, "distillation_loss": 3.3717703819274902, "epoch": 3.59, "learning_rate": 4.225109580463369e-05, "loss": 99.6287, "step": 4250, "task_loss": 1.3389135599136353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8966315680781406, "compression/movement_sparsity/importance_threshold": -0.0006708760990492427, "compression/movement_sparsity/linear_layer_sparsity": 0.8746145393570031, "compression/movement_sparsity/model_sparsity": 0.8445688393010543, "compression_loss": 95.34375762939453, "distillation_loss": 4.368921756744385, "epoch": 3.59, "learning_rate": 4.224796493425172e-05, "loss": 99.6413, "step": 4251, "task_loss": 2.3847756385803223 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8968174824794549, "compression/movement_sparsity/importance_threshold": -0.0006696694876496901, "compression/movement_sparsity/linear_layer_sparsity": 0.8747701974413217, "compression/movement_sparsity/model_sparsity": 0.8447191500513163, "compression_loss": 95.36289978027344, "distillation_loss": 4.249969959259033, "epoch": 3.59, "learning_rate": 4.224483406386976e-05, "loss": 99.6254, "step": 4252, "task_loss": 2.729175567626953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.897003173828125, "compression/movement_sparsity/importance_threshold": -0.0006684643238941135, "compression/movement_sparsity/linear_layer_sparsity": 0.8748233673048099, "compression/movement_sparsity/model_sparsity": 0.8447704933664241, "compression_loss": 95.38201141357422, "distillation_loss": 4.773062705993652, "epoch": 3.59, "learning_rate": 4.224170319348779e-05, "loss": 99.4387, "step": 4253, "task_loss": 5.511834621429443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.897188642258036, "compression/movement_sparsity/importance_threshold": -0.0006672606069135804, "compression/movement_sparsity/linear_layer_sparsity": 0.8749570849206785, "compression/movement_sparsity/model_sparsity": 0.8448996173708241, "compression_loss": 95.40106964111328, "distillation_loss": 2.5988430976867676, "epoch": 3.6, "learning_rate": 4.223857232310582e-05, "loss": 99.4857, "step": 4254, "task_loss": 0.7499180436134338 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8973738879030729, "compression/movement_sparsity/importance_threshold": -0.0006660583358391557, "compression/movement_sparsity/linear_layer_sparsity": 0.8750574744880049, "compression/movement_sparsity/model_sparsity": 0.8449965582476784, "compression_loss": 95.42012786865234, "distillation_loss": 3.3703951835632324, "epoch": 3.6, "learning_rate": 4.223544145272386e-05, "loss": 99.8327, "step": 4255, "task_loss": 1.4295319318771362 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8975589108971211, "compression/movement_sparsity/importance_threshold": -0.000664857509801906, "compression/movement_sparsity/linear_layer_sparsity": 0.8751336102983599, "compression/movement_sparsity/model_sparsity": 0.8450700785587265, "compression_loss": 95.43910217285156, "distillation_loss": 4.3338117599487305, "epoch": 3.6, "learning_rate": 4.223231058234189e-05, "loss": 99.2037, "step": 4256, "task_loss": 2.954158306121826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8977437113740656, "compression/movement_sparsity/importance_threshold": -0.000663658127932897, "compression/movement_sparsity/linear_layer_sparsity": 0.8751822251298113, "compression/movement_sparsity/model_sparsity": 0.8451170233211608, "compression_loss": 95.45804595947266, "distillation_loss": 4.749607086181641, "epoch": 3.6, "learning_rate": 4.2229179711959925e-05, "loss": 100.274, "step": 4257, "task_loss": 2.6497347354888916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8979282894677915, "compression/movement_sparsity/importance_threshold": -0.0006624601893631954, "compression/movement_sparsity/linear_layer_sparsity": 0.8753075362074966, "compression/movement_sparsity/model_sparsity": 0.8452380295778257, "compression_loss": 95.47696685791016, "distillation_loss": 5.478824138641357, "epoch": 3.6, "learning_rate": 4.222604884157796e-05, "loss": 100.3019, "step": 4258, "task_loss": 2.544741153717041 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8981126453121838, "compression/movement_sparsity/importance_threshold": -0.0006612636932238677, "compression/movement_sparsity/linear_layer_sparsity": 0.8754911206924183, "compression/movement_sparsity/model_sparsity": 0.8454153073709184, "compression_loss": 95.49594116210938, "distillation_loss": 3.3852338790893555, "epoch": 3.6, "learning_rate": 4.2222917971195995e-05, "loss": 99.4357, "step": 4259, "task_loss": 2.119915246963501 }, { "compression/movement_sparsity/importance_regularization_factor": 0.898296779041128, "compression/movement_sparsity/importance_threshold": -0.0006600686386459779, "compression/movement_sparsity/linear_layer_sparsity": 0.8754947694877149, "compression/movement_sparsity/model_sparsity": 0.8454188308188715, "compression_loss": 95.5147933959961, "distillation_loss": 2.999950408935547, "epoch": 3.6, "learning_rate": 4.221978710081403e-05, "loss": 99.8205, "step": 4260, "task_loss": 2.178978443145752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.898480690788509, "compression/movement_sparsity/importance_threshold": -0.0006588750247605944, "compression/movement_sparsity/linear_layer_sparsity": 0.875637668712663, "compression/movement_sparsity/model_sparsity": 0.8455568210158333, "compression_loss": 95.53369140625, "distillation_loss": 3.670858860015869, "epoch": 3.6, "learning_rate": 4.221665623043206e-05, "loss": 99.9322, "step": 4261, "task_loss": 2.033389091491699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.898664380688212, "compression/movement_sparsity/importance_threshold": -0.000657682850698782, "compression/movement_sparsity/linear_layer_sparsity": 0.8757606188051564, "compression/movement_sparsity/model_sparsity": 0.8456755473944109, "compression_loss": 95.5525894165039, "distillation_loss": 4.820476531982422, "epoch": 3.6, "learning_rate": 4.221352536005009e-05, "loss": 99.5578, "step": 4262, "task_loss": 2.1456522941589355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8988478488741221, "compression/movement_sparsity/importance_threshold": -0.0006564921155916064, "compression/movement_sparsity/linear_layer_sparsity": 0.8758923927817004, "compression/movement_sparsity/model_sparsity": 0.8458027945294763, "compression_loss": 95.57147216796875, "distillation_loss": 3.805536985397339, "epoch": 3.6, "learning_rate": 4.221039448966813e-05, "loss": 100.1049, "step": 4263, "task_loss": 2.066575527191162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8990310954801245, "compression/movement_sparsity/importance_threshold": -0.0006553028185701344, "compression/movement_sparsity/linear_layer_sparsity": 0.8759947259883514, "compression/movement_sparsity/model_sparsity": 0.8459016122756651, "compression_loss": 95.59026336669922, "distillation_loss": 4.488194465637207, "epoch": 3.6, "learning_rate": 4.220726361928616e-05, "loss": 99.9044, "step": 4264, "task_loss": 1.6095024347305298 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8992141206401043, "compression/movement_sparsity/importance_threshold": -0.0006541149587654315, "compression/movement_sparsity/linear_layer_sparsity": 0.8760412302421312, "compression/movement_sparsity/model_sparsity": 0.8459465189652637, "compression_loss": 95.60910034179688, "distillation_loss": 5.335603713989258, "epoch": 3.6, "learning_rate": 4.220413274890419e-05, "loss": 100.6506, "step": 4265, "task_loss": 2.5768790245056152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8993969244879467, "compression/movement_sparsity/importance_threshold": -0.0006529285353085645, "compression/movement_sparsity/linear_layer_sparsity": 0.8761004098861079, "compression/movement_sparsity/model_sparsity": 0.846003665606412, "compression_loss": 95.62785339355469, "distillation_loss": 3.387989044189453, "epoch": 3.61, "learning_rate": 4.220100187852223e-05, "loss": 99.4938, "step": 4266, "task_loss": 1.1762162446975708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8995795071575368, "compression/movement_sparsity/importance_threshold": -0.000651743547330599, "compression/movement_sparsity/linear_layer_sparsity": 0.8761958032271949, "compression/movement_sparsity/model_sparsity": 0.8460957818927684, "compression_loss": 95.6465835571289, "distillation_loss": 4.01251220703125, "epoch": 3.61, "learning_rate": 4.219787100814026e-05, "loss": 100.2394, "step": 4267, "task_loss": 3.469855308532715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8997618687827597, "compression/movement_sparsity/importance_threshold": -0.0006505599939625999, "compression/movement_sparsity/linear_layer_sparsity": 0.8762651184136622, "compression/movement_sparsity/model_sparsity": 0.846162715889342, "compression_loss": 95.66527557373047, "distillation_loss": 4.632805824279785, "epoch": 3.61, "learning_rate": 4.21947401377583e-05, "loss": 100.4697, "step": 4268, "task_loss": 2.695558547973633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.8999440094975006, "compression/movement_sparsity/importance_threshold": -0.0006493778743356347, "compression/movement_sparsity/linear_layer_sparsity": 0.8763707903872512, "compression/movement_sparsity/model_sparsity": 0.8462647577055533, "compression_loss": 95.68401336669922, "distillation_loss": 7.498427391052246, "epoch": 3.61, "learning_rate": 4.2191609267376333e-05, "loss": 100.9716, "step": 4269, "task_loss": 3.4921481609344482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9001259294356445, "compression/movement_sparsity/importance_threshold": -0.0006481971875807709, "compression/movement_sparsity/linear_layer_sparsity": 0.8764288372353025, "compression/movement_sparsity/model_sparsity": 0.8463208104658011, "compression_loss": 95.70269012451172, "distillation_loss": 4.223951816558838, "epoch": 3.61, "learning_rate": 4.218847839699437e-05, "loss": 99.8102, "step": 4270, "task_loss": 1.8262317180633545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9003076287310768, "compression/movement_sparsity/importance_threshold": -0.0006470179328290707, "compression/movement_sparsity/linear_layer_sparsity": 0.8765577971082844, "compression/movement_sparsity/model_sparsity": 0.8464453401704191, "compression_loss": 95.72134399414062, "distillation_loss": 5.032177925109863, "epoch": 3.61, "learning_rate": 4.2185347526612404e-05, "loss": 100.7993, "step": 4271, "task_loss": 3.08632755279541 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9004891075176825, "compression/movement_sparsity/importance_threshold": -0.0006458401092116024, "compression/movement_sparsity/linear_layer_sparsity": 0.876710605316538, "compression/movement_sparsity/model_sparsity": 0.846592898946626, "compression_loss": 95.739990234375, "distillation_loss": 3.521442413330078, "epoch": 3.61, "learning_rate": 4.2182216656230436e-05, "loss": 99.8942, "step": 4272, "task_loss": 2.7971725463867188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9006703659293467, "compression/movement_sparsity/importance_threshold": -0.0006446637158594327, "compression/movement_sparsity/linear_layer_sparsity": 0.8768504877270743, "compression/movement_sparsity/model_sparsity": 0.8467279759660319, "compression_loss": 95.75856018066406, "distillation_loss": 3.2244911193847656, "epoch": 3.61, "learning_rate": 4.217908578584847e-05, "loss": 99.7082, "step": 4273, "task_loss": 1.8379762172698975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9008514040999546, "compression/movement_sparsity/importance_threshold": -0.0006434887519036255, "compression/movement_sparsity/linear_layer_sparsity": 0.8768959665024375, "compression/movement_sparsity/model_sparsity": 0.8467718924055522, "compression_loss": 95.77713012695312, "distillation_loss": 4.491857051849365, "epoch": 3.61, "learning_rate": 4.2175954915466506e-05, "loss": 100.425, "step": 4274, "task_loss": 2.762059450149536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9010322221633913, "compression/movement_sparsity/importance_threshold": -0.0006423152164752493, "compression/movement_sparsity/linear_layer_sparsity": 0.8770020319735585, "compression/movement_sparsity/model_sparsity": 0.8468743142014447, "compression_loss": 95.7956771850586, "distillation_loss": 4.882683753967285, "epoch": 3.61, "learning_rate": 4.217282404508454e-05, "loss": 100.402, "step": 4275, "task_loss": 3.360321283340454 }, { "compression/movement_sparsity/importance_regularization_factor": 0.901212820253542, "compression/movement_sparsity/importance_threshold": -0.0006411431087053679, "compression/movement_sparsity/linear_layer_sparsity": 0.8771642245017415, "compression/movement_sparsity/model_sparsity": 0.847030934917322, "compression_loss": 95.81420135498047, "distillation_loss": 5.185707092285156, "epoch": 3.61, "learning_rate": 4.216969317470257e-05, "loss": 100.3494, "step": 4276, "task_loss": 2.0263848304748535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9013931985042917, "compression/movement_sparsity/importance_threshold": -0.0006399724277250489, "compression/movement_sparsity/linear_layer_sparsity": 0.8773165199707867, "compression/movement_sparsity/model_sparsity": 0.8471779985684899, "compression_loss": 95.83268737792969, "distillation_loss": 4.471525192260742, "epoch": 3.61, "learning_rate": 4.21665623043206e-05, "loss": 99.4344, "step": 4277, "task_loss": 1.5072228908538818 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9015733570495258, "compression/movement_sparsity/importance_threshold": -0.0006388031726653571, "compression/movement_sparsity/linear_layer_sparsity": 0.8774695666623931, "compression/movement_sparsity/model_sparsity": 0.8473257876354128, "compression_loss": 95.85120391845703, "distillation_loss": 2.734431028366089, "epoch": 3.62, "learning_rate": 4.216343143393864e-05, "loss": 99.3786, "step": 4278, "task_loss": 1.6592931747436523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9017532960231291, "compression/movement_sparsity/importance_threshold": -0.0006376353426573601, "compression/movement_sparsity/linear_layer_sparsity": 0.8776750796915972, "compression/movement_sparsity/model_sparsity": 0.8475242406598317, "compression_loss": 95.86955261230469, "distillation_loss": 4.0735673904418945, "epoch": 3.62, "learning_rate": 4.216030056355667e-05, "loss": 99.8613, "step": 4279, "task_loss": 2.1526927947998047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9019330155589871, "compression/movement_sparsity/importance_threshold": -0.0006364689368321226, "compression/movement_sparsity/linear_layer_sparsity": 0.877724660380627, "compression/movement_sparsity/model_sparsity": 0.8475721180996654, "compression_loss": 95.88796997070312, "distillation_loss": 4.31545352935791, "epoch": 3.62, "learning_rate": 4.21571696931747e-05, "loss": 100.4589, "step": 4280, "task_loss": 1.4806214570999146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9021125157909847, "compression/movement_sparsity/importance_threshold": -0.0006353039543207113, "compression/movement_sparsity/linear_layer_sparsity": 0.8779002437490652, "compression/movement_sparsity/model_sparsity": 0.84774166963924, "compression_loss": 95.90640258789062, "distillation_loss": 5.853517532348633, "epoch": 3.62, "learning_rate": 4.215403882279274e-05, "loss": 100.2337, "step": 4281, "task_loss": 2.7003841400146484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9022917968530071, "compression/movement_sparsity/importance_threshold": -0.0006341403942541918, "compression/movement_sparsity/linear_layer_sparsity": 0.8779514938215641, "compression/movement_sparsity/model_sparsity": 0.847791159114085, "compression_loss": 95.92474365234375, "distillation_loss": 5.2670135498046875, "epoch": 3.62, "learning_rate": 4.2150907952410774e-05, "loss": 100.6998, "step": 4282, "task_loss": 2.2572126388549805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9024708588789395, "compression/movement_sparsity/importance_threshold": -0.0006329782557636301, "compression/movement_sparsity/linear_layer_sparsity": 0.8780523126589254, "compression/movement_sparsity/model_sparsity": 0.8478885145142279, "compression_loss": 95.94304656982422, "distillation_loss": 4.087342262268066, "epoch": 3.62, "learning_rate": 4.2147777082028805e-05, "loss": 99.9375, "step": 4283, "task_loss": 1.7885308265686035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9026497020026669, "compression/movement_sparsity/importance_threshold": -0.0006318175379800934, "compression/movement_sparsity/linear_layer_sparsity": 0.8781243823281165, "compression/movement_sparsity/model_sparsity": 0.84795810836857, "compression_loss": 95.96129608154297, "distillation_loss": 4.712551116943359, "epoch": 3.62, "learning_rate": 4.214464621164684e-05, "loss": 99.9999, "step": 4284, "task_loss": 3.1486480236053467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9028283263580746, "compression/movement_sparsity/importance_threshold": -0.0006306582400346459, "compression/movement_sparsity/linear_layer_sparsity": 0.8781676193599641, "compression/movement_sparsity/model_sparsity": 0.847999860075361, "compression_loss": 95.97956848144531, "distillation_loss": 3.4320778846740723, "epoch": 3.62, "learning_rate": 4.2141515341264876e-05, "loss": 99.3783, "step": 4285, "task_loss": 1.830579400062561 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9030067320790476, "compression/movement_sparsity/importance_threshold": -0.0006295003610583558, "compression/movement_sparsity/linear_layer_sparsity": 0.878328571774713, "compression/movement_sparsity/model_sparsity": 0.8481552832795157, "compression_loss": 95.99778747558594, "distillation_loss": 3.3862199783325195, "epoch": 3.62, "learning_rate": 4.213838447088291e-05, "loss": 99.8403, "step": 4286, "task_loss": 2.0016465187072754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9031849192994712, "compression/movement_sparsity/importance_threshold": -0.0006283439001822872, "compression/movement_sparsity/linear_layer_sparsity": 0.878425109835893, "compression/movement_sparsity/model_sparsity": 0.8482485049613083, "compression_loss": 96.0159912109375, "distillation_loss": 3.9049577713012695, "epoch": 3.62, "learning_rate": 4.213525360050094e-05, "loss": 100.3118, "step": 4287, "task_loss": 2.112264633178711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9033628881532304, "compression/movement_sparsity/importance_threshold": -0.0006271888565375075, "compression/movement_sparsity/linear_layer_sparsity": 0.8784745832074141, "compression/movement_sparsity/model_sparsity": 0.8482962787703199, "compression_loss": 96.03413391113281, "distillation_loss": 5.00982666015625, "epoch": 3.62, "learning_rate": 4.213212273011897e-05, "loss": 99.9274, "step": 4288, "task_loss": 2.3987033367156982 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9035406387742103, "compression/movement_sparsity/importance_threshold": -0.0006260352292550815, "compression/movement_sparsity/linear_layer_sparsity": 0.8785874812265905, "compression/movement_sparsity/model_sparsity": 0.8484052983952226, "compression_loss": 96.0523681640625, "distillation_loss": 3.795233726501465, "epoch": 3.63, "learning_rate": 4.212899185973701e-05, "loss": 100.595, "step": 4289, "task_loss": 2.6020684242248535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9037181712962963, "compression/movement_sparsity/importance_threshold": -0.000624883017466076, "compression/movement_sparsity/linear_layer_sparsity": 0.878697326658852, "compression/movement_sparsity/model_sparsity": 0.848511370298962, "compression_loss": 96.07058715820312, "distillation_loss": 4.493381023406982, "epoch": 3.63, "learning_rate": 4.212586098935504e-05, "loss": 99.7688, "step": 4290, "task_loss": 2.3481900691986084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9038954858533733, "compression/movement_sparsity/importance_threshold": -0.0006237322203015566, "compression/movement_sparsity/linear_layer_sparsity": 0.8788035948408228, "compression/movement_sparsity/model_sparsity": 0.8486139878419628, "compression_loss": 96.08873748779297, "distillation_loss": 4.990851879119873, "epoch": 3.63, "learning_rate": 4.212273011897307e-05, "loss": 100.5513, "step": 4291, "task_loss": 2.3577311038970947 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9040725825793265, "compression/movement_sparsity/importance_threshold": -0.00062258283689259, "compression/movement_sparsity/linear_layer_sparsity": 0.8789344387322912, "compression/movement_sparsity/model_sparsity": 0.8487403368432364, "compression_loss": 96.10690307617188, "distillation_loss": 3.852588176727295, "epoch": 3.63, "learning_rate": 4.211959924859111e-05, "loss": 100.2185, "step": 4292, "task_loss": 2.1109354496002197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9042494616080411, "compression/movement_sparsity/importance_threshold": -0.0006214348663702417, "compression/movement_sparsity/linear_layer_sparsity": 0.8789860584539868, "compression/movement_sparsity/model_sparsity": 0.8487901832686909, "compression_loss": 96.12506866455078, "distillation_loss": 4.9056572914123535, "epoch": 3.63, "learning_rate": 4.2116468378209144e-05, "loss": 100.6182, "step": 4293, "task_loss": 3.6781704425811768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9044261230734021, "compression/movement_sparsity/importance_threshold": -0.0006202883078655786, "compression/movement_sparsity/linear_layer_sparsity": 0.878936179660766, "compression/movement_sparsity/model_sparsity": 0.8487420179654623, "compression_loss": 96.1431655883789, "distillation_loss": 5.2501020431518555, "epoch": 3.63, "learning_rate": 4.2113337507827175e-05, "loss": 100.4118, "step": 4294, "task_loss": 2.4796574115753174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9046025671092948, "compression/movement_sparsity/importance_threshold": -0.0006191431605096654, "compression/movement_sparsity/linear_layer_sparsity": 0.8790082254816219, "compression/movement_sparsity/model_sparsity": 0.8488115887907329, "compression_loss": 96.16131591796875, "distillation_loss": 3.7623085975646973, "epoch": 3.63, "learning_rate": 4.211020663744521e-05, "loss": 100.1041, "step": 4295, "task_loss": 1.4153671264648438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9047787938496042, "compression/movement_sparsity/importance_threshold": -0.0006179994234335687, "compression/movement_sparsity/linear_layer_sparsity": 0.8790934475077155, "compression/movement_sparsity/model_sparsity": 0.8488938831780566, "compression_loss": 96.17938995361328, "distillation_loss": 3.4800171852111816, "epoch": 3.63, "learning_rate": 4.2107075767063246e-05, "loss": 100.5117, "step": 4296, "task_loss": 1.7151365280151367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9049548034282157, "compression/movement_sparsity/importance_threshold": -0.0006168570957683552, "compression/movement_sparsity/linear_layer_sparsity": 0.8791710499906896, "compression/movement_sparsity/model_sparsity": 0.8489688197770074, "compression_loss": 96.19744873046875, "distillation_loss": 4.788797378540039, "epoch": 3.63, "learning_rate": 4.210394489668128e-05, "loss": 99.6073, "step": 4297, "task_loss": 2.1019344329833984 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9051305959790141, "compression/movement_sparsity/importance_threshold": -0.0006157161766450897, "compression/movement_sparsity/linear_layer_sparsity": 0.8791784310504562, "compression/movement_sparsity/model_sparsity": 0.8489759472746643, "compression_loss": 96.21549224853516, "distillation_loss": 3.334200143814087, "epoch": 3.63, "learning_rate": 4.210081402629931e-05, "loss": 100.7063, "step": 4298, "task_loss": 1.9109152555465698 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9053061716358847, "compression/movement_sparsity/importance_threshold": -0.0006145766651948405, "compression/movement_sparsity/linear_layer_sparsity": 0.8792916987188293, "compression/movement_sparsity/model_sparsity": 0.8490853238501765, "compression_loss": 96.23358154296875, "distillation_loss": 6.146849155426025, "epoch": 3.63, "learning_rate": 4.209768315591735e-05, "loss": 100.5551, "step": 4299, "task_loss": 2.598219156265259 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9054815305327127, "compression/movement_sparsity/importance_threshold": -0.00061343856054867, "compression/movement_sparsity/linear_layer_sparsity": 0.8793958920956314, "compression/movement_sparsity/model_sparsity": 0.8491859378639492, "compression_loss": 96.25151824951172, "distillation_loss": 4.260625839233398, "epoch": 3.63, "learning_rate": 4.209455228553538e-05, "loss": 100.3798, "step": 4300, "task_loss": 2.2677230834960938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9056566728033831, "compression/movement_sparsity/importance_threshold": -0.0006123018618376481, "compression/movement_sparsity/linear_layer_sparsity": 0.8794423009560702, "compression/movement_sparsity/model_sparsity": 0.8492307524372616, "compression_loss": 96.26953125, "distillation_loss": 5.2991108894348145, "epoch": 3.64, "learning_rate": 4.209142141515341e-05, "loss": 100.5922, "step": 4301, "task_loss": 3.0424745082855225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9058315985817812, "compression/movement_sparsity/importance_threshold": -0.0006111665681928381, "compression/movement_sparsity/linear_layer_sparsity": 0.8795368119087521, "compression/movement_sparsity/model_sparsity": 0.8493220166479691, "compression_loss": 96.28750610351562, "distillation_loss": 4.7527008056640625, "epoch": 3.64, "learning_rate": 4.208829054477145e-05, "loss": 100.517, "step": 4302, "task_loss": 3.907273292541504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.906006308001792, "compression/movement_sparsity/importance_threshold": -0.0006100326787453064, "compression/movement_sparsity/linear_layer_sparsity": 0.8796589988545167, "compression/movement_sparsity/model_sparsity": 0.8494400060962558, "compression_loss": 96.30542755126953, "distillation_loss": 3.7930173873901367, "epoch": 3.64, "learning_rate": 4.208515967438948e-05, "loss": 100.5231, "step": 4303, "task_loss": 1.9624422788619995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9061808011973007, "compression/movement_sparsity/importance_threshold": -0.0006089001926261215, "compression/movement_sparsity/linear_layer_sparsity": 0.8797338707031024, "compression/movement_sparsity/model_sparsity": 0.8495123058665097, "compression_loss": 96.32332611083984, "distillation_loss": 3.6561367511749268, "epoch": 3.64, "learning_rate": 4.208202880400752e-05, "loss": 100.2967, "step": 4304, "task_loss": 2.3792808055877686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9063550783021925, "compression/movement_sparsity/importance_threshold": -0.0006077691089663457, "compression/movement_sparsity/linear_layer_sparsity": 0.8797886861017244, "compression/movement_sparsity/model_sparsity": 0.8495652381875572, "compression_loss": 96.34119415283203, "distillation_loss": 3.9473748207092285, "epoch": 3.64, "learning_rate": 4.207889793362555e-05, "loss": 99.933, "step": 4305, "task_loss": 2.664632558822632 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9065291394503525, "compression/movement_sparsity/importance_threshold": -0.0006066394268970472, "compression/movement_sparsity/linear_layer_sparsity": 0.8798451231871449, "compression/movement_sparsity/model_sparsity": 0.8496197364854728, "compression_loss": 96.35900115966797, "distillation_loss": 2.577601671218872, "epoch": 3.64, "learning_rate": 4.2075767063243584e-05, "loss": 100.0795, "step": 4306, "task_loss": 1.3493257761001587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9067029847756658, "compression/movement_sparsity/importance_threshold": -0.0006055111455492927, "compression/movement_sparsity/linear_layer_sparsity": 0.880014553685083, "compression/movement_sparsity/model_sparsity": 0.8497833465245774, "compression_loss": 96.37677001953125, "distillation_loss": 4.209787368774414, "epoch": 3.64, "learning_rate": 4.207263619286162e-05, "loss": 100.4291, "step": 4307, "task_loss": 2.704225778579712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9068766144120175, "compression/movement_sparsity/importance_threshold": -0.000604384264054147, "compression/movement_sparsity/linear_layer_sparsity": 0.8801728112379461, "compression/movement_sparsity/model_sparsity": 0.8499361674436426, "compression_loss": 96.3945083618164, "distillation_loss": 4.990023612976074, "epoch": 3.64, "learning_rate": 4.2069505322479654e-05, "loss": 101.0574, "step": 4308, "task_loss": 2.2152626514434814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9070500284932929, "compression/movement_sparsity/importance_threshold": -0.0006032587815426758, "compression/movement_sparsity/linear_layer_sparsity": 0.8802282347691176, "compression/movement_sparsity/model_sparsity": 0.8499896870060156, "compression_loss": 96.41231536865234, "distillation_loss": 3.7624025344848633, "epoch": 3.64, "learning_rate": 4.2066374452097686e-05, "loss": 100.7277, "step": 4309, "task_loss": 2.496370315551758 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9072232271533769, "compression/movement_sparsity/importance_threshold": -0.0006021346971459467, "compression/movement_sparsity/linear_layer_sparsity": 0.880426009013526, "compression/movement_sparsity/model_sparsity": 0.8501806670967039, "compression_loss": 96.42992401123047, "distillation_loss": 5.317628383636475, "epoch": 3.64, "learning_rate": 4.206324358171572e-05, "loss": 100.5187, "step": 4310, "task_loss": 2.4477925300598145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9073962105261548, "compression/movement_sparsity/importance_threshold": -0.0006010120099950245, "compression/movement_sparsity/linear_layer_sparsity": 0.8805420907854611, "compression/movement_sparsity/model_sparsity": 0.8502927611026637, "compression_loss": 96.44759368896484, "distillation_loss": 4.242101669311523, "epoch": 3.64, "learning_rate": 4.2060112711333756e-05, "loss": 101.108, "step": 4311, "task_loss": 2.3674464225769043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9075689787455119, "compression/movement_sparsity/importance_threshold": -0.0005998907192209749, "compression/movement_sparsity/linear_layer_sparsity": 0.8806176542357697, "compression/movement_sparsity/model_sparsity": 0.8503657287159937, "compression_loss": 96.46527099609375, "distillation_loss": 3.8948607444763184, "epoch": 3.64, "learning_rate": 4.205698184095179e-05, "loss": 100.3901, "step": 4312, "task_loss": 2.1207435131073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.907741531945333, "compression/movement_sparsity/importance_threshold": -0.0005987708239548655, "compression/movement_sparsity/linear_layer_sparsity": 0.8807086475589989, "compression/movement_sparsity/model_sparsity": 0.8504535961386419, "compression_loss": 96.48290252685547, "distillation_loss": 4.038198947906494, "epoch": 3.65, "learning_rate": 4.205385097056982e-05, "loss": 100.449, "step": 4313, "task_loss": 2.1432290077209473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9079138702595035, "compression/movement_sparsity/importance_threshold": -0.0005976523233277601, "compression/movement_sparsity/linear_layer_sparsity": 0.8808054360276992, "compression/movement_sparsity/model_sparsity": 0.8505470596256861, "compression_loss": 96.50049591064453, "distillation_loss": 4.052416801452637, "epoch": 3.65, "learning_rate": 4.205072010018785e-05, "loss": 100.3637, "step": 4314, "task_loss": 3.251610517501831 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9080859938219084, "compression/movement_sparsity/importance_threshold": -0.0005965352164707281, "compression/movement_sparsity/linear_layer_sparsity": 0.8810362521406265, "compression/movement_sparsity/model_sparsity": 0.8507699464950611, "compression_loss": 96.51801300048828, "distillation_loss": 4.785122394561768, "epoch": 3.65, "learning_rate": 4.204758922980589e-05, "loss": 101.2388, "step": 4315, "task_loss": 2.4968762397766113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9082579027664329, "compression/movement_sparsity/importance_threshold": -0.0005954195025148316, "compression/movement_sparsity/linear_layer_sparsity": 0.8811603350290453, "compression/movement_sparsity/model_sparsity": 0.8508897667545391, "compression_loss": 96.53549194335938, "distillation_loss": 4.631425857543945, "epoch": 3.65, "learning_rate": 4.204445835942392e-05, "loss": 101.1321, "step": 4316, "task_loss": 2.849801778793335 }, { "compression/movement_sparsity/importance_regularization_factor": 0.908429597226962, "compression/movement_sparsity/importance_threshold": -0.0005943051805911408, "compression/movement_sparsity/linear_layer_sparsity": 0.8812971052318287, "compression/movement_sparsity/model_sparsity": 0.8510218384801025, "compression_loss": 96.55299377441406, "distillation_loss": 6.467907905578613, "epoch": 3.65, "learning_rate": 4.2041327489041954e-05, "loss": 101.2315, "step": 4317, "task_loss": 3.5941951274871826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9086010773373813, "compression/movement_sparsity/importance_threshold": -0.0005931922498307171, "compression/movement_sparsity/linear_layer_sparsity": 0.8813168635776013, "compression/movement_sparsity/model_sparsity": 0.851040918065914, "compression_loss": 96.57044982910156, "distillation_loss": 4.281374931335449, "epoch": 3.65, "learning_rate": 4.203819661865999e-05, "loss": 100.732, "step": 4318, "task_loss": 2.8736584186553955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9087723432315753, "compression/movement_sparsity/importance_threshold": -0.0005920807093646304, "compression/movement_sparsity/linear_layer_sparsity": 0.8814327426386867, "compression/movement_sparsity/model_sparsity": 0.8511528163247654, "compression_loss": 96.587890625, "distillation_loss": 3.8908591270446777, "epoch": 3.65, "learning_rate": 4.2035065748278024e-05, "loss": 101.0879, "step": 4319, "task_loss": 1.1192673444747925 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9089433950434296, "compression/movement_sparsity/importance_threshold": -0.0005909705583239449, "compression/movement_sparsity/linear_layer_sparsity": 0.8815244394878065, "compression/movement_sparsity/model_sparsity": 0.8512413631050254, "compression_loss": 96.6053237915039, "distillation_loss": 4.841855525970459, "epoch": 3.65, "learning_rate": 4.2031934877896056e-05, "loss": 100.7316, "step": 4320, "task_loss": 2.419362783432007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9091142329068291, "compression/movement_sparsity/importance_threshold": -0.0005898617958397263, "compression/movement_sparsity/linear_layer_sparsity": 0.8816983177002726, "compression/movement_sparsity/model_sparsity": 0.8514092680659814, "compression_loss": 96.622802734375, "distillation_loss": 4.2830424308776855, "epoch": 3.65, "learning_rate": 4.202880400751409e-05, "loss": 100.4289, "step": 4321, "task_loss": 2.207615375518799 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9092848569556591, "compression/movement_sparsity/importance_threshold": -0.0005887544210430428, "compression/movement_sparsity/linear_layer_sparsity": 0.8818069110949324, "compression/movement_sparsity/model_sparsity": 0.8515141309434623, "compression_loss": 96.64022827148438, "distillation_loss": 4.132948875427246, "epoch": 3.65, "learning_rate": 4.2025673137132126e-05, "loss": 102.114, "step": 4322, "task_loss": 1.2385127544403076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9094552673238047, "compression/movement_sparsity/importance_threshold": -0.0005876484330649577, "compression/movement_sparsity/linear_layer_sparsity": 0.8818753796654976, "compression/movement_sparsity/model_sparsity": 0.8515802474079945, "compression_loss": 96.65766906738281, "distillation_loss": 4.800093650817871, "epoch": 3.65, "learning_rate": 4.202254226675016e-05, "loss": 100.5375, "step": 4323, "task_loss": 2.6972572803497314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9096254641451509, "compression/movement_sparsity/importance_threshold": -0.0005865438310365392, "compression/movement_sparsity/linear_layer_sparsity": 0.8819471035338272, "compression/movement_sparsity/model_sparsity": 0.8516495073407987, "compression_loss": 96.67507934570312, "distillation_loss": 4.552634239196777, "epoch": 3.65, "learning_rate": 4.201941139636819e-05, "loss": 101.4338, "step": 4324, "task_loss": 2.5775485038757324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.909795447553583, "compression/movement_sparsity/importance_threshold": -0.0005854406140888523, "compression/movement_sparsity/linear_layer_sparsity": 0.8819989378905403, "compression/movement_sparsity/model_sparsity": 0.8516995610278976, "compression_loss": 96.69246673583984, "distillation_loss": 5.765250205993652, "epoch": 3.66, "learning_rate": 4.201628052598622e-05, "loss": 101.5823, "step": 4325, "task_loss": 2.454050064086914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9099652176829861, "compression/movement_sparsity/importance_threshold": -0.0005843387813529626, "compression/movement_sparsity/linear_layer_sparsity": 0.8821674263792351, "compression/movement_sparsity/model_sparsity": 0.8518622614186744, "compression_loss": 96.70987701416016, "distillation_loss": 6.688716411590576, "epoch": 3.66, "learning_rate": 4.201314965560426e-05, "loss": 101.2711, "step": 4326, "task_loss": 3.3912973403930664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9101347746672453, "compression/movement_sparsity/importance_threshold": -0.0005832383319599376, "compression/movement_sparsity/linear_layer_sparsity": 0.8821816161387218, "compression/movement_sparsity/model_sparsity": 0.85187596371627, "compression_loss": 96.7271728515625, "distillation_loss": 4.694073677062988, "epoch": 3.66, "learning_rate": 4.201001878522229e-05, "loss": 100.7636, "step": 4327, "task_loss": 2.1892993450164795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.910304118640246, "compression/movement_sparsity/importance_threshold": -0.0005821392650408415, "compression/movement_sparsity/linear_layer_sparsity": 0.8821827966313177, "compression/movement_sparsity/model_sparsity": 0.8518771036553136, "compression_loss": 96.7444839477539, "distillation_loss": 4.3036322593688965, "epoch": 3.66, "learning_rate": 4.2006887914840324e-05, "loss": 101.0217, "step": 4328, "task_loss": 2.1989595890045166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9104732497358727, "compression/movement_sparsity/importance_threshold": -0.0005810415797267432, "compression/movement_sparsity/linear_layer_sparsity": 0.8823548146736326, "compression/movement_sparsity/model_sparsity": 0.8520432123486856, "compression_loss": 96.76175689697266, "distillation_loss": 3.504223346710205, "epoch": 3.66, "learning_rate": 4.200375704445836e-05, "loss": 101.3154, "step": 4329, "task_loss": 1.1153539419174194 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9106421680880112, "compression/movement_sparsity/importance_threshold": -0.0005799452751487051, "compression/movement_sparsity/linear_layer_sparsity": 0.8825816004178991, "compression/movement_sparsity/model_sparsity": 0.8522622073049619, "compression_loss": 96.77901458740234, "distillation_loss": 4.579192161560059, "epoch": 3.66, "learning_rate": 4.2000626174076394e-05, "loss": 101.9749, "step": 4330, "task_loss": 2.8037500381469727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9108108738305466, "compression/movement_sparsity/importance_threshold": -0.0005788503504377948, "compression/movement_sparsity/linear_layer_sparsity": 0.8826863900030831, "compression/movement_sparsity/model_sparsity": 0.8523633970455244, "compression_loss": 96.79629516601562, "distillation_loss": 4.310407638549805, "epoch": 3.66, "learning_rate": 4.1997495303694426e-05, "loss": 101.0785, "step": 4331, "task_loss": 1.4206949472427368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9109793670973635, "compression/movement_sparsity/importance_threshold": -0.0005777568047250796, "compression/movement_sparsity/linear_layer_sparsity": 0.8827498027265707, "compression/movement_sparsity/model_sparsity": 0.8524246313468797, "compression_loss": 96.8134765625, "distillation_loss": 5.884371757507324, "epoch": 3.66, "learning_rate": 4.199436443331246e-05, "loss": 101.285, "step": 4332, "task_loss": 2.822718620300293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9111476480223475, "compression/movement_sparsity/importance_threshold": -0.0005766646371416244, "compression/movement_sparsity/linear_layer_sparsity": 0.8828020305808157, "compression/movement_sparsity/model_sparsity": 0.8524750650136599, "compression_loss": 96.83069610595703, "distillation_loss": 3.914440631866455, "epoch": 3.66, "learning_rate": 4.1991233562930496e-05, "loss": 100.8764, "step": 4333, "task_loss": 2.064237356185913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9113157167393836, "compression/movement_sparsity/importance_threshold": -0.0005755738468184958, "compression/movement_sparsity/linear_layer_sparsity": 0.8827848836277553, "compression/movement_sparsity/model_sparsity": 0.8524585071111873, "compression_loss": 96.8478775024414, "distillation_loss": 6.022558689117432, "epoch": 3.66, "learning_rate": 4.198810269254853e-05, "loss": 101.8064, "step": 4334, "task_loss": 2.7832705974578857 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9114835733823571, "compression/movement_sparsity/importance_threshold": -0.0005744844328867578, "compression/movement_sparsity/linear_layer_sparsity": 0.8829194240111908, "compression/movement_sparsity/model_sparsity": 0.8525884256185571, "compression_loss": 96.86495208740234, "distillation_loss": 4.92145299911499, "epoch": 3.66, "learning_rate": 4.1984971822166566e-05, "loss": 101.0311, "step": 4335, "task_loss": 2.6590216159820557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9116512180851528, "compression/movement_sparsity/importance_threshold": -0.0005733963944774796, "compression/movement_sparsity/linear_layer_sparsity": 0.8829916129220583, "compression/movement_sparsity/model_sparsity": 0.8526581346182572, "compression_loss": 96.88201141357422, "distillation_loss": 5.81411600112915, "epoch": 3.66, "learning_rate": 4.19818409517846e-05, "loss": 101.4386, "step": 4336, "task_loss": 2.5298779010772705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9118186509816562, "compression/movement_sparsity/importance_threshold": -0.0005723097307217245, "compression/movement_sparsity/linear_layer_sparsity": 0.883108755944913, "compression/movement_sparsity/model_sparsity": 0.8527712534179028, "compression_loss": 96.89899444580078, "distillation_loss": 4.021148681640625, "epoch": 3.67, "learning_rate": 4.197871008140264e-05, "loss": 100.5066, "step": 4337, "task_loss": 1.919948697090149 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9119858722057521, "compression/movement_sparsity/importance_threshold": -0.0005712244407505606, "compression/movement_sparsity/linear_layer_sparsity": 0.8832198534147764, "compression/movement_sparsity/model_sparsity": 0.8528785343479006, "compression_loss": 96.9159927368164, "distillation_loss": 4.482931613922119, "epoch": 3.67, "learning_rate": 4.197557921102067e-05, "loss": 101.2027, "step": 4338, "task_loss": 2.683563709259033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9121528818913259, "compression/movement_sparsity/importance_threshold": -0.0005701405236950528, "compression/movement_sparsity/linear_layer_sparsity": 0.8832610156414553, "compression/movement_sparsity/model_sparsity": 0.8529182825254633, "compression_loss": 96.93302154541016, "distillation_loss": 5.9787092208862305, "epoch": 3.67, "learning_rate": 4.19724483406387e-05, "loss": 102.1842, "step": 4339, "task_loss": 2.9021189212799072 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9123196801722628, "compression/movement_sparsity/importance_threshold": -0.000569057978686267, "compression/movement_sparsity/linear_layer_sparsity": 0.8834152666739928, "compression/movement_sparsity/model_sparsity": 0.8530672345605015, "compression_loss": 96.94998931884766, "distillation_loss": 5.772254943847656, "epoch": 3.67, "learning_rate": 4.196931747025674e-05, "loss": 101.6021, "step": 4340, "task_loss": 2.674520492553711 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9124862671824476, "compression/movement_sparsity/importance_threshold": -0.0005679768048552705, "compression/movement_sparsity/linear_layer_sparsity": 0.883485404628027, "compression/movement_sparsity/model_sparsity": 0.853134963060045, "compression_loss": 96.96688079833984, "distillation_loss": 4.887200355529785, "epoch": 3.67, "learning_rate": 4.196618659987477e-05, "loss": 101.3655, "step": 4341, "task_loss": 2.5680458545684814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9126526430557658, "compression/movement_sparsity/importance_threshold": -0.0005668970013331274, "compression/movement_sparsity/linear_layer_sparsity": 0.8835791166614773, "compression/movement_sparsity/model_sparsity": 0.8532254557968543, "compression_loss": 96.98380279541016, "distillation_loss": 3.75598406791687, "epoch": 3.67, "learning_rate": 4.19630557294928e-05, "loss": 101.3902, "step": 4342, "task_loss": 1.6282644271850586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9128188079261023, "compression/movement_sparsity/importance_threshold": -0.000565818567250906, "compression/movement_sparsity/linear_layer_sparsity": 0.8835980999363535, "compression/movement_sparsity/model_sparsity": 0.8532437869378392, "compression_loss": 97.00067138671875, "distillation_loss": 3.6483004093170166, "epoch": 3.67, "learning_rate": 4.1959924859110834e-05, "loss": 101.911, "step": 4343, "task_loss": 2.813040256500244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9129847619273423, "compression/movement_sparsity/importance_threshold": -0.0005647415017396695, "compression/movement_sparsity/linear_layer_sparsity": 0.8836963550776731, "compression/movement_sparsity/model_sparsity": 0.8533386667127862, "compression_loss": 97.01757049560547, "distillation_loss": 4.927999496459961, "epoch": 3.67, "learning_rate": 4.195679398872887e-05, "loss": 101.6098, "step": 4344, "task_loss": 3.442683696746826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.913150505193371, "compression/movement_sparsity/importance_threshold": -0.0005636658039304869, "compression/movement_sparsity/linear_layer_sparsity": 0.8837455541933387, "compression/movement_sparsity/model_sparsity": 0.8533861756874745, "compression_loss": 97.03443908691406, "distillation_loss": 4.4375739097595215, "epoch": 3.67, "learning_rate": 4.1953663118346905e-05, "loss": 101.5258, "step": 4345, "task_loss": 2.7160181999206543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9133160378580736, "compression/movement_sparsity/importance_threshold": -0.0005625914729544225, "compression/movement_sparsity/linear_layer_sparsity": 0.8838357009006658, "compression/movement_sparsity/model_sparsity": 0.8534732255780813, "compression_loss": 97.05126953125, "distillation_loss": 5.682361602783203, "epoch": 3.67, "learning_rate": 4.1950532247964936e-05, "loss": 101.2532, "step": 4346, "task_loss": 3.25177001953125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9134813600553351, "compression/movement_sparsity/importance_threshold": -0.0005615185079425426, "compression/movement_sparsity/linear_layer_sparsity": 0.8838753368338874, "compression/movement_sparsity/model_sparsity": 0.8535114998950623, "compression_loss": 97.068115234375, "distillation_loss": 4.006507873535156, "epoch": 3.67, "learning_rate": 4.194740137758297e-05, "loss": 100.3977, "step": 4347, "task_loss": 1.7248294353485107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9136464719190405, "compression/movement_sparsity/importance_threshold": -0.0005604469080259149, "compression/movement_sparsity/linear_layer_sparsity": 0.8839785285806081, "compression/movement_sparsity/model_sparsity": 0.8536111466878282, "compression_loss": 97.08489990234375, "distillation_loss": 4.613685607910156, "epoch": 3.67, "learning_rate": 4.194427050720101e-05, "loss": 101.7201, "step": 4348, "task_loss": 1.91983962059021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9138113735830754, "compression/movement_sparsity/importance_threshold": -0.0005593766723356016, "compression/movement_sparsity/linear_layer_sparsity": 0.8840176040779508, "compression/movement_sparsity/model_sparsity": 0.853648879821627, "compression_loss": 97.10169219970703, "distillation_loss": 4.542597770690918, "epoch": 3.68, "learning_rate": 4.194113963681904e-05, "loss": 101.1649, "step": 4349, "task_loss": 2.2444815635681152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9139760651813245, "compression/movement_sparsity/importance_threshold": -0.0005583078000026727, "compression/movement_sparsity/linear_layer_sparsity": 0.8840177471679624, "compression/movement_sparsity/model_sparsity": 0.8536490179960565, "compression_loss": 97.11843872070312, "distillation_loss": 3.3814854621887207, "epoch": 3.68, "learning_rate": 4.193800876643707e-05, "loss": 101.273, "step": 4350, "task_loss": 2.5894250869750977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9141405468476731, "compression/movement_sparsity/importance_threshold": -0.0005572402901581923, "compression/movement_sparsity/linear_layer_sparsity": 0.8839954132019805, "compression/movement_sparsity/model_sparsity": 0.8536274512705133, "compression_loss": 97.13513946533203, "distillation_loss": 4.200962543487549, "epoch": 3.68, "learning_rate": 4.193487789605511e-05, "loss": 101.3352, "step": 4351, "task_loss": 2.1755878925323486 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9143048187160064, "compression/movement_sparsity/importance_threshold": -0.000556174141933226, "compression/movement_sparsity/linear_layer_sparsity": 0.8840718232681911, "compression/movement_sparsity/model_sparsity": 0.8537012364158847, "compression_loss": 97.1518325805664, "distillation_loss": 4.604788303375244, "epoch": 3.68, "learning_rate": 4.193174702567314e-05, "loss": 101.8065, "step": 4352, "task_loss": 2.292379140853882 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9144688809202095, "compression/movement_sparsity/importance_threshold": -0.0005551093544588414, "compression/movement_sparsity/linear_layer_sparsity": 0.8841869630308831, "compression/movement_sparsity/model_sparsity": 0.8538124207735168, "compression_loss": 97.16851806640625, "distillation_loss": 4.1818647384643555, "epoch": 3.68, "learning_rate": 4.192861615529117e-05, "loss": 101.1367, "step": 4353, "task_loss": 2.174238443374634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9146327335941675, "compression/movement_sparsity/importance_threshold": -0.0005540459268661042, "compression/movement_sparsity/linear_layer_sparsity": 0.8843219088360181, "compression/movement_sparsity/model_sparsity": 0.8539427307751036, "compression_loss": 97.18516540527344, "distillation_loss": 4.464888572692871, "epoch": 3.68, "learning_rate": 4.1925485284909204e-05, "loss": 101.1956, "step": 4354, "task_loss": 2.009399890899658 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9147963768717657, "compression/movement_sparsity/importance_threshold": -0.0005529838582860784, "compression/movement_sparsity/linear_layer_sparsity": 0.8844118766808307, "compression/movement_sparsity/model_sparsity": 0.8540296079476735, "compression_loss": 97.20182800292969, "distillation_loss": 3.8485963344573975, "epoch": 3.68, "learning_rate": 4.192235441452724e-05, "loss": 102.0802, "step": 4355, "task_loss": 2.2801365852355957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.914959810886889, "compression/movement_sparsity/importance_threshold": -0.0005519231478498324, "compression/movement_sparsity/linear_layer_sparsity": 0.884493569153304, "compression/movement_sparsity/model_sparsity": 0.8541084940324019, "compression_loss": 97.21839904785156, "distillation_loss": 5.375701904296875, "epoch": 3.68, "learning_rate": 4.1919223544145275e-05, "loss": 101.8602, "step": 4356, "task_loss": 3.081441879272461 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9151230357734226, "compression/movement_sparsity/importance_threshold": -0.0005508637946884326, "compression/movement_sparsity/linear_layer_sparsity": 0.8845638144248469, "compression/movement_sparsity/model_sparsity": 0.8541763261627675, "compression_loss": 97.2349624633789, "distillation_loss": 4.869045734405518, "epoch": 3.68, "learning_rate": 4.1916092673763306e-05, "loss": 101.1027, "step": 4357, "task_loss": 2.8468616008758545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9152860516652519, "compression/movement_sparsity/importance_threshold": -0.0005498057979329415, "compression/movement_sparsity/linear_layer_sparsity": 0.8847209988026228, "compression/movement_sparsity/model_sparsity": 0.8543281107736111, "compression_loss": 97.25157165527344, "distillation_loss": 3.9670209884643555, "epoch": 3.68, "learning_rate": 4.191296180338134e-05, "loss": 100.8708, "step": 4358, "task_loss": 1.8893166780471802 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9154488586962616, "compression/movement_sparsity/importance_threshold": -0.0005487491567144299, "compression/movement_sparsity/linear_layer_sparsity": 0.8847978858355389, "compression/movement_sparsity/model_sparsity": 0.8544023565004143, "compression_loss": 97.26811981201172, "distillation_loss": 4.8611955642700195, "epoch": 3.68, "learning_rate": 4.190983093299938e-05, "loss": 101.3183, "step": 4359, "task_loss": 3.6532211303710938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9156114570003372, "compression/movement_sparsity/importance_threshold": -0.000547693870163961, "compression/movement_sparsity/linear_layer_sparsity": 0.8848987881420735, "compression/movement_sparsity/model_sparsity": 0.8544997925023077, "compression_loss": 97.28463745117188, "distillation_loss": 4.066122531890869, "epoch": 3.69, "learning_rate": 4.190670006261741e-05, "loss": 101.3366, "step": 4360, "task_loss": 1.9372087717056274 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9157738467113636, "compression/movement_sparsity/importance_threshold": -0.0005466399374126005, "compression/movement_sparsity/linear_layer_sparsity": 0.8849665531867481, "compression/movement_sparsity/model_sparsity": 0.8545652296092281, "compression_loss": 97.30116271972656, "distillation_loss": 4.872408866882324, "epoch": 3.69, "learning_rate": 4.190356919223544e-05, "loss": 102.0841, "step": 4361, "task_loss": 2.225010871887207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9159360279632263, "compression/movement_sparsity/importance_threshold": -0.000545587357591415, "compression/movement_sparsity/linear_layer_sparsity": 0.8850913276768898, "compression/movement_sparsity/model_sparsity": 0.8546857177117821, "compression_loss": 97.317626953125, "distillation_loss": 4.703734397888184, "epoch": 3.69, "learning_rate": 4.190043832185347e-05, "loss": 101.9682, "step": 4362, "task_loss": 2.504549503326416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9160980008898099, "compression/movement_sparsity/importance_threshold": -0.0005445361298314728, "compression/movement_sparsity/linear_layer_sparsity": 0.8851920988175804, "compression/movement_sparsity/model_sparsity": 0.8547830270537818, "compression_loss": 97.33407592773438, "distillation_loss": 4.815277099609375, "epoch": 3.69, "learning_rate": 4.189730745147151e-05, "loss": 101.8563, "step": 4363, "task_loss": 2.6848387718200684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.916259765625, "compression/movement_sparsity/importance_threshold": -0.0005434862532638363, "compression/movement_sparsity/linear_layer_sparsity": 0.885257443256225, "compression/movement_sparsity/model_sparsity": 0.8548461267099359, "compression_loss": 97.35047912597656, "distillation_loss": 4.592268466949463, "epoch": 3.69, "learning_rate": 4.189417658108954e-05, "loss": 101.5302, "step": 4364, "task_loss": 1.666142225265503 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9164213223026816, "compression/movement_sparsity/importance_threshold": -0.0005424377270195737, "compression/movement_sparsity/linear_layer_sparsity": 0.8853233839032513, "compression/movement_sparsity/model_sparsity": 0.8549098020928797, "compression_loss": 97.3669204711914, "distillation_loss": 5.925614356994629, "epoch": 3.69, "learning_rate": 4.1891045710707574e-05, "loss": 101.6301, "step": 4365, "task_loss": 3.085991621017456 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9165826710567396, "compression/movement_sparsity/importance_threshold": -0.0005413905502297508, "compression/movement_sparsity/linear_layer_sparsity": 0.8854050406032217, "compression/movement_sparsity/model_sparsity": 0.8549886536340008, "compression_loss": 97.3833236694336, "distillation_loss": 4.377010822296143, "epoch": 3.69, "learning_rate": 4.188791484032561e-05, "loss": 101.3827, "step": 4366, "task_loss": 2.0016560554504395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9167438120210597, "compression/movement_sparsity/importance_threshold": -0.0005403447220254324, "compression/movement_sparsity/linear_layer_sparsity": 0.8855767724655134, "compression/movement_sparsity/model_sparsity": 0.8551544859785137, "compression_loss": 97.39971160888672, "distillation_loss": 4.305171966552734, "epoch": 3.69, "learning_rate": 4.1884783969943644e-05, "loss": 101.6001, "step": 4367, "task_loss": 3.1375153064727783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9169047453295265, "compression/movement_sparsity/importance_threshold": -0.0005393002415376861, "compression/movement_sparsity/linear_layer_sparsity": 0.8857079025370049, "compression/movement_sparsity/model_sparsity": 0.8552811113286463, "compression_loss": 97.4161376953125, "distillation_loss": 4.524194717407227, "epoch": 3.69, "learning_rate": 4.1881653099561676e-05, "loss": 101.4526, "step": 4368, "task_loss": 2.849992036819458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9170654711160253, "compression/movement_sparsity/importance_threshold": -0.0005382571078975775, "compression/movement_sparsity/linear_layer_sparsity": 0.8857139838624992, "compression/movement_sparsity/model_sparsity": 0.8552869837419015, "compression_loss": 97.43246459960938, "distillation_loss": 7.137128829956055, "epoch": 3.69, "learning_rate": 4.1878522229179715e-05, "loss": 102.2182, "step": 4369, "task_loss": 4.1804986000061035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9172259895144413, "compression/movement_sparsity/importance_threshold": -0.0005372153202361724, "compression/movement_sparsity/linear_layer_sparsity": 0.8857053865376339, "compression/movement_sparsity/model_sparsity": 0.8552786817615936, "compression_loss": 97.44872283935547, "distillation_loss": 5.015692710876465, "epoch": 3.69, "learning_rate": 4.1875391358797747e-05, "loss": 102.2313, "step": 4370, "task_loss": 2.213963508605957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9173863006586598, "compression/movement_sparsity/importance_threshold": -0.0005361748776845365, "compression/movement_sparsity/linear_layer_sparsity": 0.8857923733405374, "compression/movement_sparsity/model_sparsity": 0.8553626803002148, "compression_loss": 97.46507263183594, "distillation_loss": 5.774823188781738, "epoch": 3.69, "learning_rate": 4.1872260488415785e-05, "loss": 102.0522, "step": 4371, "task_loss": 2.161466360092163 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9175464046825655, "compression/movement_sparsity/importance_threshold": -0.0005351357793737374, "compression/movement_sparsity/linear_layer_sparsity": 0.8859573441997796, "compression/movement_sparsity/model_sparsity": 0.8555219839029323, "compression_loss": 97.48133087158203, "distillation_loss": 3.842717170715332, "epoch": 3.7, "learning_rate": 4.186912961803382e-05, "loss": 101.3994, "step": 4372, "task_loss": 2.0439860820770264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.917706301720044, "compression/movement_sparsity/importance_threshold": -0.0005340980244348389, "compression/movement_sparsity/linear_layer_sparsity": 0.886020172639053, "compression/movement_sparsity/model_sparsity": 0.8555826539920337, "compression_loss": 97.49755096435547, "distillation_loss": 5.592374324798584, "epoch": 3.7, "learning_rate": 4.186599874765185e-05, "loss": 102.024, "step": 4373, "task_loss": 3.403803586959839 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9178659919049802, "compression/movement_sparsity/importance_threshold": -0.0005330616119989086, "compression/movement_sparsity/linear_layer_sparsity": 0.886144505934992, "compression/movement_sparsity/model_sparsity": 0.8557027160567634, "compression_loss": 97.51380157470703, "distillation_loss": 4.406132221221924, "epoch": 3.7, "learning_rate": 4.186286787726989e-05, "loss": 102.174, "step": 4374, "task_loss": 2.958364963531494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9180254753712593, "compression/movement_sparsity/importance_threshold": -0.0005320265411970114, "compression/movement_sparsity/linear_layer_sparsity": 0.8862695785293246, "compression/movement_sparsity/model_sparsity": 0.8558234920227124, "compression_loss": 97.52999877929688, "distillation_loss": 3.27341890335083, "epoch": 3.7, "learning_rate": 4.185973700688792e-05, "loss": 101.8161, "step": 4375, "task_loss": 3.3783702850341797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9181847522527664, "compression/movement_sparsity/importance_threshold": -0.0005309928111602146, "compression/movement_sparsity/linear_layer_sparsity": 0.886356541483893, "compression/movement_sparsity/model_sparsity": 0.8559074675322619, "compression_loss": 97.5462417602539, "distillation_loss": 3.320213794708252, "epoch": 3.7, "learning_rate": 4.185660613650595e-05, "loss": 101.3031, "step": 4376, "task_loss": 2.1410861015319824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9183438226833867, "compression/movement_sparsity/importance_threshold": -0.0005299604210195841, "compression/movement_sparsity/linear_layer_sparsity": 0.8863431744919732, "compression/movement_sparsity/model_sparsity": 0.8558945597376363, "compression_loss": 97.5623779296875, "distillation_loss": 4.1441802978515625, "epoch": 3.7, "learning_rate": 4.185347526612399e-05, "loss": 101.5346, "step": 4377, "task_loss": 2.888065814971924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9185026867970052, "compression/movement_sparsity/importance_threshold": -0.0005289293699061856, "compression/movement_sparsity/linear_layer_sparsity": 0.8864695468205781, "compression/movement_sparsity/model_sparsity": 0.8560165907879868, "compression_loss": 97.57846069335938, "distillation_loss": 3.8366575241088867, "epoch": 3.7, "learning_rate": 4.185034439574202e-05, "loss": 101.7346, "step": 4378, "task_loss": 1.9181631803512573 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9186613447275072, "compression/movement_sparsity/importance_threshold": -0.0005278996569510848, "compression/movement_sparsity/linear_layer_sparsity": 0.8865124261273967, "compression/movement_sparsity/model_sparsity": 0.856057997058704, "compression_loss": 97.5946044921875, "distillation_loss": 3.7289137840270996, "epoch": 3.7, "learning_rate": 4.184721352536005e-05, "loss": 102.1131, "step": 4379, "task_loss": 1.2654414176940918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9188197966087779, "compression/movement_sparsity/importance_threshold": -0.0005268712812853465, "compression/movement_sparsity/linear_layer_sparsity": 0.8866487074393069, "compression/movement_sparsity/model_sparsity": 0.8561895966882997, "compression_loss": 97.61064910888672, "distillation_loss": 4.2692413330078125, "epoch": 3.7, "learning_rate": 4.1844082654978085e-05, "loss": 101.9782, "step": 4380, "task_loss": 2.0687649250030518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9189780425747021, "compression/movement_sparsity/importance_threshold": -0.0005258442420400409, "compression/movement_sparsity/linear_layer_sparsity": 0.8868111861475132, "compression/movement_sparsity/model_sparsity": 0.8563464937530362, "compression_loss": 97.62673950195312, "distillation_loss": 5.149149417877197, "epoch": 3.7, "learning_rate": 4.184095178459612e-05, "loss": 102.2706, "step": 4381, "task_loss": 2.7027831077575684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9191360827591653, "compression/movement_sparsity/importance_threshold": -0.0005248185383462294, "compression/movement_sparsity/linear_layer_sparsity": 0.8868919723832461, "compression/movement_sparsity/model_sparsity": 0.8564245047330442, "compression_loss": 97.6427993774414, "distillation_loss": 4.638134002685547, "epoch": 3.7, "learning_rate": 4.1837820914214155e-05, "loss": 102.5369, "step": 4382, "task_loss": 1.5227817296981812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9192939172960525, "compression/movement_sparsity/importance_threshold": -0.0005237941693349811, "compression/movement_sparsity/linear_layer_sparsity": 0.8869811293846596, "compression/movement_sparsity/model_sparsity": 0.85651059891718, "compression_loss": 97.65876770019531, "distillation_loss": 4.2402191162109375, "epoch": 3.7, "learning_rate": 4.183469004383219e-05, "loss": 101.852, "step": 4383, "task_loss": 2.4145960807800293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9194515463192489, "compression/movement_sparsity/importance_threshold": -0.0005227711341373601, "compression/movement_sparsity/linear_layer_sparsity": 0.8870902951393659, "compression/movement_sparsity/model_sparsity": 0.856616014492379, "compression_loss": 97.67476654052734, "distillation_loss": 5.588130950927734, "epoch": 3.71, "learning_rate": 4.183155917345022e-05, "loss": 102.2192, "step": 4384, "task_loss": 2.2360007762908936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9196089699626395, "compression/movement_sparsity/importance_threshold": -0.0005217494318844338, "compression/movement_sparsity/linear_layer_sparsity": 0.8871492959208281, "compression/movement_sparsity/model_sparsity": 0.8566729884154903, "compression_loss": 97.69075012207031, "distillation_loss": 4.554101943969727, "epoch": 3.71, "learning_rate": 4.182842830306826e-05, "loss": 102.4507, "step": 4385, "task_loss": 2.494971990585327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9197661883601096, "compression/movement_sparsity/importance_threshold": -0.0005207290617072671, "compression/movement_sparsity/linear_layer_sparsity": 0.8872272084321609, "compression/movement_sparsity/model_sparsity": 0.8567482243923719, "compression_loss": 97.7066421508789, "distillation_loss": 3.2989602088928223, "epoch": 3.71, "learning_rate": 4.182529743268629e-05, "loss": 101.8799, "step": 4386, "task_loss": 1.8674286603927612 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9199232016455442, "compression/movement_sparsity/importance_threshold": -0.0005197100227369274, "compression/movement_sparsity/linear_layer_sparsity": 0.8873184760112457, "compression/movement_sparsity/model_sparsity": 0.8568363566493433, "compression_loss": 97.72259521484375, "distillation_loss": 4.761124134063721, "epoch": 3.71, "learning_rate": 4.182216656230432e-05, "loss": 102.7477, "step": 4387, "task_loss": 2.1041555404663086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9200800099528286, "compression/movement_sparsity/importance_threshold": -0.0005186923141044806, "compression/movement_sparsity/linear_layer_sparsity": 0.8874153360249518, "compression/movement_sparsity/model_sparsity": 0.8569298892236024, "compression_loss": 97.73848724365234, "distillation_loss": 4.86476469039917, "epoch": 3.71, "learning_rate": 4.181903569192236e-05, "loss": 102.3659, "step": 4388, "task_loss": 1.990662932395935 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9202366134158478, "compression/movement_sparsity/importance_threshold": -0.0005176759349409905, "compression/movement_sparsity/linear_layer_sparsity": 0.8874627584396396, "compression/movement_sparsity/model_sparsity": 0.8569756825324573, "compression_loss": 97.75440979003906, "distillation_loss": 4.658374786376953, "epoch": 3.71, "learning_rate": 4.181590482154039e-05, "loss": 101.7531, "step": 4389, "task_loss": 3.3646514415740967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9203930121684869, "compression/movement_sparsity/importance_threshold": -0.0005166608843775265, "compression/movement_sparsity/linear_layer_sparsity": 0.887500569975213, "compression/movement_sparsity/model_sparsity": 0.8570121951254618, "compression_loss": 97.77019500732422, "distillation_loss": 5.643728256225586, "epoch": 3.71, "learning_rate": 4.181277395115842e-05, "loss": 102.2382, "step": 4390, "task_loss": 3.003822088241577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9205492063446312, "compression/movement_sparsity/importance_threshold": -0.0005156471615451524, "compression/movement_sparsity/linear_layer_sparsity": 0.8875937931177902, "compression/movement_sparsity/model_sparsity": 0.8571022157663035, "compression_loss": 97.78604888916016, "distillation_loss": 5.5766401290893555, "epoch": 3.71, "learning_rate": 4.1809643080776455e-05, "loss": 102.6902, "step": 4391, "task_loss": 2.7979679107666016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9207051960781659, "compression/movement_sparsity/importance_threshold": -0.0005146347655749341, "compression/movement_sparsity/linear_layer_sparsity": 0.8875929226535527, "compression/movement_sparsity/model_sparsity": 0.8571013752051905, "compression_loss": 97.8018798828125, "distillation_loss": 5.485591411590576, "epoch": 3.71, "learning_rate": 4.180651221039449e-05, "loss": 101.9718, "step": 4392, "task_loss": 3.2402467727661133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9208609815029759, "compression/movement_sparsity/importance_threshold": -0.000513623695597939, "compression/movement_sparsity/linear_layer_sparsity": 0.8875961779513173, "compression/movement_sparsity/model_sparsity": 0.8571045186734624, "compression_loss": 97.81770324707031, "distillation_loss": 5.430769920349121, "epoch": 3.71, "learning_rate": 4.1803381340012525e-05, "loss": 102.6298, "step": 4393, "task_loss": 2.6283187866210938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9210165627529465, "compression/movement_sparsity/importance_threshold": -0.000512613950745232, "compression/movement_sparsity/linear_layer_sparsity": 0.8876752232585755, "compression/movement_sparsity/model_sparsity": 0.8571808485312444, "compression_loss": 97.83357238769531, "distillation_loss": 4.1016998291015625, "epoch": 3.71, "learning_rate": 4.180025046963056e-05, "loss": 102.7004, "step": 4394, "task_loss": 3.4419915676116943 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9211719399619628, "compression/movement_sparsity/importance_threshold": -0.0005116055301478796, "compression/movement_sparsity/linear_layer_sparsity": 0.8876862769619739, "compression/movement_sparsity/model_sparsity": 0.8571915225059259, "compression_loss": 97.84931182861328, "distillation_loss": 3.9654934406280518, "epoch": 3.71, "learning_rate": 4.179711959924859e-05, "loss": 101.7713, "step": 4395, "task_loss": 2.6779987812042236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9213271132639099, "compression/movement_sparsity/importance_threshold": -0.0005105984329369477, "compression/movement_sparsity/linear_layer_sparsity": 0.8877427498198974, "compression/movement_sparsity/model_sparsity": 0.8572460553474489, "compression_loss": 97.86511993408203, "distillation_loss": 3.3681693077087402, "epoch": 3.72, "learning_rate": 4.179398872886663e-05, "loss": 102.2288, "step": 4396, "task_loss": 1.5684648752212524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.921482082792673, "compression/movement_sparsity/importance_threshold": -0.0005095926582435037, "compression/movement_sparsity/linear_layer_sparsity": 0.8877347725517489, "compression/movement_sparsity/model_sparsity": 0.8572383521230024, "compression_loss": 97.88088989257812, "distillation_loss": 3.8956332206726074, "epoch": 3.72, "learning_rate": 4.179085785848466e-05, "loss": 102.1132, "step": 4397, "task_loss": 2.770991802215576 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9216368486821372, "compression/movement_sparsity/importance_threshold": -0.0005085882051986106, "compression/movement_sparsity/linear_layer_sparsity": 0.8878141517357009, "compression/movement_sparsity/model_sparsity": 0.8573150043877866, "compression_loss": 97.89662170410156, "distillation_loss": 2.8147692680358887, "epoch": 3.72, "learning_rate": 4.178772698810269e-05, "loss": 101.4214, "step": 4398, "task_loss": 2.5741429328918457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9217914110661877, "compression/movement_sparsity/importance_threshold": -0.0005075850729333378, "compression/movement_sparsity/linear_layer_sparsity": 0.8878661768790962, "compression/movement_sparsity/model_sparsity": 0.8573652423074581, "compression_loss": 97.91230010986328, "distillation_loss": 4.455719947814941, "epoch": 3.72, "learning_rate": 4.178459611772072e-05, "loss": 102.332, "step": 4399, "task_loss": 2.9599649906158447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9219457700787095, "compression/movement_sparsity/importance_threshold": -0.0005065832605787501, "compression/movement_sparsity/linear_layer_sparsity": 0.8878786018617728, "compression/movement_sparsity/model_sparsity": 0.8573772404537561, "compression_loss": 97.927978515625, "distillation_loss": 6.010303497314453, "epoch": 3.72, "learning_rate": 4.178146524733876e-05, "loss": 102.4676, "step": 4400, "task_loss": 2.321807384490967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9220999258535879, "compression/movement_sparsity/importance_threshold": -0.0005055827672659123, "compression/movement_sparsity/linear_layer_sparsity": 0.8879220773769732, "compression/movement_sparsity/model_sparsity": 0.857419222451263, "compression_loss": 97.94357299804688, "distillation_loss": 4.234959602355957, "epoch": 3.72, "learning_rate": 4.177833437695679e-05, "loss": 101.7321, "step": 4401, "task_loss": 4.555571556091309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9222538785247081, "compression/movement_sparsity/importance_threshold": -0.0005045835921258903, "compression/movement_sparsity/linear_layer_sparsity": 0.8880841864359826, "compression/movement_sparsity/model_sparsity": 0.8575757625653897, "compression_loss": 97.9591064453125, "distillation_loss": 4.125043869018555, "epoch": 3.72, "learning_rate": 4.177520350657483e-05, "loss": 101.691, "step": 4402, "task_loss": 1.4486937522888184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.922407628225955, "compression/movement_sparsity/importance_threshold": -0.0005035857342897522, "compression/movement_sparsity/linear_layer_sparsity": 0.8881266722452692, "compression/movement_sparsity/model_sparsity": 0.8576167888564257, "compression_loss": 97.97460174560547, "distillation_loss": 2.49403715133667, "epoch": 3.72, "learning_rate": 4.177207263619286e-05, "loss": 101.8523, "step": 4403, "task_loss": 1.909193754196167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9225611750912138, "compression/movement_sparsity/importance_threshold": -0.000502589192888564, "compression/movement_sparsity/linear_layer_sparsity": 0.8882120612097096, "compression/movement_sparsity/model_sparsity": 0.8576992444472504, "compression_loss": 97.99006652832031, "distillation_loss": 6.112577438354492, "epoch": 3.72, "learning_rate": 4.17689417658109e-05, "loss": 103.0264, "step": 4404, "task_loss": 3.271435022354126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9227145192543698, "compression/movement_sparsity/importance_threshold": -0.0005015939670533903, "compression/movement_sparsity/linear_layer_sparsity": 0.8882931932463041, "compression/movement_sparsity/model_sparsity": 0.8577775893487964, "compression_loss": 98.0054702758789, "distillation_loss": 3.937145709991455, "epoch": 3.72, "learning_rate": 4.1765810895428933e-05, "loss": 101.6285, "step": 4405, "task_loss": 1.3687411546707153 }, { "compression/movement_sparsity/importance_regularization_factor": 0.922867660849308, "compression/movement_sparsity/importance_threshold": -0.000500600055915297, "compression/movement_sparsity/linear_layer_sparsity": 0.8883928912119076, "compression/movement_sparsity/model_sparsity": 0.8578738623825746, "compression_loss": 98.0208511352539, "distillation_loss": 5.9676289558410645, "epoch": 3.72, "learning_rate": 4.1762680025046965e-05, "loss": 101.9301, "step": 4406, "task_loss": 3.008826971054077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9230206000099137, "compression/movement_sparsity/importance_threshold": -0.0004996074586053507, "compression/movement_sparsity/linear_layer_sparsity": 0.8884891073205613, "compression/movement_sparsity/model_sparsity": 0.8579667731719007, "compression_loss": 98.03624725341797, "distillation_loss": 2.7344985008239746, "epoch": 3.72, "learning_rate": 4.1759549154665004e-05, "loss": 102.2046, "step": 4407, "task_loss": 0.899051308631897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9231733368700717, "compression/movement_sparsity/importance_threshold": -0.0004986161742546189, "compression/movement_sparsity/linear_layer_sparsity": 0.8885970806585041, "compression/movement_sparsity/model_sparsity": 0.8580710372935203, "compression_loss": 98.05162811279297, "distillation_loss": 5.016868591308594, "epoch": 3.73, "learning_rate": 4.1756418284283036e-05, "loss": 103.2141, "step": 4408, "task_loss": 2.8446998596191406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9233258715636676, "compression/movement_sparsity/importance_threshold": -0.0004976262019941646, "compression/movement_sparsity/linear_layer_sparsity": 0.8887730098278036, "compression/movement_sparsity/model_sparsity": 0.858240922754633, "compression_loss": 98.06693267822266, "distillation_loss": 6.80361270904541, "epoch": 3.73, "learning_rate": 4.175328741390107e-05, "loss": 103.2296, "step": 4409, "task_loss": 3.2781262397766113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9234782042245862, "compression/movement_sparsity/importance_threshold": -0.0004966375409550571, "compression/movement_sparsity/linear_layer_sparsity": 0.8888671034346182, "compression/movement_sparsity/model_sparsity": 0.8583317839565877, "compression_loss": 98.08229064941406, "distillation_loss": 4.971750259399414, "epoch": 3.73, "learning_rate": 4.17501565435191e-05, "loss": 102.6364, "step": 4410, "task_loss": 2.107276201248169 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9236303349867128, "compression/movement_sparsity/importance_threshold": -0.0004956501902683595, "compression/movement_sparsity/linear_layer_sparsity": 0.8888993702322409, "compression/movement_sparsity/model_sparsity": 0.8583629422904477, "compression_loss": 98.0976333618164, "distillation_loss": 5.519400596618652, "epoch": 3.73, "learning_rate": 4.174702567313714e-05, "loss": 103.2221, "step": 4411, "task_loss": 3.165459394454956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9237822639839324, "compression/movement_sparsity/importance_threshold": -0.0004946641490651401, "compression/movement_sparsity/linear_layer_sparsity": 0.8889947516491601, "compression/movement_sparsity/model_sparsity": 0.8584550470622683, "compression_loss": 98.11297607421875, "distillation_loss": 3.7380154132843018, "epoch": 3.73, "learning_rate": 4.174389480275517e-05, "loss": 103.0266, "step": 4412, "task_loss": 2.3041200637817383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9239339913501302, "compression/movement_sparsity/importance_threshold": -0.0004936794164764639, "compression/movement_sparsity/linear_layer_sparsity": 0.8891301386484977, "compression/movement_sparsity/model_sparsity": 0.8585857831016794, "compression_loss": 98.12828063964844, "distillation_loss": 4.103647232055664, "epoch": 3.73, "learning_rate": 4.17407639323732e-05, "loss": 102.8395, "step": 4413, "task_loss": 2.837089776992798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9240855172191914, "compression/movement_sparsity/importance_threshold": -0.0004926959916333965, "compression/movement_sparsity/linear_layer_sparsity": 0.8891640509812541, "compression/movement_sparsity/model_sparsity": 0.8586185304414792, "compression_loss": 98.14360046386719, "distillation_loss": 4.531942367553711, "epoch": 3.73, "learning_rate": 4.173763306199124e-05, "loss": 102.607, "step": 4414, "task_loss": 2.067744731903076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9242368417250012, "compression/movement_sparsity/importance_threshold": -0.0004917138736670046, "compression/movement_sparsity/linear_layer_sparsity": 0.8892663245670669, "compression/movement_sparsity/model_sparsity": 0.8587172906149889, "compression_loss": 98.158935546875, "distillation_loss": 4.881964683532715, "epoch": 3.73, "learning_rate": 4.173450219160927e-05, "loss": 102.3974, "step": 4415, "task_loss": 2.8301937580108643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9243879650014446, "compression/movement_sparsity/importance_threshold": -0.0004907330617083547, "compression/movement_sparsity/linear_layer_sparsity": 0.8893498414371884, "compression/movement_sparsity/model_sparsity": 0.8587979384236939, "compression_loss": 98.17420959472656, "distillation_loss": 3.9609415531158447, "epoch": 3.73, "learning_rate": 4.1731371321227303e-05, "loss": 102.249, "step": 4416, "task_loss": 1.6792418956756592 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9245388871824067, "compression/movement_sparsity/importance_threshold": -0.0004897535548885118, "compression/movement_sparsity/linear_layer_sparsity": 0.8894479534884964, "compression/movement_sparsity/model_sparsity": 0.8588926800242114, "compression_loss": 98.18948364257812, "distillation_loss": 3.8716907501220703, "epoch": 3.73, "learning_rate": 4.1728240450845335e-05, "loss": 102.7259, "step": 4417, "task_loss": 3.0061421394348145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9246896084017728, "compression/movement_sparsity/importance_threshold": -0.0004887753523385423, "compression/movement_sparsity/linear_layer_sparsity": 0.8895564157173121, "compression/movement_sparsity/model_sparsity": 0.8589974162417985, "compression_loss": 98.20471954345703, "distillation_loss": 5.763439178466797, "epoch": 3.73, "learning_rate": 4.1725109580463374e-05, "loss": 102.9537, "step": 4418, "task_loss": 2.0671639442443848 }, { "compression/movement_sparsity/importance_regularization_factor": 0.924840128793428, "compression/movement_sparsity/importance_threshold": -0.00048779845318951305, "compression/movement_sparsity/linear_layer_sparsity": 0.8894878994500764, "compression/movement_sparsity/model_sparsity": 0.858931253719123, "compression_loss": 98.22003173828125, "distillation_loss": 4.3122148513793945, "epoch": 3.73, "learning_rate": 4.1721978710081406e-05, "loss": 102.4592, "step": 4419, "task_loss": 1.6163713932037354 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9249904484912574, "compression/movement_sparsity/importance_threshold": -0.00048682285657248876, "compression/movement_sparsity/linear_layer_sparsity": 0.889530778756895, "compression/movement_sparsity/model_sparsity": 0.8589726599898403, "compression_loss": 98.23526000976562, "distillation_loss": 3.806293487548828, "epoch": 3.74, "learning_rate": 4.171884783969944e-05, "loss": 102.3752, "step": 4420, "task_loss": 2.767552375793457 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9251405676291462, "compression/movement_sparsity/importance_threshold": -0.00048584856161853606, "compression/movement_sparsity/linear_layer_sparsity": 0.8896087270407307, "compression/movement_sparsity/model_sparsity": 0.8590479305103291, "compression_loss": 98.2505111694336, "distillation_loss": 4.68280553817749, "epoch": 3.74, "learning_rate": 4.171571696931747e-05, "loss": 103.0045, "step": 4421, "task_loss": 1.6856708526611328 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9252904863409794, "compression/movement_sparsity/importance_threshold": -0.00048487556745872157, "compression/movement_sparsity/linear_layer_sparsity": 0.8897175946912461, "compression/movement_sparsity/model_sparsity": 0.8591530582221333, "compression_loss": 98.2657241821289, "distillation_loss": 4.505087852478027, "epoch": 3.74, "learning_rate": 4.171258609893551e-05, "loss": 102.7886, "step": 4422, "task_loss": 2.336688280105591 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9254402047606423, "compression/movement_sparsity/importance_threshold": -0.000483903873224111, "compression/movement_sparsity/linear_layer_sparsity": 0.8897186917146686, "compression/movement_sparsity/model_sparsity": 0.8591541175594264, "compression_loss": 98.2809066772461, "distillation_loss": 4.60619592666626, "epoch": 3.74, "learning_rate": 4.170945522855354e-05, "loss": 102.7546, "step": 4423, "task_loss": 3.2721612453460693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9255897230220199, "compression/movement_sparsity/importance_threshold": -0.00048293347804577013, "compression/movement_sparsity/linear_layer_sparsity": 0.8896917192474763, "compression/movement_sparsity/model_sparsity": 0.8591280716794591, "compression_loss": 98.29607391357422, "distillation_loss": 4.358454704284668, "epoch": 3.74, "learning_rate": 4.170632435817157e-05, "loss": 102.2533, "step": 4424, "task_loss": 2.9876773357391357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9257390412589975, "compression/movement_sparsity/importance_threshold": -0.0004819643810547638, "compression/movement_sparsity/linear_layer_sparsity": 0.8897976774010885, "compression/movement_sparsity/model_sparsity": 0.8592303898445295, "compression_loss": 98.31122589111328, "distillation_loss": 5.918052673339844, "epoch": 3.74, "learning_rate": 4.170319348778961e-05, "loss": 103.1922, "step": 4425, "task_loss": 3.724876642227173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9258881596054601, "compression/movement_sparsity/importance_threshold": -0.0004809965813821612, "compression/movement_sparsity/linear_layer_sparsity": 0.8899342448930221, "compression/movement_sparsity/model_sparsity": 0.8593622658229844, "compression_loss": 98.32632446289062, "distillation_loss": 2.971859931945801, "epoch": 3.74, "learning_rate": 4.170006261740764e-05, "loss": 102.452, "step": 4426, "task_loss": 1.1929054260253906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9260370781952929, "compression/movement_sparsity/importance_threshold": -0.00048003007815902547, "compression/movement_sparsity/linear_layer_sparsity": 0.8900522822284495, "compression/movement_sparsity/model_sparsity": 0.8594762482128145, "compression_loss": 98.34139251708984, "distillation_loss": 4.454765796661377, "epoch": 3.74, "learning_rate": 4.169693174702567e-05, "loss": 102.1324, "step": 4427, "task_loss": 2.6102359294891357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.926185797162381, "compression/movement_sparsity/importance_threshold": -0.0004790648705164241, "compression/movement_sparsity/linear_layer_sparsity": 0.8901680658961937, "compression/movement_sparsity/model_sparsity": 0.8595880543553795, "compression_loss": 98.35648345947266, "distillation_loss": 4.262306213378906, "epoch": 3.74, "learning_rate": 4.1693800876643705e-05, "loss": 102.7717, "step": 4428, "task_loss": 1.855597972869873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9263343166406097, "compression/movement_sparsity/importance_threshold": -0.00047810095758542195, "compression/movement_sparsity/linear_layer_sparsity": 0.8902782690534844, "compression/movement_sparsity/model_sparsity": 0.8596944716951926, "compression_loss": 98.3715591430664, "distillation_loss": 4.33069372177124, "epoch": 3.74, "learning_rate": 4.1690670006261744e-05, "loss": 101.924, "step": 4429, "task_loss": 2.461531639099121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9264826367638639, "compression/movement_sparsity/importance_threshold": -0.0004771383384970865, "compression/movement_sparsity/linear_layer_sparsity": 0.8903079602308976, "compression/movement_sparsity/model_sparsity": 0.8597231428893211, "compression_loss": 98.3865737915039, "distillation_loss": 3.6567368507385254, "epoch": 3.74, "learning_rate": 4.1687539135879775e-05, "loss": 102.4556, "step": 4430, "task_loss": 2.5508127212524414 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9266307576660289, "compression/movement_sparsity/importance_threshold": -0.00047617701238248343, "compression/movement_sparsity/linear_layer_sparsity": 0.8903340145371821, "compression/movement_sparsity/model_sparsity": 0.8597483021500322, "compression_loss": 98.40162658691406, "distillation_loss": 4.968832015991211, "epoch": 3.75, "learning_rate": 4.168440826549781e-05, "loss": 103.4224, "step": 4431, "task_loss": 3.3280272483825684 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9267786794809898, "compression/movement_sparsity/importance_threshold": -0.00047521697837267767, "compression/movement_sparsity/linear_layer_sparsity": 0.8904519803276036, "compression/movement_sparsity/model_sparsity": 0.8598622154526475, "compression_loss": 98.41658782958984, "distillation_loss": 4.700809478759766, "epoch": 3.75, "learning_rate": 4.168127739511584e-05, "loss": 103.2138, "step": 4432, "task_loss": 3.389479160308838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9269264023426318, "compression/movement_sparsity/importance_threshold": -0.0004742582355987358, "compression/movement_sparsity/linear_layer_sparsity": 0.890525123171882, "compression/movement_sparsity/model_sparsity": 0.8599328456152112, "compression_loss": 98.431640625, "distillation_loss": 4.589573383331299, "epoch": 3.75, "learning_rate": 4.167814652473388e-05, "loss": 103.453, "step": 4433, "task_loss": 2.583601951599121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9270739263848399, "compression/movement_sparsity/importance_threshold": -0.00047330078319172443, "compression/movement_sparsity/linear_layer_sparsity": 0.8906873037758973, "compression/movement_sparsity/model_sparsity": 0.8600894548165527, "compression_loss": 98.44664001464844, "distillation_loss": 3.5141005516052246, "epoch": 3.75, "learning_rate": 4.167501565435191e-05, "loss": 102.3379, "step": 4434, "task_loss": 1.7071340084075928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9272212517414994, "compression/movement_sparsity/importance_threshold": -0.0004723446202827093, "compression/movement_sparsity/linear_layer_sparsity": 0.8907281917467208, "compression/movement_sparsity/model_sparsity": 0.8601289381597923, "compression_loss": 98.46157836914062, "distillation_loss": 4.201456069946289, "epoch": 3.75, "learning_rate": 4.167188478396994e-05, "loss": 102.5042, "step": 4435, "task_loss": 1.976671814918518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9273683785464952, "compression/movement_sparsity/importance_threshold": -0.000471389746002757, "compression/movement_sparsity/linear_layer_sparsity": 0.8908338398719745, "compression/movement_sparsity/model_sparsity": 0.8602309569469319, "compression_loss": 98.47648620605469, "distillation_loss": 4.305334091186523, "epoch": 3.75, "learning_rate": 4.166875391358798e-05, "loss": 102.9936, "step": 4436, "task_loss": 2.3879311084747314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9275153069337128, "compression/movement_sparsity/importance_threshold": -0.00047043615948293236, "compression/movement_sparsity/linear_layer_sparsity": 0.8908839452243804, "compression/movement_sparsity/model_sparsity": 0.8602793410263405, "compression_loss": 98.49138641357422, "distillation_loss": 3.095916748046875, "epoch": 3.75, "learning_rate": 4.166562304320601e-05, "loss": 102.233, "step": 4437, "task_loss": 1.4192885160446167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9276620370370371, "compression/movement_sparsity/importance_threshold": -0.0004694838598543012, "compression/movement_sparsity/linear_layer_sparsity": 0.8910771644367519, "compression/movement_sparsity/model_sparsity": 0.8604659225643553, "compression_loss": 98.50631713867188, "distillation_loss": 6.335461616516113, "epoch": 3.75, "learning_rate": 4.166249217282405e-05, "loss": 103.28, "step": 4438, "task_loss": 3.0801639556884766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9278085689903532, "compression/movement_sparsity/importance_threshold": -0.0004685328462479318, "compression/movement_sparsity/linear_layer_sparsity": 0.8911554942939519, "compression/movement_sparsity/model_sparsity": 0.8605415615499896, "compression_loss": 98.52119445800781, "distillation_loss": 4.0575690269470215, "epoch": 3.75, "learning_rate": 4.165936130244208e-05, "loss": 102.6744, "step": 4439, "task_loss": 2.0254907608032227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9279549029275463, "compression/movement_sparsity/importance_threshold": -0.000467583117794889, "compression/movement_sparsity/linear_layer_sparsity": 0.8911733566970704, "compression/movement_sparsity/model_sparsity": 0.8605588103246099, "compression_loss": 98.53596496582031, "distillation_loss": 3.3245294094085693, "epoch": 3.75, "learning_rate": 4.165623043206012e-05, "loss": 102.6602, "step": 4440, "task_loss": 1.848137378692627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9281010389825016, "compression/movement_sparsity/importance_threshold": -0.00046663467362623775, "compression/movement_sparsity/linear_layer_sparsity": 0.8912469049630485, "compression/movement_sparsity/model_sparsity": 0.8606298319813905, "compression_loss": 98.55081939697266, "distillation_loss": 5.823215484619141, "epoch": 3.75, "learning_rate": 4.165309956167815e-05, "loss": 103.5133, "step": 4441, "task_loss": 2.6198511123657227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9282469772891043, "compression/movement_sparsity/importance_threshold": -0.00046568751287304547, "compression/movement_sparsity/linear_layer_sparsity": 0.8913160651353365, "compression/movement_sparsity/model_sparsity": 0.860696616288999, "compression_loss": 98.56565856933594, "distillation_loss": 4.044127941131592, "epoch": 3.75, "learning_rate": 4.1649968691296184e-05, "loss": 102.5394, "step": 4442, "task_loss": 1.912967324256897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9283927179812393, "compression/movement_sparsity/importance_threshold": -0.0004647416346663779, "compression/movement_sparsity/linear_layer_sparsity": 0.8914352710391922, "compression/movement_sparsity/model_sparsity": 0.860811727103337, "compression_loss": 98.5804443359375, "distillation_loss": 5.277503967285156, "epoch": 3.76, "learning_rate": 4.1646837820914216e-05, "loss": 103.3955, "step": 4443, "task_loss": 2.4384348392486572 }, { "compression/movement_sparsity/importance_regularization_factor": 0.928538261192792, "compression/movement_sparsity/importance_threshold": -0.0004637970381372999, "compression/movement_sparsity/linear_layer_sparsity": 0.8915414319036542, "compression/movement_sparsity/model_sparsity": 0.8609142410155157, "compression_loss": 98.59525299072266, "distillation_loss": 3.4935193061828613, "epoch": 3.76, "learning_rate": 4.1643706950532254e-05, "loss": 102.3577, "step": 4444, "task_loss": 2.2318806648254395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9286836070576474, "compression/movement_sparsity/importance_threshold": -0.00046285372241687984, "compression/movement_sparsity/linear_layer_sparsity": 0.8916570128605487, "compression/movement_sparsity/model_sparsity": 0.8610258514109722, "compression_loss": 98.61000061035156, "distillation_loss": 4.186989784240723, "epoch": 3.76, "learning_rate": 4.1640576080150286e-05, "loss": 102.5127, "step": 4445, "task_loss": 2.5130884647369385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9288287557096907, "compression/movement_sparsity/importance_threshold": -0.0004619116866361817, "compression/movement_sparsity/linear_layer_sparsity": 0.8916810519825026, "compression/movement_sparsity/model_sparsity": 0.861049064715134, "compression_loss": 98.62467956542969, "distillation_loss": 5.431525230407715, "epoch": 3.76, "learning_rate": 4.163744520976832e-05, "loss": 103.1569, "step": 4446, "task_loss": 2.3700380325317383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.928973707282807, "compression/movement_sparsity/importance_threshold": -0.00046097092992627205, "compression/movement_sparsity/linear_layer_sparsity": 0.8917861992927156, "compression/movement_sparsity/model_sparsity": 0.8611505998917702, "compression_loss": 98.63932800292969, "distillation_loss": 4.785214900970459, "epoch": 3.76, "learning_rate": 4.163431433938635e-05, "loss": 102.9175, "step": 4447, "task_loss": 3.8161017894744873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9291184619108814, "compression/movement_sparsity/importance_threshold": -0.00046003145141821754, "compression/movement_sparsity/linear_layer_sparsity": 0.8918003532796994, "compression/movement_sparsity/model_sparsity": 0.8611642676457584, "compression_loss": 98.65400695800781, "distillation_loss": 4.242592811584473, "epoch": 3.76, "learning_rate": 4.163118346900439e-05, "loss": 102.6299, "step": 4448, "task_loss": 2.195553779602051 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9292630197277991, "compression/movement_sparsity/importance_threshold": -0.0004590932502430839, "compression/movement_sparsity/linear_layer_sparsity": 0.8919321511045786, "compression/movement_sparsity/model_sparsity": 0.8612915378098954, "compression_loss": 98.66868591308594, "distillation_loss": 3.703857183456421, "epoch": 3.76, "learning_rate": 4.162805259862242e-05, "loss": 102.3081, "step": 4449, "task_loss": 1.1028790473937988 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9294073808674452, "compression/movement_sparsity/importance_threshold": -0.00045815632553193684, "compression/movement_sparsity/linear_layer_sparsity": 0.8920437732378179, "compression/movement_sparsity/model_sparsity": 0.8613993253794682, "compression_loss": 98.68330383300781, "distillation_loss": 3.832453727722168, "epoch": 3.76, "learning_rate": 4.162492172824045e-05, "loss": 102.69, "step": 4450, "task_loss": 2.9609053134918213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.929551545463705, "compression/movement_sparsity/importance_threshold": -0.0004572206764158421, "compression/movement_sparsity/linear_layer_sparsity": 0.8920511185250816, "compression/movement_sparsity/model_sparsity": 0.8614064183335176, "compression_loss": 98.69788360595703, "distillation_loss": 6.250268936157227, "epoch": 3.76, "learning_rate": 4.162179085785849e-05, "loss": 103.1911, "step": 4451, "task_loss": 3.3039557933807373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9296955136504633, "compression/movement_sparsity/importance_threshold": -0.00045628630202586805, "compression/movement_sparsity/linear_layer_sparsity": 0.8920848758436587, "compression/movement_sparsity/model_sparsity": 0.8614390159843519, "compression_loss": 98.71240234375, "distillation_loss": 5.855362892150879, "epoch": 3.76, "learning_rate": 4.161865998747652e-05, "loss": 103.5432, "step": 4452, "task_loss": 3.1981797218322754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9298392855616057, "compression/movement_sparsity/importance_threshold": -0.0004553532014930769, "compression/movement_sparsity/linear_layer_sparsity": 0.8921912751914736, "compression/movement_sparsity/model_sparsity": 0.8615417601872466, "compression_loss": 98.72698211669922, "distillation_loss": 4.656683921813965, "epoch": 3.76, "learning_rate": 4.1615529117094554e-05, "loss": 102.7055, "step": 4453, "task_loss": 2.660473585128784 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9299828613310169, "compression/movement_sparsity/importance_threshold": -0.0004544213739485379, "compression/movement_sparsity/linear_layer_sparsity": 0.8923268410533256, "compression/movement_sparsity/model_sparsity": 0.8616726689446947, "compression_loss": 98.74144744873047, "distillation_loss": 4.6848649978637695, "epoch": 3.76, "learning_rate": 4.1612398246712586e-05, "loss": 103.3436, "step": 4454, "task_loss": 2.332627296447754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9301262410925824, "compression/movement_sparsity/importance_threshold": -0.00045349081852331507, "compression/movement_sparsity/linear_layer_sparsity": 0.8924293769708265, "compression/movement_sparsity/model_sparsity": 0.861771682437992, "compression_loss": 98.75597381591797, "distillation_loss": 4.7595930099487305, "epoch": 3.77, "learning_rate": 4.1609267376330624e-05, "loss": 103.0363, "step": 4455, "task_loss": 2.0028018951416016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.930269424980187, "compression/movement_sparsity/importance_threshold": -0.0004525615343484758, "compression/movement_sparsity/linear_layer_sparsity": 0.8924332284769728, "compression/movement_sparsity/model_sparsity": 0.8617754016330537, "compression_loss": 98.77041625976562, "distillation_loss": 3.9076905250549316, "epoch": 3.77, "learning_rate": 4.1606136505948656e-05, "loss": 102.2835, "step": 4456, "task_loss": 2.044459104537964 }, { "compression/movement_sparsity/importance_regularization_factor": 0.930412413127716, "compression/movement_sparsity/importance_threshold": -0.0004516335205550868, "compression/movement_sparsity/linear_layer_sparsity": 0.8925105447799238, "compression/movement_sparsity/model_sparsity": 0.8618500618831455, "compression_loss": 98.78485107421875, "distillation_loss": 6.431154251098633, "epoch": 3.77, "learning_rate": 4.160300563556669e-05, "loss": 103.0194, "step": 4457, "task_loss": 3.995347738265991 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9305552056690547, "compression/movement_sparsity/importance_threshold": -0.0004507067762742111, "compression/movement_sparsity/linear_layer_sparsity": 0.8927182399318052, "compression/movement_sparsity/model_sparsity": 0.8620506220676147, "compression_loss": 98.7992935180664, "distillation_loss": 5.332747459411621, "epoch": 3.77, "learning_rate": 4.159987476518472e-05, "loss": 103.0928, "step": 4458, "task_loss": 1.7774864435195923 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9306978027380881, "compression/movement_sparsity/importance_threshold": -0.00044978130063691706, "compression/movement_sparsity/linear_layer_sparsity": 0.8926850788216099, "compression/movement_sparsity/model_sparsity": 0.8620186001435701, "compression_loss": 98.813720703125, "distillation_loss": 4.489331245422363, "epoch": 3.77, "learning_rate": 4.159674389480276e-05, "loss": 102.7709, "step": 4459, "task_loss": 2.5787432193756104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9308402044687013, "compression/movement_sparsity/importance_threshold": -0.00044885709277427046, "compression/movement_sparsity/linear_layer_sparsity": 0.8927693469142925, "compression/movement_sparsity/model_sparsity": 0.8620999733680301, "compression_loss": 98.828125, "distillation_loss": 3.2915191650390625, "epoch": 3.77, "learning_rate": 4.159361302442079e-05, "loss": 103.069, "step": 4460, "task_loss": 2.7292308807373047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9309824109947794, "compression/movement_sparsity/importance_threshold": -0.0004479341518173379, "compression/movement_sparsity/linear_layer_sparsity": 0.8928089589991789, "compression/movement_sparsity/model_sparsity": 0.8621382246559396, "compression_loss": 98.84245300292969, "distillation_loss": 4.1890764236450195, "epoch": 3.77, "learning_rate": 4.159048215403882e-05, "loss": 103.4922, "step": 4461, "task_loss": 2.247316360473633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9311244224502077, "compression/movement_sparsity/importance_threshold": -0.0004470124768971833, "compression/movement_sparsity/linear_layer_sparsity": 0.8928978059722337, "compression/movement_sparsity/model_sparsity": 0.8622240194621448, "compression_loss": 98.85684967041016, "distillation_loss": 3.8742027282714844, "epoch": 3.77, "learning_rate": 4.158735128365686e-05, "loss": 103.0748, "step": 4462, "task_loss": 1.7955517768859863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9312662389688712, "compression/movement_sparsity/importance_threshold": -0.0004460920671448751, "compression/movement_sparsity/linear_layer_sparsity": 0.8929848404718078, "compression/movement_sparsity/model_sparsity": 0.8623080640589091, "compression_loss": 98.87114715576172, "distillation_loss": 3.59493350982666, "epoch": 3.77, "learning_rate": 4.158422041327489e-05, "loss": 102.8754, "step": 4463, "task_loss": 1.7760859727859497 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9314078606846552, "compression/movement_sparsity/importance_threshold": -0.00044517292169147724, "compression/movement_sparsity/linear_layer_sparsity": 0.8929836838275471, "compression/movement_sparsity/model_sparsity": 0.8623069471489371, "compression_loss": 98.88544464111328, "distillation_loss": 5.622312545776367, "epoch": 3.77, "learning_rate": 4.1581089542892924e-05, "loss": 103.3414, "step": 4464, "task_loss": 3.6341607570648193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9315492877314446, "compression/movement_sparsity/importance_threshold": -0.0004442550396680572, "compression/movement_sparsity/linear_layer_sparsity": 0.8930826424947572, "compression/movement_sparsity/model_sparsity": 0.862402506281496, "compression_loss": 98.89974212646484, "distillation_loss": 4.63020658493042, "epoch": 3.77, "learning_rate": 4.1577958672510956e-05, "loss": 103.4013, "step": 4465, "task_loss": 2.9918441772460938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9316905202431248, "compression/movement_sparsity/importance_threshold": -0.00044333842020568073, "compression/movement_sparsity/linear_layer_sparsity": 0.8930888192135925, "compression/movement_sparsity/model_sparsity": 0.8624084708110376, "compression_loss": 98.9139633178711, "distillation_loss": 3.2491817474365234, "epoch": 3.77, "learning_rate": 4.1574827802128994e-05, "loss": 103.0766, "step": 4466, "task_loss": 0.9399847984313965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9318315583535808, "compression/movement_sparsity/importance_threshold": -0.0004424230624354144, "compression/movement_sparsity/linear_layer_sparsity": 0.8931974364565877, "compression/movement_sparsity/model_sparsity": 0.86251335671759, "compression_loss": 98.92816925048828, "distillation_loss": 3.5754899978637695, "epoch": 3.78, "learning_rate": 4.1571696931747026e-05, "loss": 103.0083, "step": 4467, "task_loss": 2.222385883331299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9319724021966976, "compression/movement_sparsity/importance_threshold": -0.00044150896548832315, "compression/movement_sparsity/linear_layer_sparsity": 0.8932488057707629, "compression/movement_sparsity/model_sparsity": 0.8625629613377929, "compression_loss": 98.94235229492188, "distillation_loss": 3.746732234954834, "epoch": 3.78, "learning_rate": 4.156856606136506e-05, "loss": 103.175, "step": 4468, "task_loss": 2.5957627296447754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9321130519063607, "compression/movement_sparsity/importance_threshold": -0.0004405961284954735, "compression/movement_sparsity/linear_layer_sparsity": 0.893261409615954, "compression/movement_sparsity/model_sparsity": 0.8625751322021277, "compression_loss": 98.9565658569336, "distillation_loss": 5.249558925628662, "epoch": 3.78, "learning_rate": 4.1565435190983096e-05, "loss": 103.8044, "step": 4469, "task_loss": 3.0654940605163574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.932253507616455, "compression/movement_sparsity/importance_threshold": -0.00043968455058793123, "compression/movement_sparsity/linear_layer_sparsity": 0.8933244884627477, "compression/movement_sparsity/model_sparsity": 0.8626360440964808, "compression_loss": 98.97073364257812, "distillation_loss": 5.597637176513672, "epoch": 3.78, "learning_rate": 4.156230432060113e-05, "loss": 103.4417, "step": 4470, "task_loss": 3.0430383682250977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9323937694608657, "compression/movement_sparsity/importance_threshold": -0.00043877423089676206, "compression/movement_sparsity/linear_layer_sparsity": 0.8933505069965292, "compression/movement_sparsity/model_sparsity": 0.8626611688135845, "compression_loss": 98.98484802246094, "distillation_loss": 3.8404667377471924, "epoch": 3.78, "learning_rate": 4.1559173450219167e-05, "loss": 103.067, "step": 4471, "task_loss": 2.3698019981384277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9325338375734779, "compression/movement_sparsity/importance_threshold": -0.00043786516855303347, "compression/movement_sparsity/linear_layer_sparsity": 0.8934847015791032, "compression/movement_sparsity/model_sparsity": 0.8627907533994162, "compression_loss": 98.9989242553711, "distillation_loss": 6.135897636413574, "epoch": 3.78, "learning_rate": 4.15560425798372e-05, "loss": 103.6012, "step": 4472, "task_loss": 3.042968988418579 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9326737120881767, "compression/movement_sparsity/importance_threshold": -0.0004369573626878103, "compression/movement_sparsity/linear_layer_sparsity": 0.8936083075008165, "compression/movement_sparsity/model_sparsity": 0.8629101130774625, "compression_loss": 99.01300811767578, "distillation_loss": 4.69561767578125, "epoch": 3.78, "learning_rate": 4.155291170945523e-05, "loss": 103.534, "step": 4473, "task_loss": 1.9008699655532837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9328133931388474, "compression/movement_sparsity/importance_threshold": -0.0004360508124321592, "compression/movement_sparsity/linear_layer_sparsity": 0.8936496128175072, "compression/movement_sparsity/model_sparsity": 0.8629499994294547, "compression_loss": 99.027099609375, "distillation_loss": 4.835062026977539, "epoch": 3.78, "learning_rate": 4.154978083907327e-05, "loss": 103.2797, "step": 4474, "task_loss": 3.266671657562256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.932952880859375, "compression/movement_sparsity/importance_threshold": -0.00043514551691714587, "compression/movement_sparsity/linear_layer_sparsity": 0.8936927425318462, "compression/movement_sparsity/model_sparsity": 0.8629916475054236, "compression_loss": 99.04104614257812, "distillation_loss": 3.283834218978882, "epoch": 3.78, "learning_rate": 4.15466499686913e-05, "loss": 103.3851, "step": 4475, "task_loss": 1.1404616832733154 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9330921753836448, "compression/movement_sparsity/importance_threshold": -0.00043424147527383605, "compression/movement_sparsity/linear_layer_sparsity": 0.8937446603577327, "compression/movement_sparsity/model_sparsity": 0.863041781794273, "compression_loss": 99.0550537109375, "distillation_loss": 5.465261459350586, "epoch": 3.78, "learning_rate": 4.154351909830933e-05, "loss": 103.377, "step": 4476, "task_loss": 2.793916940689087 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9332312768455415, "compression/movement_sparsity/importance_threshold": -0.0004333386866332972, "compression/movement_sparsity/linear_layer_sparsity": 0.8938037088358655, "compression/movement_sparsity/model_sparsity": 0.8630988017755276, "compression_loss": 99.06903839111328, "distillation_loss": 3.74177885055542, "epoch": 3.78, "learning_rate": 4.154038822792737e-05, "loss": 103.5717, "step": 4477, "task_loss": 2.579580783843994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9333701853789509, "compression/movement_sparsity/importance_threshold": -0.0004324371501265934, "compression/movement_sparsity/linear_layer_sparsity": 0.8938302401088553, "compression/movement_sparsity/model_sparsity": 0.8631244216176704, "compression_loss": 99.0829849243164, "distillation_loss": 4.0651631355285645, "epoch": 3.78, "learning_rate": 4.15372573575454e-05, "loss": 102.8914, "step": 4478, "task_loss": 2.3306782245635986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9335089011177578, "compression/movement_sparsity/importance_threshold": -0.0004315368648847911, "compression/movement_sparsity/linear_layer_sparsity": 0.8938853774600035, "compression/movement_sparsity/model_sparsity": 0.8631776648311844, "compression_loss": 99.09693908691406, "distillation_loss": 3.923036813735962, "epoch": 3.79, "learning_rate": 4.1534126487163434e-05, "loss": 103.6869, "step": 4479, "task_loss": 2.1073946952819824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9336474241958472, "compression/movement_sparsity/importance_threshold": -0.00043063783003895705, "compression/movement_sparsity/linear_layer_sparsity": 0.8939400855411168, "compression/movement_sparsity/model_sparsity": 0.8632304935214098, "compression_loss": 99.11096954345703, "distillation_loss": 5.14935302734375, "epoch": 3.79, "learning_rate": 4.1530995616781466e-05, "loss": 103.6048, "step": 4480, "task_loss": 3.126893997192383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9337857547471043, "compression/movement_sparsity/importance_threshold": -0.00042974004472015777, "compression/movement_sparsity/linear_layer_sparsity": 0.8940060381123108, "compression/movement_sparsity/model_sparsity": 0.8632941804188894, "compression_loss": 99.12496948242188, "distillation_loss": 4.621941566467285, "epoch": 3.79, "learning_rate": 4.1527864746399505e-05, "loss": 103.5581, "step": 4481, "task_loss": 2.3639769554138184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9339238929054146, "compression/movement_sparsity/importance_threshold": -0.00042884350805945815, "compression/movement_sparsity/linear_layer_sparsity": 0.89411216320427, "compression/movement_sparsity/model_sparsity": 0.8633966597874608, "compression_loss": 99.13893127441406, "distillation_loss": 3.829904556274414, "epoch": 3.79, "learning_rate": 4.1524733876017536e-05, "loss": 103.3384, "step": 4482, "task_loss": 2.2156357765197754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9340618388046628, "compression/movement_sparsity/importance_threshold": -0.0004279482191879248, "compression/movement_sparsity/linear_layer_sparsity": 0.8942013082815157, "compression/movement_sparsity/model_sparsity": 0.8634827424570608, "compression_loss": 99.15289306640625, "distillation_loss": 3.5327649116516113, "epoch": 3.79, "learning_rate": 4.152160300563557e-05, "loss": 103.6933, "step": 4483, "task_loss": 1.7134519815444946 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9341995925787342, "compression/movement_sparsity/importance_threshold": -0.0004270541772366243, "compression/movement_sparsity/linear_layer_sparsity": 0.8943221954930081, "compression/movement_sparsity/model_sparsity": 0.8635994768209458, "compression_loss": 99.16687774658203, "distillation_loss": 4.055639266967773, "epoch": 3.79, "learning_rate": 4.15184721352536e-05, "loss": 103.0948, "step": 4484, "task_loss": 1.7365601062774658 }, { "compression/movement_sparsity/importance_regularization_factor": 0.934337154361514, "compression/movement_sparsity/importance_threshold": -0.0004261613813366207, "compression/movement_sparsity/linear_layer_sparsity": 0.8943548915606656, "compression/movement_sparsity/model_sparsity": 0.8636310496780945, "compression_loss": 99.18077087402344, "distillation_loss": 4.231998920440674, "epoch": 3.79, "learning_rate": 4.151534126487164e-05, "loss": 103.5486, "step": 4485, "task_loss": 2.820758581161499 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9344745242868874, "compression/movement_sparsity/importance_threshold": -0.0004252698306189814, "compression/movement_sparsity/linear_layer_sparsity": 0.8944099335184728, "compression/movement_sparsity/model_sparsity": 0.863684200775322, "compression_loss": 99.19465637207031, "distillation_loss": 4.707241058349609, "epoch": 3.79, "learning_rate": 4.151221039448967e-05, "loss": 103.1456, "step": 4486, "task_loss": 3.3994479179382324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9346117024887393, "compression/movement_sparsity/importance_threshold": -0.0004243795242147739, "compression/movement_sparsity/linear_layer_sparsity": 0.8944608854867808, "compression/movement_sparsity/model_sparsity": 0.8637334023867721, "compression_loss": 99.20853424072266, "distillation_loss": 5.219758987426758, "epoch": 3.79, "learning_rate": 4.15090795241077e-05, "loss": 103.7853, "step": 4487, "task_loss": 4.232064247131348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.934748689100955, "compression/movement_sparsity/importance_threshold": -0.0004234904612550614, "compression/movement_sparsity/linear_layer_sparsity": 0.8944888357357192, "compression/movement_sparsity/model_sparsity": 0.8637603924586745, "compression_loss": 99.22229766845703, "distillation_loss": 3.9475603103637695, "epoch": 3.79, "learning_rate": 4.150594865372574e-05, "loss": 102.9463, "step": 4488, "task_loss": 3.069196939468384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9348854842574196, "compression/movement_sparsity/importance_threshold": -0.0004226026408709121, "compression/movement_sparsity/linear_layer_sparsity": 0.8945546332927339, "compression/movement_sparsity/model_sparsity": 0.8638239296671888, "compression_loss": 99.23612213134766, "distillation_loss": 5.716420650482178, "epoch": 3.79, "learning_rate": 4.150281778334377e-05, "loss": 103.6993, "step": 4489, "task_loss": 2.2899162769317627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9350220880920183, "compression/movement_sparsity/importance_threshold": -0.000421716062193391, "compression/movement_sparsity/linear_layer_sparsity": 0.8946766413759841, "compression/movement_sparsity/model_sparsity": 0.8639417463974386, "compression_loss": 99.2499008178711, "distillation_loss": 3.531751871109009, "epoch": 3.79, "learning_rate": 4.1499686912961804e-05, "loss": 103.415, "step": 4490, "task_loss": 2.5611653327941895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9351585007386362, "compression/movement_sparsity/importance_threshold": -0.00042083072435356375, "compression/movement_sparsity/linear_layer_sparsity": 0.8948368664165072, "compression/movement_sparsity/model_sparsity": 0.8640964672149098, "compression_loss": 99.26371765136719, "distillation_loss": 3.6139590740203857, "epoch": 3.8, "learning_rate": 4.1496556042579836e-05, "loss": 103.3738, "step": 4491, "task_loss": 2.152653455734253 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9352947223311583, "compression/movement_sparsity/importance_threshold": -0.00041994662648249786, "compression/movement_sparsity/linear_layer_sparsity": 0.8948718638485185, "compression/movement_sparsity/model_sparsity": 0.8641302623774668, "compression_loss": 99.2774887084961, "distillation_loss": 5.428610801696777, "epoch": 3.8, "learning_rate": 4.1493425172197875e-05, "loss": 103.4499, "step": 4492, "task_loss": 3.5060925483703613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.93543075300347, "compression/movement_sparsity/importance_threshold": -0.0004190637677112582, "compression/movement_sparsity/linear_layer_sparsity": 0.895013964154235, "compression/movement_sparsity/model_sparsity": 0.8642674811005303, "compression_loss": 99.29115295410156, "distillation_loss": 5.494523048400879, "epoch": 3.8, "learning_rate": 4.1490294301815906e-05, "loss": 103.9601, "step": 4493, "task_loss": 2.7337887287139893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9355665928894563, "compression/movement_sparsity/importance_threshold": -0.00041818214717091046, "compression/movement_sparsity/linear_layer_sparsity": 0.895039016830438, "compression/movement_sparsity/model_sparsity": 0.8642916731402347, "compression_loss": 99.30488586425781, "distillation_loss": 3.989633083343506, "epoch": 3.8, "learning_rate": 4.148716343143394e-05, "loss": 103.1846, "step": 4494, "task_loss": 1.6610260009765625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9357022421230023, "compression/movement_sparsity/importance_threshold": -0.00041730176399252215, "compression/movement_sparsity/linear_layer_sparsity": 0.895118586801072, "compression/movement_sparsity/model_sparsity": 0.8643685096375916, "compression_loss": 99.31859588623047, "distillation_loss": 4.864150047302246, "epoch": 3.8, "learning_rate": 4.148403256105197e-05, "loss": 103.2267, "step": 4495, "task_loss": 2.2815091609954834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9358377008379933, "compression/movement_sparsity/importance_threshold": -0.00041642261730715814, "compression/movement_sparsity/linear_layer_sparsity": 0.8951761328340827, "compression/movement_sparsity/model_sparsity": 0.8644240787873361, "compression_loss": 99.33219909667969, "distillation_loss": 6.066831111907959, "epoch": 3.8, "learning_rate": 4.148090169067001e-05, "loss": 104.338, "step": 4496, "task_loss": 3.3299429416656494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9359729691683143, "compression/movement_sparsity/importance_threshold": -0.0004155447062458859, "compression/movement_sparsity/linear_layer_sparsity": 0.8951796981602058, "compression/movement_sparsity/model_sparsity": 0.8644275216335386, "compression_loss": 99.34584045410156, "distillation_loss": 2.8134894371032715, "epoch": 3.8, "learning_rate": 4.147777082028804e-05, "loss": 103.2007, "step": 4497, "task_loss": 1.627724528312683 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9361080472478505, "compression/movement_sparsity/importance_threshold": -0.0004146680299397694, "compression/movement_sparsity/linear_layer_sparsity": 0.8952712757676493, "compression/movement_sparsity/model_sparsity": 0.8645159532684407, "compression_loss": 99.3594741821289, "distillation_loss": 4.590512275695801, "epoch": 3.8, "learning_rate": 4.147463994990607e-05, "loss": 103.2655, "step": 4498, "task_loss": 2.4228553771972656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.936242935210487, "compression/movement_sparsity/importance_threshold": -0.00041379258751987613, "compression/movement_sparsity/linear_layer_sparsity": 0.8952809581917696, "compression/movement_sparsity/model_sparsity": 0.8645253030715059, "compression_loss": 99.37309265136719, "distillation_loss": 5.057433605194092, "epoch": 3.8, "learning_rate": 4.147150907952411e-05, "loss": 103.7737, "step": 4499, "task_loss": 2.084150552749634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9363776331901089, "compression/movement_sparsity/importance_threshold": -0.00041291837811727097, "compression/movement_sparsity/linear_layer_sparsity": 0.8952280387358017, "compression/movement_sparsity/model_sparsity": 0.8644742015616497, "compression_loss": 99.38665008544922, "distillation_loss": 3.54097318649292, "epoch": 3.8, "learning_rate": 4.146837820914214e-05, "loss": 103.5744, "step": 4500, "task_loss": 1.7322368621826172 }, { "epoch": 3.8, "eval_accuracy": 0.5845544554455445, "eval_loss": 103.1210708618164, "eval_runtime": 210.3364, "eval_samples_per_second": 120.046, "eval_steps_per_second": 0.941, "step": 4500 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9365121413206015, "compression/movement_sparsity/importance_threshold": -0.00041204540086302137, "compression/movement_sparsity/linear_layer_sparsity": 0.8952323314361506, "compression/movement_sparsity/model_sparsity": 0.8644783467945357, "compression_loss": 99.40016174316406, "distillation_loss": 5.261808395385742, "epoch": 3.8, "learning_rate": 4.1465247338760174e-05, "loss": 103.899, "step": 4501, "task_loss": 2.3957977294921875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9366464597358498, "compression/movement_sparsity/importance_threshold": -0.0004111736548881922, "compression/movement_sparsity/linear_layer_sparsity": 0.8952796703816649, "compression/movement_sparsity/model_sparsity": 0.86452405950164, "compression_loss": 99.41363525390625, "distillation_loss": 3.613737106323242, "epoch": 3.81, "learning_rate": 4.1462116468378206e-05, "loss": 103.6102, "step": 4502, "task_loss": 2.689242124557495 }, { "compression/movement_sparsity/importance_regularization_factor": 0.936780588569739, "compression/movement_sparsity/importance_threshold": -0.0004103031393238501, "compression/movement_sparsity/linear_layer_sparsity": 0.8953778539779786, "compression/movement_sparsity/model_sparsity": 0.8646188701893723, "compression_loss": 99.42716217041016, "distillation_loss": 4.073810577392578, "epoch": 3.81, "learning_rate": 4.1458985597996245e-05, "loss": 103.797, "step": 4503, "task_loss": 2.607189416885376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9369145279561542, "compression/movement_sparsity/importance_threshold": -0.0004094338533010616, "compression/movement_sparsity/linear_layer_sparsity": 0.895417954953738, "compression/movement_sparsity/model_sparsity": 0.8646575935732493, "compression_loss": 99.44070434570312, "distillation_loss": 3.822589874267578, "epoch": 3.81, "learning_rate": 4.1455854727614276e-05, "loss": 103.086, "step": 4504, "task_loss": 1.613885760307312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9370482780289806, "compression/movement_sparsity/importance_threshold": -0.00040856579595089076, "compression/movement_sparsity/linear_layer_sparsity": 0.8955035347048607, "compression/movement_sparsity/model_sparsity": 0.8647402333966467, "compression_loss": 99.4541244506836, "distillation_loss": 3.6968088150024414, "epoch": 3.81, "learning_rate": 4.1452723857232315e-05, "loss": 103.2327, "step": 4505, "task_loss": 1.9439810514450073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9371818389221033, "compression/movement_sparsity/importance_threshold": -0.00040769896640440676, "compression/movement_sparsity/linear_layer_sparsity": 0.8955605918469982, "compression/movement_sparsity/model_sparsity": 0.8647953304504237, "compression_loss": 99.46756744384766, "distillation_loss": 4.251734733581543, "epoch": 3.81, "learning_rate": 4.144959298685035e-05, "loss": 103.8147, "step": 4506, "task_loss": 2.1952884197235107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9373152107694074, "compression/movement_sparsity/importance_threshold": -0.00040683336379267274, "compression/movement_sparsity/linear_layer_sparsity": 0.8957116114301065, "compression/movement_sparsity/model_sparsity": 0.8649411620462615, "compression_loss": 99.4809799194336, "distillation_loss": 5.458251953125, "epoch": 3.81, "learning_rate": 4.1446462116468385e-05, "loss": 104.1068, "step": 4507, "task_loss": 3.1676697731018066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9374483937047781, "compression/movement_sparsity/importance_threshold": -0.0004059689872467562, "compression/movement_sparsity/linear_layer_sparsity": 0.895822732748305, "compression/movement_sparsity/model_sparsity": 0.8650484660053308, "compression_loss": 99.49435424804688, "distillation_loss": 4.807710647583008, "epoch": 3.81, "learning_rate": 4.144333124608642e-05, "loss": 104.1579, "step": 4508, "task_loss": 2.6813502311706543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9375813878621005, "compression/movement_sparsity/importance_threshold": -0.0004051058358977228, "compression/movement_sparsity/linear_layer_sparsity": 0.8959035786048762, "compression/movement_sparsity/model_sparsity": 0.8651265345580178, "compression_loss": 99.50768280029297, "distillation_loss": 6.0586347579956055, "epoch": 3.81, "learning_rate": 4.144020037570445e-05, "loss": 104.2858, "step": 4509, "task_loss": 2.9718077182769775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9377141933752599, "compression/movement_sparsity/importance_threshold": -0.0004042439088766383, "compression/movement_sparsity/linear_layer_sparsity": 0.8959849848973263, "compression/movement_sparsity/model_sparsity": 0.8652051442938872, "compression_loss": 99.52103424072266, "distillation_loss": 5.754048824310303, "epoch": 3.81, "learning_rate": 4.143706950532248e-05, "loss": 103.4928, "step": 4510, "task_loss": 2.371142864227295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9378468103781411, "compression/movement_sparsity/importance_threshold": -0.00040338320531457025, "compression/movement_sparsity/linear_layer_sparsity": 0.8959650119165362, "compression/movement_sparsity/model_sparsity": 0.8651858574464313, "compression_loss": 99.53422546386719, "distillation_loss": 5.329265117645264, "epoch": 3.81, "learning_rate": 4.143393863494052e-05, "loss": 104.1329, "step": 4511, "task_loss": 2.609250068664551 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9379792390046297, "compression/movement_sparsity/importance_threshold": -0.0004025237243425817, "compression/movement_sparsity/linear_layer_sparsity": 0.8959999258793739, "compression/movement_sparsity/model_sparsity": 0.8652195720072378, "compression_loss": 99.54750061035156, "distillation_loss": 3.084571123123169, "epoch": 3.81, "learning_rate": 4.143080776455855e-05, "loss": 103.3043, "step": 4512, "task_loss": 2.57709002494812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9381114793886104, "compression/movement_sparsity/importance_threshold": -0.00040166546509174186, "compression/movement_sparsity/linear_layer_sparsity": 0.8960387271208611, "compression/movement_sparsity/model_sparsity": 0.8652570403067131, "compression_loss": 99.5606918334961, "distillation_loss": 5.930343151092529, "epoch": 3.81, "learning_rate": 4.142767689417658e-05, "loss": 104.0844, "step": 4513, "task_loss": 2.5774576663970947 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9382435316639686, "compression/movement_sparsity/importance_threshold": -0.00040080842669311476, "compression/movement_sparsity/linear_layer_sparsity": 0.8961936459067862, "compression/movement_sparsity/model_sparsity": 0.8654066371557558, "compression_loss": 99.5739517211914, "distillation_loss": 5.6854705810546875, "epoch": 3.82, "learning_rate": 4.142454602379462e-05, "loss": 105.2846, "step": 4514, "task_loss": 2.750324249267578 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9383753959645893, "compression/movement_sparsity/importance_threshold": -0.00039995260827776786, "compression/movement_sparsity/linear_layer_sparsity": 0.8962645470075491, "compression/movement_sparsity/model_sparsity": 0.8654751025855901, "compression_loss": 99.58717346191406, "distillation_loss": 4.3391265869140625, "epoch": 3.82, "learning_rate": 4.142141515341265e-05, "loss": 103.77, "step": 4515, "task_loss": 2.3447625637054443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9385070724243578, "compression/movement_sparsity/importance_threshold": -0.000399098008976766, "compression/movement_sparsity/linear_layer_sparsity": 0.8962963368384662, "compression/movement_sparsity/model_sparsity": 0.8655058003380184, "compression_loss": 99.6003646850586, "distillation_loss": 3.403304100036621, "epoch": 3.82, "learning_rate": 4.1418284283030685e-05, "loss": 103.1246, "step": 4516, "task_loss": 1.5900479555130005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9386385611771592, "compression/movement_sparsity/importance_threshold": -0.000398244627921175, "compression/movement_sparsity/linear_layer_sparsity": 0.8963946158281211, "compression/movement_sparsity/model_sparsity": 0.865600703142037, "compression_loss": 99.61360931396484, "distillation_loss": 3.952195167541504, "epoch": 3.82, "learning_rate": 4.1415153412648717e-05, "loss": 103.8213, "step": 4517, "task_loss": 1.780625581741333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9387698623568784, "compression/movement_sparsity/importance_threshold": -0.0003973924642420622, "compression/movement_sparsity/linear_layer_sparsity": 0.8964542366663003, "compression/movement_sparsity/model_sparsity": 0.8656582758210097, "compression_loss": 99.62678527832031, "distillation_loss": 5.036187171936035, "epoch": 3.82, "learning_rate": 4.1412022542266755e-05, "loss": 104.3613, "step": 4518, "task_loss": 1.9454435110092163 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9389009760974009, "compression/movement_sparsity/importance_threshold": -0.00039654151707049345, "compression/movement_sparsity/linear_layer_sparsity": 0.8965298120407764, "compression/movement_sparsity/model_sparsity": 0.8657312549488755, "compression_loss": 99.63996124267578, "distillation_loss": 3.2574825286865234, "epoch": 3.82, "learning_rate": 4.140889167188479e-05, "loss": 103.5242, "step": 4519, "task_loss": 1.916933536529541 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9390319025326116, "compression/movement_sparsity/importance_threshold": -0.00039569178553753356, "compression/movement_sparsity/linear_layer_sparsity": 0.8966541214883803, "compression/movement_sparsity/model_sparsity": 0.8658512939845336, "compression_loss": 99.65312194824219, "distillation_loss": 4.220731735229492, "epoch": 3.82, "learning_rate": 4.140576080150282e-05, "loss": 104.2491, "step": 4520, "task_loss": 1.577354073524475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9391626417963957, "compression/movement_sparsity/importance_threshold": -0.00039484326877425, "compression/movement_sparsity/linear_layer_sparsity": 0.8966541453367156, "compression/movement_sparsity/model_sparsity": 0.8658513170136052, "compression_loss": 99.66622924804688, "distillation_loss": 4.958728313446045, "epoch": 3.82, "learning_rate": 4.140262993112085e-05, "loss": 104.0817, "step": 4521, "task_loss": 3.313663959503174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9392931940226384, "compression/movement_sparsity/importance_threshold": -0.00039399596591170767, "compression/movement_sparsity/linear_layer_sparsity": 0.896668442413711, "compression/movement_sparsity/model_sparsity": 0.8658651229420229, "compression_loss": 99.67936706542969, "distillation_loss": 3.7188241481781006, "epoch": 3.82, "learning_rate": 4.139949906073889e-05, "loss": 103.6273, "step": 4522, "task_loss": 1.6565232276916504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9394235593452248, "compression/movement_sparsity/importance_threshold": -0.00039314987608097315, "compression/movement_sparsity/linear_layer_sparsity": 0.8967161748567574, "compression/movement_sparsity/model_sparsity": 0.8659112156288084, "compression_loss": 99.69247436523438, "distillation_loss": 5.640217304229736, "epoch": 3.82, "learning_rate": 4.139636819035692e-05, "loss": 104.5955, "step": 4523, "task_loss": 2.803363800048828 }, { "compression/movement_sparsity/importance_regularization_factor": 0.93955373789804, "compression/movement_sparsity/importance_threshold": -0.00039230499841311304, "compression/movement_sparsity/linear_layer_sparsity": 0.8967791225377071, "compression/movement_sparsity/model_sparsity": 0.8659720008632678, "compression_loss": 99.70559692382812, "distillation_loss": 4.650172233581543, "epoch": 3.82, "learning_rate": 4.139323731997495e-05, "loss": 104.868, "step": 4524, "task_loss": 3.247243881225586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9396837298149692, "compression/movement_sparsity/importance_threshold": -0.0003914613320391922, "compression/movement_sparsity/linear_layer_sparsity": 0.8968520984436386, "compression/movement_sparsity/model_sparsity": 0.8660424698223304, "compression_loss": 99.71873474121094, "distillation_loss": 4.9277496337890625, "epoch": 3.82, "learning_rate": 4.139010644959299e-05, "loss": 104.2309, "step": 4525, "task_loss": 2.6433651447296143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9398135352298975, "compression/movement_sparsity/importance_threshold": -0.0003906188760902773, "compression/movement_sparsity/linear_layer_sparsity": 0.8969409692650286, "compression/movement_sparsity/model_sparsity": 0.8661282876576071, "compression_loss": 99.73179626464844, "distillation_loss": 3.7554702758789062, "epoch": 3.83, "learning_rate": 4.138697557921102e-05, "loss": 104.2121, "step": 4526, "task_loss": 2.212545871734619 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9399431542767102, "compression/movement_sparsity/importance_threshold": -0.00038977762969743396, "compression/movement_sparsity/linear_layer_sparsity": 0.8969943537635344, "compression/movement_sparsity/model_sparsity": 0.8661798382343593, "compression_loss": 99.74490356445312, "distillation_loss": 4.692024230957031, "epoch": 3.83, "learning_rate": 4.1383844708829055e-05, "loss": 103.819, "step": 4527, "task_loss": 3.0202016830444336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9400725870892921, "compression/movement_sparsity/importance_threshold": -0.00038893759199172886, "compression/movement_sparsity/linear_layer_sparsity": 0.8971558189174916, "compression/movement_sparsity/model_sparsity": 0.8663357565635531, "compression_loss": 99.75799560546875, "distillation_loss": 2.4251763820648193, "epoch": 3.83, "learning_rate": 4.1380713838447087e-05, "loss": 103.6276, "step": 4528, "task_loss": 2.419297456741333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9402018338015287, "compression/movement_sparsity/importance_threshold": -0.00038809876210422773, "compression/movement_sparsity/linear_layer_sparsity": 0.8971788325610288, "compression/movement_sparsity/model_sparsity": 0.8663579796176366, "compression_loss": 99.77102661132812, "distillation_loss": 4.812941551208496, "epoch": 3.83, "learning_rate": 4.1377582968065125e-05, "loss": 104.4546, "step": 4529, "task_loss": 2.3952527046203613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9403308945473048, "compression/movement_sparsity/importance_threshold": -0.00038726113916599716, "compression/movement_sparsity/linear_layer_sparsity": 0.8972307026902449, "compression/movement_sparsity/model_sparsity": 0.8664080678483428, "compression_loss": 99.7840576171875, "distillation_loss": 5.288850784301758, "epoch": 3.83, "learning_rate": 4.137445209768316e-05, "loss": 104.242, "step": 4530, "task_loss": 3.619065523147583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9404597694605058, "compression/movement_sparsity/importance_threshold": -0.000386424722308102, "compression/movement_sparsity/linear_layer_sparsity": 0.897285708875549, "compression/movement_sparsity/model_sparsity": 0.8664611844019631, "compression_loss": 99.7970962524414, "distillation_loss": 4.36412239074707, "epoch": 3.83, "learning_rate": 4.137132122730119e-05, "loss": 104.1718, "step": 4531, "task_loss": 2.793273687362671 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9405884586750167, "compression/movement_sparsity/importance_threshold": -0.00038558951066160977, "compression/movement_sparsity/linear_layer_sparsity": 0.8973018422743604, "compression/movement_sparsity/model_sparsity": 0.866476763568893, "compression_loss": 99.81007385253906, "distillation_loss": 2.4807024002075195, "epoch": 3.83, "learning_rate": 4.136819035691922e-05, "loss": 103.77, "step": 4532, "task_loss": 0.7518596053123474 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9407169623247228, "compression/movement_sparsity/importance_threshold": -0.0003847555033575844, "compression/movement_sparsity/linear_layer_sparsity": 0.8973112623667927, "compression/movement_sparsity/model_sparsity": 0.8664858600521708, "compression_loss": 99.8230209350586, "distillation_loss": 4.341029167175293, "epoch": 3.83, "learning_rate": 4.136505948653726e-05, "loss": 104.6602, "step": 4533, "task_loss": 1.3949520587921143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9408452805435091, "compression/movement_sparsity/importance_threshold": -0.0003839226995270943, "compression/movement_sparsity/linear_layer_sparsity": 0.8974687210004243, "compression/movement_sparsity/model_sparsity": 0.8666379094973377, "compression_loss": 99.83597564697266, "distillation_loss": 3.3386058807373047, "epoch": 3.83, "learning_rate": 4.136192861615529e-05, "loss": 104.4929, "step": 4534, "task_loss": 1.7800004482269287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9409734134652609, "compression/movement_sparsity/importance_threshold": -0.0003830910983012043, "compression/movement_sparsity/linear_layer_sparsity": 0.8975224989964621, "compression/movement_sparsity/model_sparsity": 0.866689840053771, "compression_loss": 99.84892272949219, "distillation_loss": 5.5670294761657715, "epoch": 3.83, "learning_rate": 4.135879774577332e-05, "loss": 104.3963, "step": 4535, "task_loss": 3.430874824523926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9411013612238631, "compression/movement_sparsity/importance_threshold": -0.00038226069881098015, "compression/movement_sparsity/linear_layer_sparsity": 0.8976079714300759, "compression/movement_sparsity/model_sparsity": 0.8667723762463463, "compression_loss": 99.86177825927734, "distillation_loss": 5.224770545959473, "epoch": 3.83, "learning_rate": 4.135566687539136e-05, "loss": 103.7354, "step": 4536, "task_loss": 2.0580203533172607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.941229123953201, "compression/movement_sparsity/importance_threshold": -0.00038143150018748843, "compression/movement_sparsity/linear_layer_sparsity": 0.8976311043152895, "compression/movement_sparsity/model_sparsity": 0.8667947144457877, "compression_loss": 99.87467193603516, "distillation_loss": 3.8436005115509033, "epoch": 3.83, "learning_rate": 4.135253600500939e-05, "loss": 104.2316, "step": 4537, "task_loss": 1.3202227354049683 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9413567017871597, "compression/movement_sparsity/importance_threshold": -0.0003806035015617949, "compression/movement_sparsity/linear_layer_sparsity": 0.8976739597737727, "compression/movement_sparsity/model_sparsity": 0.8668360976874333, "compression_loss": 99.88752746582031, "distillation_loss": 4.008355140686035, "epoch": 3.84, "learning_rate": 4.134940513462743e-05, "loss": 104.0785, "step": 4538, "task_loss": 2.2898292541503906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9414840948596243, "compression/movement_sparsity/importance_threshold": -0.000379776702064967, "compression/movement_sparsity/linear_layer_sparsity": 0.8977142873087173, "compression/movement_sparsity/model_sparsity": 0.8668750398474905, "compression_loss": 99.90036010742188, "distillation_loss": 5.604606628417969, "epoch": 3.84, "learning_rate": 4.134627426424546e-05, "loss": 104.886, "step": 4539, "task_loss": 2.8869292736053467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9416113033044801, "compression/movement_sparsity/importance_threshold": -0.0003789511008280679, "compression/movement_sparsity/linear_layer_sparsity": 0.8977356554171207, "compression/movement_sparsity/model_sparsity": 0.8668956738956343, "compression_loss": 99.91320037841797, "distillation_loss": 4.1502790451049805, "epoch": 3.84, "learning_rate": 4.1343143393863495e-05, "loss": 103.9656, "step": 4540, "task_loss": 2.759917974472046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.941738327255612, "compression/movement_sparsity/importance_threshold": -0.0003781266969821659, "compression/movement_sparsity/linear_layer_sparsity": 0.8977763883737648, "compression/movement_sparsity/model_sparsity": 0.8669350075499084, "compression_loss": 99.92605590820312, "distillation_loss": 5.73134183883667, "epoch": 3.84, "learning_rate": 4.1340012523481534e-05, "loss": 103.8283, "step": 4541, "task_loss": 3.781029224395752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9418651668469054, "compression/movement_sparsity/importance_threshold": -0.00037730348965832675, "compression/movement_sparsity/linear_layer_sparsity": 0.8978145695585349, "compression/movement_sparsity/model_sparsity": 0.8669718770935225, "compression_loss": 99.93888092041016, "distillation_loss": 3.626842498779297, "epoch": 3.84, "learning_rate": 4.1336881653099565e-05, "loss": 103.9691, "step": 4542, "task_loss": 1.3053442239761353 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9419918222122452, "compression/movement_sparsity/importance_threshold": -0.0003764814779876153, "compression/movement_sparsity/linear_layer_sparsity": 0.8979005189588541, "compression/movement_sparsity/model_sparsity": 0.8670548738675296, "compression_loss": 99.95175170898438, "distillation_loss": 4.115375518798828, "epoch": 3.84, "learning_rate": 4.13337507827176e-05, "loss": 104.4646, "step": 4543, "task_loss": 3.405061721801758 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9421182934855167, "compression/movement_sparsity/importance_threshold": -0.00037566066110109996, "compression/movement_sparsity/linear_layer_sparsity": 0.897969357178616, "compression/movement_sparsity/model_sparsity": 0.8671213472826714, "compression_loss": 99.96453094482422, "distillation_loss": 4.602190017700195, "epoch": 3.84, "learning_rate": 4.1330619912335636e-05, "loss": 104.3347, "step": 4544, "task_loss": 3.1808714866638184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9422445808006049, "compression/movement_sparsity/importance_threshold": -0.00037484103812984466, "compression/movement_sparsity/linear_layer_sparsity": 0.8979742818598496, "compression/movement_sparsity/model_sparsity": 0.8671261027859546, "compression_loss": 99.97740173339844, "distillation_loss": 4.896975517272949, "epoch": 3.84, "learning_rate": 4.132748904195367e-05, "loss": 105.3298, "step": 4545, "task_loss": 1.8886102437973022 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9423706842913951, "compression/movement_sparsity/importance_threshold": -0.00037402260820491515, "compression/movement_sparsity/linear_layer_sparsity": 0.8979507197046012, "compression/movement_sparsity/model_sparsity": 0.8671033500632246, "compression_loss": 99.9901351928711, "distillation_loss": 4.649235725402832, "epoch": 3.84, "learning_rate": 4.13243581715717e-05, "loss": 105.2147, "step": 4546, "task_loss": 3.061427116394043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9424966040917724, "compression/movement_sparsity/importance_threshold": -0.0003732053704573789, "compression/movement_sparsity/linear_layer_sparsity": 0.8980508827127424, "compression/movement_sparsity/model_sparsity": 0.8672000721638987, "compression_loss": 100.00291442871094, "distillation_loss": 5.351968765258789, "epoch": 3.84, "learning_rate": 4.132122730118973e-05, "loss": 104.5068, "step": 4547, "task_loss": 2.408273935317993 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9426223403356219, "compression/movement_sparsity/importance_threshold": -0.00037238932401830165, "compression/movement_sparsity/linear_layer_sparsity": 0.8981358662554833, "compression/movement_sparsity/model_sparsity": 0.8672821362605064, "compression_loss": 100.01567077636719, "distillation_loss": 4.905787467956543, "epoch": 3.84, "learning_rate": 4.131809643080777e-05, "loss": 104.5165, "step": 4548, "task_loss": 3.524320125579834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9427478931568287, "compression/movement_sparsity/importance_threshold": -0.00037157446801874914, "compression/movement_sparsity/linear_layer_sparsity": 0.8981467053238642, "compression/movement_sparsity/model_sparsity": 0.8672926029735437, "compression_loss": 100.02838134765625, "distillation_loss": 3.9205758571624756, "epoch": 3.84, "learning_rate": 4.13149655604258e-05, "loss": 104.4322, "step": 4549, "task_loss": 2.078174591064453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9428732626892781, "compression/movement_sparsity/importance_threshold": -0.0003707608015897871, "compression/movement_sparsity/linear_layer_sparsity": 0.8982612846506772, "compression/movement_sparsity/model_sparsity": 0.8674032461479934, "compression_loss": 100.0409927368164, "distillation_loss": 4.900690078735352, "epoch": 3.85, "learning_rate": 4.131183469004383e-05, "loss": 103.9821, "step": 4550, "task_loss": 1.6776789426803589 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9429984490668549, "compression/movement_sparsity/importance_threshold": -0.000369948323862483, "compression/movement_sparsity/linear_layer_sparsity": 0.8983195580579136, "compression/movement_sparsity/model_sparsity": 0.8674595176844213, "compression_loss": 100.05361938476562, "distillation_loss": 3.9833860397338867, "epoch": 3.85, "learning_rate": 4.130870381966187e-05, "loss": 104.5564, "step": 4551, "task_loss": 2.3233702182769775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9431234524234448, "compression/movement_sparsity/importance_threshold": -0.0003691370339679008, "compression/movement_sparsity/linear_layer_sparsity": 0.8983713327937886, "compression/movement_sparsity/model_sparsity": 0.8675095137988412, "compression_loss": 100.06617736816406, "distillation_loss": 4.009873390197754, "epoch": 3.85, "learning_rate": 4.1305572949279903e-05, "loss": 104.141, "step": 4552, "task_loss": 2.5289177894592285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9432482728929324, "compression/movement_sparsity/importance_threshold": -0.0003683269310371072, "compression/movement_sparsity/linear_layer_sparsity": 0.8984311324944825, "compression/movement_sparsity/model_sparsity": 0.8675672591958509, "compression_loss": 100.07875061035156, "distillation_loss": 4.01463508605957, "epoch": 3.85, "learning_rate": 4.1302442078897935e-05, "loss": 104.8957, "step": 4553, "task_loss": 2.5166492462158203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.943372910609203, "compression/movement_sparsity/importance_threshold": -0.00036751801420117046, "compression/movement_sparsity/linear_layer_sparsity": 0.8984823706428138, "compression/movement_sparsity/model_sparsity": 0.86761673715616, "compression_loss": 100.09130096435547, "distillation_loss": 5.691941261291504, "epoch": 3.85, "learning_rate": 4.129931120851597e-05, "loss": 105.0518, "step": 4554, "task_loss": 2.3328723907470703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.943497365706142, "compression/movement_sparsity/importance_threshold": -0.00036671028259115286, "compression/movement_sparsity/linear_layer_sparsity": 0.8986325555341874, "compression/movement_sparsity/model_sparsity": 0.8677617627344922, "compression_loss": 100.1038589477539, "distillation_loss": 4.320714950561523, "epoch": 3.85, "learning_rate": 4.1296180338134006e-05, "loss": 104.1126, "step": 4555, "task_loss": 3.6645591259002686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9436216383176342, "compression/movement_sparsity/importance_threshold": -0.00036590373533812275, "compression/movement_sparsity/linear_layer_sparsity": 0.8987534308215122, "compression/movement_sparsity/model_sparsity": 0.8678784855838415, "compression_loss": 100.11631774902344, "distillation_loss": 4.818197250366211, "epoch": 3.85, "learning_rate": 4.129304946775204e-05, "loss": 104.1675, "step": 4556, "task_loss": 2.3907017707824707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9437457285775649, "compression/movement_sparsity/importance_threshold": -0.0003650983715731467, "compression/movement_sparsity/linear_layer_sparsity": 0.8988051936332194, "compression/movement_sparsity/model_sparsity": 0.8679284701837255, "compression_loss": 100.1288070678711, "distillation_loss": 4.874901294708252, "epoch": 3.85, "learning_rate": 4.128991859737007e-05, "loss": 104.3821, "step": 4557, "task_loss": 2.270557403564453 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9438696366198193, "compression/movement_sparsity/importance_threshold": -0.00036429419042728877, "compression/movement_sparsity/linear_layer_sparsity": 0.8988432794246484, "compression/movement_sparsity/model_sparsity": 0.8679652476110533, "compression_loss": 100.1412124633789, "distillation_loss": 4.5893940925598145, "epoch": 3.85, "learning_rate": 4.12867877269881e-05, "loss": 105.0313, "step": 4558, "task_loss": 2.685469150543213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9439933625782824, "compression/movement_sparsity/importance_threshold": -0.0003634911910316164, "compression/movement_sparsity/linear_layer_sparsity": 0.8988861825798022, "compression/movement_sparsity/model_sparsity": 0.8680066769108421, "compression_loss": 100.15364837646484, "distillation_loss": 4.357693672180176, "epoch": 3.85, "learning_rate": 4.128365685660614e-05, "loss": 105.1541, "step": 4559, "task_loss": 2.295363426208496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9441169065868393, "compression/movement_sparsity/importance_threshold": -0.00036268937251719615, "compression/movement_sparsity/linear_layer_sparsity": 0.898944312897027, "compression/movement_sparsity/model_sparsity": 0.8680628102728405, "compression_loss": 100.16596984863281, "distillation_loss": 5.853177070617676, "epoch": 3.85, "learning_rate": 4.128052598622417e-05, "loss": 105.9258, "step": 4560, "task_loss": 2.8003523349761963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9442402687793753, "compression/movement_sparsity/importance_threshold": -0.00036188873401509294, "compression/movement_sparsity/linear_layer_sparsity": 0.8989871683555104, "compression/movement_sparsity/model_sparsity": 0.8681041935144861, "compression_loss": 100.17832946777344, "distillation_loss": 4.245377540588379, "epoch": 3.85, "learning_rate": 4.12773951158422e-05, "loss": 103.8754, "step": 4561, "task_loss": 1.61696457862854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9443634492897756, "compression/movement_sparsity/importance_threshold": -0.00036108927465637163, "compression/movement_sparsity/linear_layer_sparsity": 0.8990208899015846, "compression/movement_sparsity/model_sparsity": 0.8681367566217131, "compression_loss": 100.19066619873047, "distillation_loss": 4.66460657119751, "epoch": 3.86, "learning_rate": 4.127426424546024e-05, "loss": 103.7898, "step": 4562, "task_loss": 2.746424913406372 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9444864482519251, "compression/movement_sparsity/importance_threshold": -0.00036029099357210055, "compression/movement_sparsity/linear_layer_sparsity": 0.8991120382389931, "compression/movement_sparsity/model_sparsity": 0.8682247737333265, "compression_loss": 100.20296478271484, "distillation_loss": 3.660703420639038, "epoch": 3.86, "learning_rate": 4.1271133375078273e-05, "loss": 103.5956, "step": 4563, "task_loss": 1.2595893144607544 }, { "compression/movement_sparsity/importance_regularization_factor": 0.944609265799709, "compression/movement_sparsity/importance_threshold": -0.00035949388989334544, "compression/movement_sparsity/linear_layer_sparsity": 0.8991426475773144, "compression/movement_sparsity/model_sparsity": 0.8682543315467111, "compression_loss": 100.21521759033203, "distillation_loss": 5.062169075012207, "epoch": 3.86, "learning_rate": 4.1268002504696305e-05, "loss": 104.7582, "step": 4564, "task_loss": 3.332879066467285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9447319020670126, "compression/movement_sparsity/importance_threshold": -0.00035869796275117116, "compression/movement_sparsity/linear_layer_sparsity": 0.8992251389690193, "compression/movement_sparsity/model_sparsity": 0.8683339891053378, "compression_loss": 100.22746276855469, "distillation_loss": 4.148728847503662, "epoch": 3.86, "learning_rate": 4.126487163431434e-05, "loss": 104.5881, "step": 4565, "task_loss": 0.9361843466758728 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9448543571877209, "compression/movement_sparsity/importance_threshold": -0.0003579032112766443, "compression/movement_sparsity/linear_layer_sparsity": 0.8992721678861751, "compression/movement_sparsity/model_sparsity": 0.8683794024345114, "compression_loss": 100.23966979980469, "distillation_loss": 4.7280073165893555, "epoch": 3.86, "learning_rate": 4.1261740763932376e-05, "loss": 104.3875, "step": 4566, "task_loss": 2.028907299041748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9449766312957192, "compression/movement_sparsity/importance_threshold": -0.0003571096346008315, "compression/movement_sparsity/linear_layer_sparsity": 0.8993758246754336, "compression/movement_sparsity/model_sparsity": 0.8684794982941734, "compression_loss": 100.25194549560547, "distillation_loss": 4.436640739440918, "epoch": 3.86, "learning_rate": 4.125860989355041e-05, "loss": 104.7009, "step": 4567, "task_loss": 3.1713123321533203 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9450987245248924, "compression/movement_sparsity/importance_threshold": -0.0003563172318547976, "compression/movement_sparsity/linear_layer_sparsity": 0.8994896885521886, "compression/movement_sparsity/model_sparsity": 0.8685894505964754, "compression_loss": 100.26408386230469, "distillation_loss": 5.375784873962402, "epoch": 3.86, "learning_rate": 4.125547902316844e-05, "loss": 105.1612, "step": 4568, "task_loss": 2.9243879318237305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9452206370091258, "compression/movement_sparsity/importance_threshold": -0.0003555260021696092, "compression/movement_sparsity/linear_layer_sparsity": 0.8995513841955365, "compression/movement_sparsity/model_sparsity": 0.8686490268046764, "compression_loss": 100.27619171142578, "distillation_loss": 3.658583641052246, "epoch": 3.86, "learning_rate": 4.125234815278647e-05, "loss": 104.612, "step": 4569, "task_loss": 1.1587982177734375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9453423688823045, "compression/movement_sparsity/importance_threshold": -0.0003547359446763338, "compression/movement_sparsity/linear_layer_sparsity": 0.8996039220781401, "compression/movement_sparsity/model_sparsity": 0.8686997598493871, "compression_loss": 100.288330078125, "distillation_loss": 4.0002875328063965, "epoch": 3.86, "learning_rate": 4.124921728240451e-05, "loss": 104.5945, "step": 4570, "task_loss": 2.273941993713379 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9454639202783137, "compression/movement_sparsity/importance_threshold": -0.0003539470585060345, "compression/movement_sparsity/linear_layer_sparsity": 0.8996744773780415, "compression/movement_sparsity/model_sparsity": 0.8687678913576834, "compression_loss": 100.3004379272461, "distillation_loss": 5.991607189178467, "epoch": 3.86, "learning_rate": 4.124608641202254e-05, "loss": 105.5962, "step": 4571, "task_loss": 4.033348083496094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9455852913310384, "compression/movement_sparsity/importance_threshold": -0.00035315934278977966, "compression/movement_sparsity/linear_layer_sparsity": 0.8997737699219454, "compression/movement_sparsity/model_sparsity": 0.8688637728972446, "compression_loss": 100.31263732910156, "distillation_loss": 5.201147079467773, "epoch": 3.86, "learning_rate": 4.124295554164058e-05, "loss": 105.0856, "step": 4572, "task_loss": 2.4417216777801514 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9457064821743638, "compression/movement_sparsity/importance_threshold": -0.000352372796658635, "compression/movement_sparsity/linear_layer_sparsity": 0.8998573583370727, "compression/movement_sparsity/model_sparsity": 0.8689444897931643, "compression_loss": 100.32476043701172, "distillation_loss": 3.877317428588867, "epoch": 3.87, "learning_rate": 4.123982467125861e-05, "loss": 104.9015, "step": 4573, "task_loss": 2.971978187561035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9458274929421752, "compression/movement_sparsity/importance_threshold": -0.0003515874192436654, "compression/movement_sparsity/linear_layer_sparsity": 0.8999612655338517, "compression/movement_sparsity/model_sparsity": 0.8690448274580779, "compression_loss": 100.33687591552734, "distillation_loss": 3.4000258445739746, "epoch": 3.87, "learning_rate": 4.123669380087665e-05, "loss": 104.8266, "step": 4574, "task_loss": 1.7719323635101318 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9459483237683575, "compression/movement_sparsity/importance_threshold": -0.0003508032096759383, "compression/movement_sparsity/linear_layer_sparsity": 0.8999911474979472, "compression/movement_sparsity/model_sparsity": 0.869073682884779, "compression_loss": 100.3489761352539, "distillation_loss": 3.732917070388794, "epoch": 3.87, "learning_rate": 4.123356293049468e-05, "loss": 104.3986, "step": 4575, "task_loss": 2.575914144515991 }, { "compression/movement_sparsity/importance_regularization_factor": 0.946068974786796, "compression/movement_sparsity/importance_threshold": -0.0003500201670865186, "compression/movement_sparsity/linear_layer_sparsity": 0.9000782177700242, "compression/movement_sparsity/model_sparsity": 0.8691577620251507, "compression_loss": 100.36109924316406, "distillation_loss": 4.489272117614746, "epoch": 3.87, "learning_rate": 4.1230432060112714e-05, "loss": 104.2431, "step": 4576, "task_loss": 2.3525326251983643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9461894461313758, "compression/movement_sparsity/importance_threshold": -0.0003492382906064729, "compression/movement_sparsity/linear_layer_sparsity": 0.9001487969182609, "compression/movement_sparsity/model_sparsity": 0.8692259165625187, "compression_loss": 100.37312316894531, "distillation_loss": 4.997652053833008, "epoch": 3.87, "learning_rate": 4.122730118973075e-05, "loss": 104.3963, "step": 4577, "task_loss": 3.355085611343384 }, { "compression/movement_sparsity/importance_regularization_factor": 0.946309737935982, "compression/movement_sparsity/importance_threshold": -0.0003484575793668669, "compression/movement_sparsity/linear_layer_sparsity": 0.900218934872295, "compression/movement_sparsity/model_sparsity": 0.8692936450620622, "compression_loss": 100.38511657714844, "distillation_loss": 4.706137657165527, "epoch": 3.87, "learning_rate": 4.1224170319348784e-05, "loss": 104.9286, "step": 4578, "task_loss": 2.4296185970306396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9464298503344998, "compression/movement_sparsity/importance_threshold": -0.00034767803249876727, "compression/movement_sparsity/linear_layer_sparsity": 0.900243462885122, "compression/movement_sparsity/model_sparsity": 0.8693173304621915, "compression_loss": 100.39705657958984, "distillation_loss": 4.33980655670166, "epoch": 3.87, "learning_rate": 4.1221039448966816e-05, "loss": 105.2083, "step": 4579, "task_loss": 2.0307490825653076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9465497834608143, "compression/movement_sparsity/importance_threshold": -0.0003468996491332397, "compression/movement_sparsity/linear_layer_sparsity": 0.9003223293298656, "compression/movement_sparsity/model_sparsity": 0.8693934876019366, "compression_loss": 100.40902709960938, "distillation_loss": 3.6383535861968994, "epoch": 3.87, "learning_rate": 4.121790857858485e-05, "loss": 104.3387, "step": 4580, "task_loss": 1.8129959106445312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9466695374488105, "compression/movement_sparsity/importance_threshold": -0.0003461224284013508, "compression/movement_sparsity/linear_layer_sparsity": 0.9003786948702803, "compression/movement_sparsity/model_sparsity": 0.8694479168126374, "compression_loss": 100.4209976196289, "distillation_loss": 5.92132568359375, "epoch": 3.87, "learning_rate": 4.1214777708202886e-05, "loss": 104.3941, "step": 4581, "task_loss": 3.2964248657226562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.946789112432374, "compression/movement_sparsity/importance_threshold": -0.0003453463694341654, "compression/movement_sparsity/linear_layer_sparsity": 0.9004415113853861, "compression/movement_sparsity/model_sparsity": 0.869508575387203, "compression_loss": 100.43290710449219, "distillation_loss": 4.115144729614258, "epoch": 3.87, "learning_rate": 4.121164683782092e-05, "loss": 104.2581, "step": 4582, "task_loss": 3.125844955444336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9469085085453894, "compression/movement_sparsity/importance_threshold": -0.00034457147136275016, "compression/movement_sparsity/linear_layer_sparsity": 0.9005042563554859, "compression/movement_sparsity/model_sparsity": 0.8695691648745539, "compression_loss": 100.44480895996094, "distillation_loss": 5.145026206970215, "epoch": 3.87, "learning_rate": 4.120851596743895e-05, "loss": 104.6408, "step": 4583, "task_loss": 2.313863515853882 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9470277259217422, "compression/movement_sparsity/importance_threshold": -0.0003437977333181708, "compression/movement_sparsity/linear_layer_sparsity": 0.900423470119753, "compression/movement_sparsity/model_sparsity": 0.8694911538945459, "compression_loss": 100.45662689208984, "distillation_loss": 3.6565868854522705, "epoch": 3.87, "learning_rate": 4.120538509705699e-05, "loss": 104.6371, "step": 4584, "task_loss": 2.1390221118927 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9471467646953173, "compression/movement_sparsity/importance_threshold": -0.0003430251544314939, "compression/movement_sparsity/linear_layer_sparsity": 0.9004642626972352, "compression/movement_sparsity/model_sparsity": 0.8695305451214991, "compression_loss": 100.46849060058594, "distillation_loss": 3.6213202476501465, "epoch": 3.88, "learning_rate": 4.120225422667502e-05, "loss": 104.3434, "step": 4585, "task_loss": 2.5940160751342773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.947265625, "compression/movement_sparsity/importance_threshold": -0.0003422537338337861, "compression/movement_sparsity/linear_layer_sparsity": 0.900489875809317, "compression/movement_sparsity/model_sparsity": 0.8695552783443857, "compression_loss": 100.48020935058594, "distillation_loss": 5.036615371704102, "epoch": 3.88, "learning_rate": 4.119912335629305e-05, "loss": 105.0686, "step": 4586, "task_loss": 2.619068145751953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9473843069696755, "compression/movement_sparsity/importance_threshold": -0.00034148347065611137, "compression/movement_sparsity/linear_layer_sparsity": 0.900506676961516, "compression/movement_sparsity/model_sparsity": 0.8695715023253202, "compression_loss": 100.49200439453125, "distillation_loss": 4.005194664001465, "epoch": 3.88, "learning_rate": 4.1195992485911084e-05, "loss": 104.3189, "step": 4587, "task_loss": 1.719730019569397 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9475028107382286, "compression/movement_sparsity/importance_threshold": -0.0003407143640295389, "compression/movement_sparsity/linear_layer_sparsity": 0.9005644853262147, "compression/movement_sparsity/model_sparsity": 0.8696273247948522, "compression_loss": 100.50372314453125, "distillation_loss": 5.418208122253418, "epoch": 3.88, "learning_rate": 4.119286161552912e-05, "loss": 104.9933, "step": 4588, "task_loss": 2.9065704345703125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9476211364395448, "compression/movement_sparsity/importance_threshold": -0.00033994641308513103, "compression/movement_sparsity/linear_layer_sparsity": 0.9006821768607807, "compression/movement_sparsity/model_sparsity": 0.8697409732631443, "compression_loss": 100.51543426513672, "distillation_loss": 4.689842224121094, "epoch": 3.88, "learning_rate": 4.1189730745147154e-05, "loss": 105.8599, "step": 4589, "task_loss": 3.018312454223633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9477392842075092, "compression/movement_sparsity/importance_threshold": -0.000339179616953956, "compression/movement_sparsity/linear_layer_sparsity": 0.9007372665152583, "compression/movement_sparsity/model_sparsity": 0.869794170418515, "compression_loss": 100.5272445678711, "distillation_loss": 4.772582530975342, "epoch": 3.88, "learning_rate": 4.1186599874765186e-05, "loss": 104.9087, "step": 4590, "task_loss": 3.070620059967041 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9478572541760069, "compression/movement_sparsity/importance_threshold": -0.00033841397476707964, "compression/movement_sparsity/linear_layer_sparsity": 0.9007993079594677, "compression/movement_sparsity/model_sparsity": 0.8698540805482541, "compression_loss": 100.53894805908203, "distillation_loss": 3.4688515663146973, "epoch": 3.88, "learning_rate": 4.118346900438322e-05, "loss": 104.2897, "step": 4591, "task_loss": 2.6497347354888916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9479750464789227, "compression/movement_sparsity/importance_threshold": -0.0003376494856555685, "compression/movement_sparsity/linear_layer_sparsity": 0.9008555423340385, "compression/movement_sparsity/model_sparsity": 0.869908383099061, "compression_loss": 100.55064392089844, "distillation_loss": 7.072486400604248, "epoch": 3.88, "learning_rate": 4.1180338134001256e-05, "loss": 105.4531, "step": 4592, "task_loss": 3.8340654373168945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9480926612501424, "compression/movement_sparsity/importance_threshold": -0.00033688614875048654, "compression/movement_sparsity/linear_layer_sparsity": 0.9009026785687031, "compression/movement_sparsity/model_sparsity": 0.8699539000590569, "compression_loss": 100.56240844726562, "distillation_loss": 5.129284381866455, "epoch": 3.88, "learning_rate": 4.117720726361929e-05, "loss": 105.2016, "step": 4593, "task_loss": 2.416576862335205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9482100986235505, "compression/movement_sparsity/importance_threshold": -0.00033612396318290216, "compression/movement_sparsity/linear_layer_sparsity": 0.9009855992304429, "compression/movement_sparsity/model_sparsity": 0.8700339721409721, "compression_loss": 100.57411193847656, "distillation_loss": 5.969608783721924, "epoch": 3.88, "learning_rate": 4.117407639323732e-05, "loss": 104.6138, "step": 4594, "task_loss": 3.9726576805114746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9483273587330325, "compression/movement_sparsity/importance_threshold": -0.0003353629280838811, "compression/movement_sparsity/linear_layer_sparsity": 0.9010679713804713, "compression/movement_sparsity/model_sparsity": 0.8701135145542408, "compression_loss": 100.58582305908203, "distillation_loss": 3.8323535919189453, "epoch": 3.88, "learning_rate": 4.117094552285535e-05, "loss": 104.6597, "step": 4595, "task_loss": 2.0595638751983643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9484444417124736, "compression/movement_sparsity/importance_threshold": -0.0003346030425844873, "compression/movement_sparsity/linear_layer_sparsity": 0.90110953902885, "compression/movement_sparsity/model_sparsity": 0.8701536542260206, "compression_loss": 100.59749603271484, "distillation_loss": 3.95436429977417, "epoch": 3.88, "learning_rate": 4.116781465247339e-05, "loss": 104.109, "step": 4596, "task_loss": 3.818039894104004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9485613476957586, "compression/movement_sparsity/importance_threshold": -0.00033384430581578917, "compression/movement_sparsity/linear_layer_sparsity": 0.9011157395960206, "compression/movement_sparsity/model_sparsity": 0.8701596417846338, "compression_loss": 100.60913848876953, "distillation_loss": 4.414793014526367, "epoch": 3.89, "learning_rate": 4.116468378209142e-05, "loss": 105.2556, "step": 4597, "task_loss": 2.3718819618225098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.948678076816773, "compression/movement_sparsity/importance_threshold": -0.0003330867169088515, "compression/movement_sparsity/linear_layer_sparsity": 0.901223104801414, "compression/movement_sparsity/model_sparsity": 0.8702633186649278, "compression_loss": 100.62083435058594, "distillation_loss": 4.25513219833374, "epoch": 3.89, "learning_rate": 4.1161552911709453e-05, "loss": 105.0215, "step": 4598, "task_loss": 1.6051677465438843 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9487946292094017, "compression/movement_sparsity/importance_threshold": -0.0003323302749947401, "compression/movement_sparsity/linear_layer_sparsity": 0.9012030721997857, "compression/movement_sparsity/model_sparsity": 0.8702439742447929, "compression_loss": 100.63252258300781, "distillation_loss": 3.5074031352996826, "epoch": 3.89, "learning_rate": 4.115842204132749e-05, "loss": 104.2617, "step": 4599, "task_loss": 1.444356083869934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9489110050075299, "compression/movement_sparsity/importance_threshold": -0.00033157497920452243, "compression/movement_sparsity/linear_layer_sparsity": 0.9012818551753559, "compression/movement_sparsity/model_sparsity": 0.8703200507827875, "compression_loss": 100.64411926269531, "distillation_loss": 5.495142936706543, "epoch": 3.89, "learning_rate": 4.1155291170945524e-05, "loss": 105.5399, "step": 4600, "task_loss": 2.333914041519165 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9490272043450427, "compression/movement_sparsity/importance_threshold": -0.0003308208286692642, "compression/movement_sparsity/linear_layer_sparsity": 0.9014357007861937, "compression/movement_sparsity/model_sparsity": 0.8704686113236086, "compression_loss": 100.65570068359375, "distillation_loss": 3.959157943725586, "epoch": 3.89, "learning_rate": 4.1152160300563556e-05, "loss": 105.127, "step": 4601, "task_loss": 2.703773021697998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9491432273558256, "compression/movement_sparsity/importance_threshold": -0.0003300678225200285, "compression/movement_sparsity/linear_layer_sparsity": 0.9014590721547601, "compression/movement_sparsity/model_sparsity": 0.870491179813766, "compression_loss": 100.66728973388672, "distillation_loss": 2.9800703525543213, "epoch": 3.89, "learning_rate": 4.114902943018159e-05, "loss": 104.2281, "step": 4602, "task_loss": 1.5689337253570557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9492590741737632, "compression/movement_sparsity/importance_threshold": -0.0003293159598878855, "compression/movement_sparsity/linear_layer_sparsity": 0.9014810006990424, "compression/movement_sparsity/model_sparsity": 0.8705123550450922, "compression_loss": 100.67884063720703, "distillation_loss": 3.7003986835479736, "epoch": 3.89, "learning_rate": 4.1145898559799626e-05, "loss": 104.4454, "step": 4603, "task_loss": 1.427895426750183 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9493747449327409, "compression/movement_sparsity/importance_threshold": -0.0003285652399039, "compression/movement_sparsity/linear_layer_sparsity": 0.9015121108524043, "compression/movement_sparsity/model_sparsity": 0.8705423964689801, "compression_loss": 100.69029998779297, "distillation_loss": 3.2990903854370117, "epoch": 3.89, "learning_rate": 4.114276768941766e-05, "loss": 104.1736, "step": 4604, "task_loss": 3.822112798690796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9494902397666438, "compression/movement_sparsity/importance_threshold": -0.00032781566169913685, "compression/movement_sparsity/linear_layer_sparsity": 0.9014936283925687, "compression/movement_sparsity/model_sparsity": 0.8705245489384985, "compression_loss": 100.70176696777344, "distillation_loss": 5.413264751434326, "epoch": 3.89, "learning_rate": 4.113963681903569e-05, "loss": 105.3005, "step": 4605, "task_loss": 2.7830452919006348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9496055588093572, "compression/movement_sparsity/importance_threshold": -0.0003270672244046636, "compression/movement_sparsity/linear_layer_sparsity": 0.9014972652636977, "compression/movement_sparsity/model_sparsity": 0.8705280608719159, "compression_loss": 100.7132339477539, "distillation_loss": 3.6144630908966064, "epoch": 3.89, "learning_rate": 4.113650594865373e-05, "loss": 104.4845, "step": 4606, "task_loss": 3.4208884239196777 }, { "compression/movement_sparsity/importance_regularization_factor": 0.949720702194766, "compression/movement_sparsity/importance_threshold": -0.00032631992715154417, "compression/movement_sparsity/linear_layer_sparsity": 0.9014663339728503, "compression/movement_sparsity/model_sparsity": 0.8704981921660648, "compression_loss": 100.72468566894531, "distillation_loss": 4.369352340698242, "epoch": 3.89, "learning_rate": 4.113337507827176e-05, "loss": 104.9842, "step": 4607, "task_loss": 2.761279582977295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9498356700567555, "compression/movement_sparsity/importance_threshold": -0.00032557376907084695, "compression/movement_sparsity/linear_layer_sparsity": 0.9014761475628146, "compression/movement_sparsity/model_sparsity": 0.8705076686290237, "compression_loss": 100.73605346679688, "distillation_loss": 5.178452014923096, "epoch": 3.89, "learning_rate": 4.11302442078898e-05, "loss": 104.7333, "step": 4608, "task_loss": 2.9123711585998535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9499504625292108, "compression/movement_sparsity/importance_threshold": -0.0003248287492936359, "compression/movement_sparsity/linear_layer_sparsity": 0.9015435071857896, "compression/movement_sparsity/model_sparsity": 0.8705727142417271, "compression_loss": 100.74738311767578, "distillation_loss": 5.050841331481934, "epoch": 3.9, "learning_rate": 4.112711333750783e-05, "loss": 105.0343, "step": 4609, "task_loss": 1.991439938545227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.950065079746017, "compression/movement_sparsity/importance_threshold": -0.0003240848669509794, "compression/movement_sparsity/linear_layer_sparsity": 0.9015571365093974, "compression/movement_sparsity/model_sparsity": 0.8705858753561403, "compression_loss": 100.75880432128906, "distillation_loss": 2.588820219039917, "epoch": 3.9, "learning_rate": 4.112398246712587e-05, "loss": 104.6764, "step": 4610, "task_loss": 1.285973072052002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9501795218410594, "compression/movement_sparsity/importance_threshold": -0.0003233421211739406, "compression/movement_sparsity/linear_layer_sparsity": 0.9016238999239906, "compression/movement_sparsity/model_sparsity": 0.8706503452420539, "compression_loss": 100.77017211914062, "distillation_loss": 4.522249698638916, "epoch": 3.9, "learning_rate": 4.11208515967439e-05, "loss": 104.4881, "step": 4611, "task_loss": 3.599580764770508 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9502937889482228, "compression/movement_sparsity/importance_threshold": -0.0003226005110935886, "compression/movement_sparsity/linear_layer_sparsity": 0.9017028856104106, "compression/movement_sparsity/model_sparsity": 0.870726617527157, "compression_loss": 100.78146362304688, "distillation_loss": 3.5847699642181396, "epoch": 3.9, "learning_rate": 4.111772072636193e-05, "loss": 105.1797, "step": 4612, "task_loss": 2.415595054626465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9504078812013926, "compression/movement_sparsity/importance_threshold": -0.0003218600358409875, "compression/movement_sparsity/linear_layer_sparsity": 0.9018123017726373, "compression/movement_sparsity/model_sparsity": 0.8708322749076076, "compression_loss": 100.79280853271484, "distillation_loss": 5.465292930603027, "epoch": 3.9, "learning_rate": 4.1114589855979964e-05, "loss": 105.5101, "step": 4613, "task_loss": 3.386854410171509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.950521798734454, "compression/movement_sparsity/importance_threshold": -0.00032112069454720386, "compression/movement_sparsity/linear_layer_sparsity": 0.9019231607591478, "compression/movement_sparsity/model_sparsity": 0.8709393255468895, "compression_loss": 100.80410766601562, "distillation_loss": 4.792135238647461, "epoch": 3.9, "learning_rate": 4.1111458985598e-05, "loss": 105.2835, "step": 4614, "task_loss": 2.58837890625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9506355416812919, "compression/movement_sparsity/importance_threshold": -0.00032038248634330344, "compression/movement_sparsity/linear_layer_sparsity": 0.9019595652429401, "compression/movement_sparsity/model_sparsity": 0.8709744794246702, "compression_loss": 100.81539916992188, "distillation_loss": 4.06812047958374, "epoch": 3.9, "learning_rate": 4.1108328115216034e-05, "loss": 105.02, "step": 4615, "task_loss": 2.676288366317749 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9507491101757917, "compression/movement_sparsity/importance_threshold": -0.0003196454103603528, "compression/movement_sparsity/linear_layer_sparsity": 0.9020554474749001, "compression/movement_sparsity/model_sparsity": 0.8710670678069942, "compression_loss": 100.82669067382812, "distillation_loss": 4.225161075592041, "epoch": 3.9, "learning_rate": 4.1105197244834066e-05, "loss": 104.7487, "step": 4616, "task_loss": 2.0559780597686768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9508625043518383, "compression/movement_sparsity/importance_threshold": -0.00031890946572941774, "compression/movement_sparsity/linear_layer_sparsity": 0.9020545889348304, "compression/movement_sparsity/model_sparsity": 0.8710662387604169, "compression_loss": 100.83789825439453, "distillation_loss": 4.894096374511719, "epoch": 3.9, "learning_rate": 4.11020663744521e-05, "loss": 105.7969, "step": 4617, "task_loss": 2.5070176124572754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.950975724343317, "compression/movement_sparsity/importance_threshold": -0.00031817465158156394, "compression/movement_sparsity/linear_layer_sparsity": 0.9020945468205781, "compression/movement_sparsity/model_sparsity": 0.8711048239698644, "compression_loss": 100.8490982055664, "distillation_loss": 5.51568603515625, "epoch": 3.9, "learning_rate": 4.1098935504070137e-05, "loss": 105.8471, "step": 4618, "task_loss": 3.4415700435638428 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9510887702841129, "compression/movement_sparsity/importance_threshold": -0.000317440967047858, "compression/movement_sparsity/linear_layer_sparsity": 0.9021871618306059, "compression/movement_sparsity/model_sparsity": 0.8711942573693806, "compression_loss": 100.86029815673828, "distillation_loss": 6.360787868499756, "epoch": 3.9, "learning_rate": 4.109580463368817e-05, "loss": 105.6298, "step": 4619, "task_loss": 3.17024827003479 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9512016423081112, "compression/movement_sparsity/importance_threshold": -0.0003167084112593657, "compression/movement_sparsity/linear_layer_sparsity": 0.9021614056285124, "compression/movement_sparsity/model_sparsity": 0.8711693859720644, "compression_loss": 100.87141418457031, "distillation_loss": 3.6564440727233887, "epoch": 3.9, "learning_rate": 4.10926737633062e-05, "loss": 105.4128, "step": 4620, "task_loss": 2.247269868850708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.951314340549197, "compression/movement_sparsity/importance_threshold": -0.00031597698334715277, "compression/movement_sparsity/linear_layer_sparsity": 0.9023022419724596, "compression/movement_sparsity/model_sparsity": 0.8713053841543338, "compression_loss": 100.88259887695312, "distillation_loss": 4.738427639007568, "epoch": 3.91, "learning_rate": 4.108954289292424e-05, "loss": 105.2189, "step": 4621, "task_loss": 3.011610746383667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9514268651412552, "compression/movement_sparsity/importance_threshold": -0.00031524668244228664, "compression/movement_sparsity/linear_layer_sparsity": 0.902331014988965, "compression/movement_sparsity/model_sparsity": 0.871333168729206, "compression_loss": 100.89369201660156, "distillation_loss": 4.730954170227051, "epoch": 3.91, "learning_rate": 4.108641202254227e-05, "loss": 105.615, "step": 4622, "task_loss": 2.5221590995788574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9515392162181713, "compression/movement_sparsity/importance_threshold": -0.00031451750767583133, "compression/movement_sparsity/linear_layer_sparsity": 0.902381179962209, "compression/movement_sparsity/model_sparsity": 0.8713816103812936, "compression_loss": 100.90480041503906, "distillation_loss": 3.5265092849731445, "epoch": 3.91, "learning_rate": 4.10832811521603e-05, "loss": 105.1077, "step": 4623, "task_loss": 1.5801843404769897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9516513939138304, "compression/movement_sparsity/importance_threshold": -0.00031378945817885345, "compression/movement_sparsity/linear_layer_sparsity": 0.9024240234965246, "compression/movement_sparsity/model_sparsity": 0.8714229821084034, "compression_loss": 100.91588592529297, "distillation_loss": 4.965545177459717, "epoch": 3.91, "learning_rate": 4.1080150281778334e-05, "loss": 105.0291, "step": 4624, "task_loss": 2.4195353984832764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9517633983621173, "compression/movement_sparsity/importance_threshold": -0.00031306253308242046, "compression/movement_sparsity/linear_layer_sparsity": 0.9025250092722328, "compression/movement_sparsity/model_sparsity": 0.8715204987120474, "compression_loss": 100.92691040039062, "distillation_loss": 4.469172954559326, "epoch": 3.91, "learning_rate": 4.107701941139637e-05, "loss": 104.3236, "step": 4625, "task_loss": 3.2882766723632812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9518752296969175, "compression/movement_sparsity/importance_threshold": -0.00031233673151759723, "compression/movement_sparsity/linear_layer_sparsity": 0.9025964469605392, "compression/movement_sparsity/model_sparsity": 0.8715894822959924, "compression_loss": 100.93798065185547, "distillation_loss": 4.811069965362549, "epoch": 3.91, "learning_rate": 4.1073888541014404e-05, "loss": 105.7166, "step": 4626, "task_loss": 2.2579410076141357 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9519868880521161, "compression/movement_sparsity/importance_threshold": -0.0003116120526154495, "compression/movement_sparsity/linear_layer_sparsity": 0.9026064871096886, "compression/movement_sparsity/model_sparsity": 0.8715991775351315, "compression_loss": 100.94898223876953, "distillation_loss": 5.136645317077637, "epoch": 3.91, "learning_rate": 4.1070757670632436e-05, "loss": 105.6778, "step": 4627, "task_loss": 2.320871114730835 }, { "compression/movement_sparsity/importance_regularization_factor": 0.952098373561598, "compression/movement_sparsity/importance_threshold": -0.00031088849550704473, "compression/movement_sparsity/linear_layer_sparsity": 0.9026689578239329, "compression/movement_sparsity/model_sparsity": 0.871659502188159, "compression_loss": 100.96002197265625, "distillation_loss": 4.077414512634277, "epoch": 3.91, "learning_rate": 4.106762680025047e-05, "loss": 104.5942, "step": 4628, "task_loss": 3.1586973667144775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9522096863592486, "compression/movement_sparsity/importance_threshold": -0.00031016605932344694, "compression/movement_sparsity/linear_layer_sparsity": 0.9026585837980897, "compression/movement_sparsity/model_sparsity": 0.8716494845420178, "compression_loss": 100.9710693359375, "distillation_loss": 5.904484748840332, "epoch": 3.91, "learning_rate": 4.1064495929868506e-05, "loss": 105.2009, "step": 4629, "task_loss": 2.8846917152404785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.952320826578953, "compression/movement_sparsity/importance_threshold": -0.0003094447431957227, "compression/movement_sparsity/linear_layer_sparsity": 0.9026794153189496, "compression/movement_sparsity/model_sparsity": 0.8716696004360509, "compression_loss": 100.98210144042969, "distillation_loss": 4.881065368652344, "epoch": 3.91, "learning_rate": 4.106136505948654e-05, "loss": 105.7797, "step": 4630, "task_loss": 3.0778565406799316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9524317943545961, "compression/movement_sparsity/importance_threshold": -0.0003087245462549404, "compression/movement_sparsity/linear_layer_sparsity": 0.9026763985045376, "compression/movement_sparsity/model_sparsity": 0.8716666872584948, "compression_loss": 100.9931411743164, "distillation_loss": 3.926767349243164, "epoch": 3.91, "learning_rate": 4.105823418910457e-05, "loss": 104.7998, "step": 4631, "task_loss": 1.8330944776535034 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9525425898200633, "compression/movement_sparsity/importance_threshold": -0.0003080054676321631, "compression/movement_sparsity/linear_layer_sparsity": 0.9027012842423937, "compression/movement_sparsity/model_sparsity": 0.871690718094698, "compression_loss": 101.0041275024414, "distillation_loss": 4.834381103515625, "epoch": 3.91, "learning_rate": 4.10551033187226e-05, "loss": 104.9776, "step": 4632, "task_loss": 3.03425931930542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9526532131092398, "compression/movement_sparsity/importance_threshold": -0.0003072875064584575, "compression/movement_sparsity/linear_layer_sparsity": 0.902796403327625, "compression/movement_sparsity/model_sparsity": 0.8717825695467311, "compression_loss": 101.01509094238281, "distillation_loss": 6.7089738845825195, "epoch": 3.92, "learning_rate": 4.105197244834064e-05, "loss": 106.0798, "step": 4633, "task_loss": 3.4084396362304688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9527636643560105, "compression/movement_sparsity/importance_threshold": -0.00030657066186489095, "compression/movement_sparsity/linear_layer_sparsity": 0.9029178986716668, "compression/movement_sparsity/model_sparsity": 0.8718998911519417, "compression_loss": 101.02601623535156, "distillation_loss": 5.702532768249512, "epoch": 3.92, "learning_rate": 4.104884157795867e-05, "loss": 105.5913, "step": 4634, "task_loss": 2.5024216175079346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9528739436942606, "compression/movement_sparsity/importance_threshold": -0.0003058549329825293, "compression/movement_sparsity/linear_layer_sparsity": 0.9028976752833564, "compression/movement_sparsity/model_sparsity": 0.8718803624992342, "compression_loss": 101.03695678710938, "distillation_loss": 3.1224308013916016, "epoch": 3.92, "learning_rate": 4.1045710707576704e-05, "loss": 105.4553, "step": 4635, "task_loss": 1.9786745309829712 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9529840512578753, "compression/movement_sparsity/importance_threshold": -0.00030514031894243645, "compression/movement_sparsity/linear_layer_sparsity": 0.9029027549787693, "compression/movement_sparsity/model_sparsity": 0.8718852676914827, "compression_loss": 101.04791259765625, "distillation_loss": 3.4683737754821777, "epoch": 3.92, "learning_rate": 4.104257983719474e-05, "loss": 105.2427, "step": 4636, "task_loss": 1.6941593885421753 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9530939871807399, "compression/movement_sparsity/importance_threshold": -0.00030442681887567994, "compression/movement_sparsity/linear_layer_sparsity": 0.9029685763841192, "compression/movement_sparsity/model_sparsity": 0.8719488279290685, "compression_loss": 101.05874633789062, "distillation_loss": 3.1010584831237793, "epoch": 3.92, "learning_rate": 4.1039448966812774e-05, "loss": 104.8862, "step": 4637, "task_loss": 1.8707209825515747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9532037515967392, "compression/movement_sparsity/importance_threshold": -0.0003037144319133272, "compression/movement_sparsity/linear_layer_sparsity": 0.9030972023804072, "compression/movement_sparsity/model_sparsity": 0.8720730352266842, "compression_loss": 101.0696029663086, "distillation_loss": 3.9027957916259766, "epoch": 3.92, "learning_rate": 4.1036318096430806e-05, "loss": 105.6905, "step": 4638, "task_loss": 2.948152780532837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9533133446397585, "compression/movement_sparsity/importance_threshold": -0.0003030031571864423, "compression/movement_sparsity/linear_layer_sparsity": 0.9031494660071553, "compression/movement_sparsity/model_sparsity": 0.8721235034370717, "compression_loss": 101.08052825927734, "distillation_loss": 4.928401947021484, "epoch": 3.92, "learning_rate": 4.1033187226048845e-05, "loss": 105.0611, "step": 4639, "task_loss": 1.8813085556030273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.953422766443683, "compression/movement_sparsity/importance_threshold": -0.00030229299382609175, "compression/movement_sparsity/linear_layer_sparsity": 0.9032642599689857, "compression/movement_sparsity/model_sparsity": 0.8722343538731657, "compression_loss": 101.0912857055664, "distillation_loss": 4.713939666748047, "epoch": 3.92, "learning_rate": 4.1030056355666876e-05, "loss": 105.1799, "step": 4640, "task_loss": 3.034125566482544 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9535320171423978, "compression/movement_sparsity/importance_threshold": -0.0003015839409633422, "compression/movement_sparsity/linear_layer_sparsity": 0.9033682864074409, "compression/movement_sparsity/model_sparsity": 0.8723348066834373, "compression_loss": 101.1020736694336, "distillation_loss": 3.92116117477417, "epoch": 3.92, "learning_rate": 4.1026925485284915e-05, "loss": 104.9877, "step": 4641, "task_loss": 1.6019446849822998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.953641096869788, "compression/movement_sparsity/importance_threshold": -0.00030087599772925765, "compression/movement_sparsity/linear_layer_sparsity": 0.9034129424152373, "compression/movement_sparsity/model_sparsity": 0.8723779286199879, "compression_loss": 101.11283874511719, "distillation_loss": 5.278929233551025, "epoch": 3.92, "learning_rate": 4.102379461490295e-05, "loss": 105.6137, "step": 4642, "task_loss": 1.9051085710525513 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9537500057597388, "compression/movement_sparsity/importance_threshold": -0.0003001691632549073, "compression/movement_sparsity/linear_layer_sparsity": 0.9034070161039223, "compression/movement_sparsity/model_sparsity": 0.872372205895698, "compression_loss": 101.12357330322266, "distillation_loss": 4.100269317626953, "epoch": 3.92, "learning_rate": 4.102066374452098e-05, "loss": 105.4378, "step": 4643, "task_loss": 2.258197546005249 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9538587439461352, "compression/movement_sparsity/importance_threshold": -0.00029946343667135514, "compression/movement_sparsity/linear_layer_sparsity": 0.9034668992737895, "compression/movement_sparsity/model_sparsity": 0.8724300318944581, "compression_loss": 101.13428497314453, "distillation_loss": 4.243312358856201, "epoch": 3.93, "learning_rate": 4.101753287413902e-05, "loss": 105.2812, "step": 4644, "task_loss": 2.2319164276123047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9539673115628625, "compression/movement_sparsity/importance_threshold": -0.0002987588171096678, "compression/movement_sparsity/linear_layer_sparsity": 0.9034837719709943, "compression/movement_sparsity/model_sparsity": 0.8724463249626074, "compression_loss": 101.14497375488281, "distillation_loss": 4.16384744644165, "epoch": 3.93, "learning_rate": 4.101440200375705e-05, "loss": 105.1358, "step": 4645, "task_loss": 2.694849967956543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9540757087438059, "compression/movement_sparsity/importance_threshold": -0.00029805530370091097, "compression/movement_sparsity/linear_layer_sparsity": 0.903461712260868, "compression/movement_sparsity/model_sparsity": 0.8724250230713875, "compression_loss": 101.15569305419922, "distillation_loss": 3.699084520339966, "epoch": 3.93, "learning_rate": 4.101127113337508e-05, "loss": 105.917, "step": 4646, "task_loss": 2.155829429626465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9541839356228503, "compression/movement_sparsity/importance_threshold": -0.0002973528955761513, "compression/movement_sparsity/linear_layer_sparsity": 0.9035150609868708, "compression/movement_sparsity/model_sparsity": 0.8724765391045323, "compression_loss": 101.1663818359375, "distillation_loss": 5.5700154304504395, "epoch": 3.93, "learning_rate": 4.100814026299312e-05, "loss": 105.2181, "step": 4647, "task_loss": 2.4601385593414307 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9542919923338811, "compression/movement_sparsity/importance_threshold": -0.0002966515918664536, "compression/movement_sparsity/linear_layer_sparsity": 0.9035075129887573, "compression/movement_sparsity/model_sparsity": 0.8724692504033743, "compression_loss": 101.17709350585938, "distillation_loss": 4.9870147705078125, "epoch": 3.93, "learning_rate": 4.100500939261115e-05, "loss": 105.9235, "step": 4648, "task_loss": 3.279409885406494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9543998790107833, "compression/movement_sparsity/importance_threshold": -0.00029595139170288456, "compression/movement_sparsity/linear_layer_sparsity": 0.9035552096593008, "compression/movement_sparsity/model_sparsity": 0.8725153085465525, "compression_loss": 101.1877670288086, "distillation_loss": 4.386382102966309, "epoch": 3.93, "learning_rate": 4.100187852222918e-05, "loss": 105.4513, "step": 4649, "task_loss": 1.932860255241394 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9545075957874419, "compression/movement_sparsity/importance_threshold": -0.0002952522942165116, "compression/movement_sparsity/linear_layer_sparsity": 0.9035431781741562, "compression/movement_sparsity/model_sparsity": 0.8725036903799358, "compression_loss": 101.19841766357422, "distillation_loss": 6.38057804107666, "epoch": 3.93, "learning_rate": 4.0998747651847215e-05, "loss": 105.8229, "step": 4650, "task_loss": 3.206296920776367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9546151427977424, "compression/movement_sparsity/importance_threshold": -0.0002945542985383987, "compression/movement_sparsity/linear_layer_sparsity": 0.9036144846966186, "compression/movement_sparsity/model_sparsity": 0.8725725473039871, "compression_loss": 101.20905303955078, "distillation_loss": 3.6160573959350586, "epoch": 3.93, "learning_rate": 4.099561678146525e-05, "loss": 105.3239, "step": 4651, "task_loss": 2.227640151977539 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9547225201755696, "compression/movement_sparsity/importance_threshold": -0.00029385740379961334, "compression/movement_sparsity/linear_layer_sparsity": 0.9036634095564285, "compression/movement_sparsity/model_sparsity": 0.8726197914443522, "compression_loss": 101.21965789794922, "distillation_loss": 4.053164005279541, "epoch": 3.93, "learning_rate": 4.0992485911083285e-05, "loss": 105.5276, "step": 4652, "task_loss": 1.522519826889038 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9548297280548088, "compression/movement_sparsity/importance_threshold": -0.0002931616091312204, "compression/movement_sparsity/linear_layer_sparsity": 0.9036579244393161, "compression/movement_sparsity/model_sparsity": 0.8726144947578867, "compression_loss": 101.23029327392578, "distillation_loss": 4.704656600952148, "epoch": 3.93, "learning_rate": 4.098935504070132e-05, "loss": 105.3343, "step": 4653, "task_loss": 2.4082140922546387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9549367665693451, "compression/movement_sparsity/importance_threshold": -0.0002924669136642865, "compression/movement_sparsity/linear_layer_sparsity": 0.9037693915583761, "compression/movement_sparsity/model_sparsity": 0.8727221326384941, "compression_loss": 101.2408447265625, "distillation_loss": 4.057040214538574, "epoch": 3.93, "learning_rate": 4.098622417031935e-05, "loss": 105.7469, "step": 4654, "task_loss": 2.716005325317383 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9550436358530636, "compression/movement_sparsity/importance_threshold": -0.0002917733165298791, "compression/movement_sparsity/linear_layer_sparsity": 0.9039174897204135, "compression/movement_sparsity/model_sparsity": 0.8728651431730622, "compression_loss": 101.25138854980469, "distillation_loss": 5.200481414794922, "epoch": 3.93, "learning_rate": 4.098309329993739e-05, "loss": 105.0197, "step": 4655, "task_loss": 2.737880229949951 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9551503360398497, "compression/movement_sparsity/importance_threshold": -0.0002910808168590613, "compression/movement_sparsity/linear_layer_sparsity": 0.9040610209262464, "compression/movement_sparsity/model_sparsity": 0.8730037436404211, "compression_loss": 101.26190185546875, "distillation_loss": 5.224555969238281, "epoch": 3.94, "learning_rate": 4.097996242955542e-05, "loss": 105.9385, "step": 4656, "task_loss": 2.7645864486694336 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9552568672635882, "compression/movement_sparsity/importance_threshold": -0.0002903894137829006, "compression/movement_sparsity/linear_layer_sparsity": 0.904050134161195, "compression/movement_sparsity/model_sparsity": 0.8729932308692407, "compression_loss": 101.27246856689453, "distillation_loss": 3.113921642303467, "epoch": 3.94, "learning_rate": 4.097683155917345e-05, "loss": 105.0287, "step": 4657, "task_loss": 2.2281739711761475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9553632296581643, "compression/movement_sparsity/importance_threshold": -0.00028969910643246357, "compression/movement_sparsity/linear_layer_sparsity": 0.9040714545729278, "compression/movement_sparsity/model_sparsity": 0.8730138188592413, "compression_loss": 101.28302001953125, "distillation_loss": 6.1024065017700195, "epoch": 3.94, "learning_rate": 4.097370068879149e-05, "loss": 105.4812, "step": 4658, "task_loss": 4.122501850128174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9554694233574633, "compression/movement_sparsity/importance_threshold": -0.000289009893938816, "compression/movement_sparsity/linear_layer_sparsity": 0.9041143219555787, "compression/movement_sparsity/model_sparsity": 0.8730552136154227, "compression_loss": 101.2935791015625, "distillation_loss": 3.9201693534851074, "epoch": 3.94, "learning_rate": 4.097056981840952e-05, "loss": 105.8023, "step": 4659, "task_loss": 2.0746264457702637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9555754484953703, "compression/movement_sparsity/importance_threshold": -0.0002883217754330236, "compression/movement_sparsity/linear_layer_sparsity": 0.9041998301616956, "compression/movement_sparsity/model_sparsity": 0.8731377843516054, "compression_loss": 101.30413818359375, "distillation_loss": 3.8624472618103027, "epoch": 3.94, "learning_rate": 4.096743894802755e-05, "loss": 104.924, "step": 4660, "task_loss": 1.7499442100524902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9556813052057704, "compression/movement_sparsity/importance_threshold": -0.0002876347500461521, "compression/movement_sparsity/linear_layer_sparsity": 0.9043138609767973, "compression/movement_sparsity/model_sparsity": 0.8732478978574085, "compression_loss": 101.31464385986328, "distillation_loss": 5.104468822479248, "epoch": 3.94, "learning_rate": 4.0964308077645584e-05, "loss": 105.175, "step": 4661, "task_loss": 2.553333044052124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9557869936225487, "compression/movement_sparsity/importance_threshold": -0.0002869488169092681, "compression/movement_sparsity/linear_layer_sparsity": 0.9043723132465483, "compression/movement_sparsity/model_sparsity": 0.8733043421118734, "compression_loss": 101.32512664794922, "distillation_loss": 1.6891396045684814, "epoch": 3.94, "learning_rate": 4.096117720726362e-05, "loss": 105.3076, "step": 4662, "task_loss": 0.9750037789344788 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9558925138795905, "compression/movement_sparsity/importance_threshold": -0.0002862639751534373, "compression/movement_sparsity/linear_layer_sparsity": 0.9044154906575578, "compression/movement_sparsity/model_sparsity": 0.8733460362459854, "compression_loss": 101.33556365966797, "distillation_loss": 2.956501007080078, "epoch": 3.94, "learning_rate": 4.0958046336881655e-05, "loss": 105.6095, "step": 4663, "task_loss": 2.1116254329681396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9559978661107807, "compression/movement_sparsity/importance_threshold": -0.0002855802239097264, "compression/movement_sparsity/linear_layer_sparsity": 0.9044363221784176, "compression/movement_sparsity/model_sparsity": 0.8733661521400186, "compression_loss": 101.34598541259766, "distillation_loss": 4.750007152557373, "epoch": 3.94, "learning_rate": 4.0954915466499687e-05, "loss": 105.2613, "step": 4664, "task_loss": 2.1608355045318604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9561030504500045, "compression/movement_sparsity/importance_threshold": -0.00028489756230920104, "compression/movement_sparsity/linear_layer_sparsity": 0.9044306343504553, "compression/movement_sparsity/model_sparsity": 0.8733606597064445, "compression_loss": 101.35643005371094, "distillation_loss": 4.676169395446777, "epoch": 3.94, "learning_rate": 4.095178459611772e-05, "loss": 105.7568, "step": 4665, "task_loss": 3.349846363067627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9562080670311472, "compression/movement_sparsity/importance_threshold": -0.000284215989482927, "compression/movement_sparsity/linear_layer_sparsity": 0.9045288894917748, "compression/movement_sparsity/model_sparsity": 0.8734555394813915, "compression_loss": 101.36677551269531, "distillation_loss": 3.909311532974243, "epoch": 3.94, "learning_rate": 4.094865372573576e-05, "loss": 105.4214, "step": 4666, "task_loss": 1.5802407264709473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.956312915988094, "compression/movement_sparsity/importance_threshold": -0.00028353550456197, "compression/movement_sparsity/linear_layer_sparsity": 0.9045355670256509, "compression/movement_sparsity/model_sparsity": 0.8734619876214365, "compression_loss": 101.3771743774414, "distillation_loss": 4.0445146560668945, "epoch": 3.94, "learning_rate": 4.094552285535379e-05, "loss": 105.4979, "step": 4667, "task_loss": 2.406919240951538 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9564175974547297, "compression/movement_sparsity/importance_threshold": -0.00028285610667739755, "compression/movement_sparsity/linear_layer_sparsity": 0.9046111185517918, "compression/movement_sparsity/model_sparsity": 0.8735349437202307, "compression_loss": 101.38754272460938, "distillation_loss": 4.612880706787109, "epoch": 3.95, "learning_rate": 4.094239198497182e-05, "loss": 106.1755, "step": 4668, "task_loss": 3.124878168106079 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9565221115649396, "compression/movement_sparsity/importance_threshold": -0.00028217779496027447, "compression/movement_sparsity/linear_layer_sparsity": 0.9045926837886267, "compression/movement_sparsity/model_sparsity": 0.8735171422478923, "compression_loss": 101.39789581298828, "distillation_loss": 3.3281517028808594, "epoch": 3.95, "learning_rate": 4.093926111458985e-05, "loss": 105.6907, "step": 4669, "task_loss": 2.3057453632354736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.956626458452609, "compression/movement_sparsity/importance_threshold": -0.0002815005685416665, "compression/movement_sparsity/linear_layer_sparsity": 0.9046014361276714, "compression/movement_sparsity/model_sparsity": 0.8735255939171656, "compression_loss": 101.4082260131836, "distillation_loss": 5.149050235748291, "epoch": 3.95, "learning_rate": 4.093613024420789e-05, "loss": 105.8883, "step": 4670, "task_loss": 1.7692673206329346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9567306382516227, "compression/movement_sparsity/importance_threshold": -0.00028082442655264113, "compression/movement_sparsity/linear_layer_sparsity": 0.9046018177010358, "compression/movement_sparsity/model_sparsity": 0.873525962382311, "compression_loss": 101.41852569580078, "distillation_loss": 4.295919418334961, "epoch": 3.95, "learning_rate": 4.093299937382592e-05, "loss": 105.7549, "step": 4671, "task_loss": 2.827812910079956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9568346510958662, "compression/movement_sparsity/importance_threshold": -0.0002801493681242632, "compression/movement_sparsity/linear_layer_sparsity": 0.9046683068597733, "compression/movement_sparsity/model_sparsity": 0.8735901674339013, "compression_loss": 101.42880249023438, "distillation_loss": 4.039035320281982, "epoch": 3.95, "learning_rate": 4.0929868503443954e-05, "loss": 105.3921, "step": 4672, "task_loss": 2.123638153076172 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9569384971192245, "compression/movement_sparsity/importance_threshold": -0.0002794753923875985, "compression/movement_sparsity/linear_layer_sparsity": 0.9046413224684133, "compression/movement_sparsity/model_sparsity": 0.8735641100393983, "compression_loss": 101.43904876708984, "distillation_loss": 6.634730339050293, "epoch": 3.95, "learning_rate": 4.092673763306199e-05, "loss": 105.8612, "step": 4673, "task_loss": 3.991177558898926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9570421764555826, "compression/movement_sparsity/importance_threshold": -0.00027880249847371443, "compression/movement_sparsity/linear_layer_sparsity": 0.9048130185582022, "compression/movement_sparsity/model_sparsity": 0.8737299078403038, "compression_loss": 101.44927215576172, "distillation_loss": 4.888792037963867, "epoch": 3.95, "learning_rate": 4.0923606762680025e-05, "loss": 105.974, "step": 4674, "task_loss": 2.743758201599121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9571456892388259, "compression/movement_sparsity/importance_threshold": -0.00027813068551367504, "compression/movement_sparsity/linear_layer_sparsity": 0.904840956882973, "compression/movement_sparsity/model_sparsity": 0.8737568863976705, "compression_loss": 101.45947265625, "distillation_loss": 3.7873692512512207, "epoch": 3.95, "learning_rate": 4.092047589229806e-05, "loss": 105.4576, "step": 4675, "task_loss": 2.4788427352905273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9572490356028394, "compression/movement_sparsity/importance_threshold": -0.00027745995263854864, "compression/movement_sparsity/linear_layer_sparsity": 0.9048799846836452, "compression/movement_sparsity/model_sparsity": 0.873794573473326, "compression_loss": 101.4697036743164, "distillation_loss": 6.498286247253418, "epoch": 3.95, "learning_rate": 4.0917345021916095e-05, "loss": 106.2659, "step": 4676, "task_loss": 3.209911346435547 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9573522156815082, "compression/movement_sparsity/importance_threshold": -0.0002767902989794001, "compression/movement_sparsity/linear_layer_sparsity": 0.9047927713215564, "compression/movement_sparsity/model_sparsity": 0.8737103561585247, "compression_loss": 101.47989654541016, "distillation_loss": 2.8484814167022705, "epoch": 3.95, "learning_rate": 4.0914214151534134e-05, "loss": 104.8972, "step": 4677, "task_loss": 2.139747381210327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9574552296087175, "compression/movement_sparsity/importance_threshold": -0.0002761217236672952, "compression/movement_sparsity/linear_layer_sparsity": 0.9048352809791783, "compression/movement_sparsity/model_sparsity": 0.8737514054786323, "compression_loss": 101.49007415771484, "distillation_loss": 2.802572727203369, "epoch": 3.95, "learning_rate": 4.0911083281152165e-05, "loss": 104.9176, "step": 4678, "task_loss": 2.5844101905822754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9575580775183525, "compression/movement_sparsity/importance_threshold": -0.0002754542258333005, "compression/movement_sparsity/linear_layer_sparsity": 0.9048105741038368, "compression/movement_sparsity/model_sparsity": 0.873727547360466, "compression_loss": 101.50020599365234, "distillation_loss": 4.564849853515625, "epoch": 3.95, "learning_rate": 4.09079524107702e-05, "loss": 105.4872, "step": 4679, "task_loss": 2.487870454788208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9576607595442982, "compression/movement_sparsity/importance_threshold": -0.00027478780460848257, "compression/movement_sparsity/linear_layer_sparsity": 0.9048629450480935, "compression/movement_sparsity/model_sparsity": 0.8737781192016756, "compression_loss": 101.51033020019531, "distillation_loss": 5.171796798706055, "epoch": 3.96, "learning_rate": 4.090482154038823e-05, "loss": 106.2071, "step": 4680, "task_loss": 2.852546453475952 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9577632758204399, "compression/movement_sparsity/importance_threshold": -0.00027412245912390547, "compression/movement_sparsity/linear_layer_sparsity": 0.9048723770646935, "compression/movement_sparsity/model_sparsity": 0.8737872271994891, "compression_loss": 101.52049255371094, "distillation_loss": 4.86936092376709, "epoch": 3.96, "learning_rate": 4.090169067000627e-05, "loss": 106.0816, "step": 4681, "task_loss": 2.1388707160949707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9578656264806625, "compression/movement_sparsity/importance_threshold": -0.0002734581885106375, "compression/movement_sparsity/linear_layer_sparsity": 0.9048879858001289, "compression/movement_sparsity/model_sparsity": 0.8738022997268441, "compression_loss": 101.5306396484375, "distillation_loss": 3.4570508003234863, "epoch": 3.96, "learning_rate": 4.08985597996243e-05, "loss": 105.8264, "step": 4682, "task_loss": 1.7082393169403076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9579678116588514, "compression/movement_sparsity/importance_threshold": -0.0002727949918997436, "compression/movement_sparsity/linear_layer_sparsity": 0.9049754376455702, "compression/movement_sparsity/model_sparsity": 0.8738867473323613, "compression_loss": 101.54072570800781, "distillation_loss": 4.082529067993164, "epoch": 3.96, "learning_rate": 4.089542892924233e-05, "loss": 105.8747, "step": 4683, "task_loss": 1.6747971773147583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9580698314888917, "compression/movement_sparsity/importance_threshold": -0.0002721328684222894, "compression/movement_sparsity/linear_layer_sparsity": 0.9050229077569286, "compression/movement_sparsity/model_sparsity": 0.8739325866993594, "compression_loss": 101.55084991455078, "distillation_loss": 4.240242958068848, "epoch": 3.96, "learning_rate": 4.089229805886037e-05, "loss": 105.8156, "step": 4684, "task_loss": 3.3034210205078125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9581716861046683, "compression/movement_sparsity/importance_threshold": -0.0002714718172093424, "compression/movement_sparsity/linear_layer_sparsity": 0.905064189225284, "compression/movement_sparsity/model_sparsity": 0.8739724500222801, "compression_loss": 101.5609130859375, "distillation_loss": 3.0009407997131348, "epoch": 3.96, "learning_rate": 4.08891671884784e-05, "loss": 105.216, "step": 4685, "task_loss": 1.8811537027359009 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9582733756400666, "compression/movement_sparsity/importance_threshold": -0.0002708118373919667, "compression/movement_sparsity/linear_layer_sparsity": 0.9051581635904222, "compression/movement_sparsity/model_sparsity": 0.8740631960788768, "compression_loss": 101.5709457397461, "distillation_loss": 4.949099063873291, "epoch": 3.96, "learning_rate": 4.088603631809643e-05, "loss": 106.603, "step": 4686, "task_loss": 3.711230754852295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9583749002289718, "compression/movement_sparsity/importance_threshold": -0.0002701529281012296, "compression/movement_sparsity/linear_layer_sparsity": 0.905089337294828, "compression/movement_sparsity/model_sparsity": 0.8739967341782707, "compression_loss": 101.5809555053711, "distillation_loss": 4.084665775299072, "epoch": 3.96, "learning_rate": 4.0882905447714465e-05, "loss": 105.8096, "step": 4687, "task_loss": 2.823697328567505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9584762600052686, "compression/movement_sparsity/importance_threshold": -0.00026949508846819785, "compression/movement_sparsity/linear_layer_sparsity": 0.9051699088955435, "compression/movement_sparsity/model_sparsity": 0.8740745378966345, "compression_loss": 101.59095764160156, "distillation_loss": 3.305032253265381, "epoch": 3.96, "learning_rate": 4.0879774577332504e-05, "loss": 106.1126, "step": 4688, "task_loss": 1.9472458362579346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9585774551028428, "compression/movement_sparsity/importance_threshold": -0.0002688383176239354, "compression/movement_sparsity/linear_layer_sparsity": 0.9052716458938127, "compression/movement_sparsity/model_sparsity": 0.8741727799160335, "compression_loss": 101.60093688964844, "distillation_loss": 4.458347320556641, "epoch": 3.96, "learning_rate": 4.0876643706950535e-05, "loss": 105.8699, "step": 4689, "task_loss": 1.9441804885864258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9586784856555789, "compression/movement_sparsity/importance_threshold": -0.00026818261469950967, "compression/movement_sparsity/linear_layer_sparsity": 0.9053010985878733, "compression/movement_sparsity/model_sparsity": 0.874201220819446, "compression_loss": 101.61085510253906, "distillation_loss": 4.028903007507324, "epoch": 3.96, "learning_rate": 4.087351283656857e-05, "loss": 105.1906, "step": 4690, "task_loss": 2.892592430114746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9587793517973624, "compression/movement_sparsity/importance_threshold": -0.0002675279788259873, "compression/movement_sparsity/linear_layer_sparsity": 0.9054183847007397, "compression/movement_sparsity/model_sparsity": 0.8743144777935211, "compression_loss": 101.62077331542969, "distillation_loss": 3.4157872200012207, "epoch": 3.96, "learning_rate": 4.08703819661866e-05, "loss": 105.996, "step": 4691, "task_loss": 2.4577319622039795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9588800536620783, "compression/movement_sparsity/importance_threshold": -0.0002668744091344332, "compression/movement_sparsity/linear_layer_sparsity": 0.9054883318680916, "compression/movement_sparsity/model_sparsity": 0.8743820220604919, "compression_loss": 101.6307144165039, "distillation_loss": 5.2552008628845215, "epoch": 3.97, "learning_rate": 4.086725109580464e-05, "loss": 105.8727, "step": 4692, "task_loss": 3.7607932090759277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9589805913836119, "compression/movement_sparsity/importance_threshold": -0.00026622190475591306, "compression/movement_sparsity/linear_layer_sparsity": 0.9055631321716714, "compression/movement_sparsity/model_sparsity": 0.8744542527435311, "compression_loss": 101.64057922363281, "distillation_loss": 4.123255252838135, "epoch": 3.97, "learning_rate": 4.086412022542267e-05, "loss": 105.8067, "step": 4693, "task_loss": 2.3276443481445312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9590809650958482, "compression/movement_sparsity/importance_threshold": -0.00026557046482149436, "compression/movement_sparsity/linear_layer_sparsity": 0.9055388545663647, "compression/movement_sparsity/model_sparsity": 0.8744308091486533, "compression_loss": 101.65045166015625, "distillation_loss": 5.366290092468262, "epoch": 3.97, "learning_rate": 4.08609893550407e-05, "loss": 106.294, "step": 4694, "task_loss": 2.2536306381225586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9591811749326724, "compression/movement_sparsity/importance_threshold": -0.0002649200884622411, "compression/movement_sparsity/linear_layer_sparsity": 0.9056016472331352, "compression/movement_sparsity/model_sparsity": 0.8744914446941474, "compression_loss": 101.66034698486328, "distillation_loss": 3.157841682434082, "epoch": 3.97, "learning_rate": 4.085785848465874e-05, "loss": 105.1714, "step": 4695, "task_loss": 2.8524930477142334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9592812210279696, "compression/movement_sparsity/importance_threshold": -0.0002642707748092216, "compression/movement_sparsity/linear_layer_sparsity": 0.9056109003872207, "compression/movement_sparsity/model_sparsity": 0.8745003799739239, "compression_loss": 101.6701889038086, "distillation_loss": 3.8445475101470947, "epoch": 3.97, "learning_rate": 4.085472761427677e-05, "loss": 105.8849, "step": 4696, "task_loss": 1.5538086891174316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.959381103515625, "compression/movement_sparsity/importance_threshold": -0.00026362252299350075, "compression/movement_sparsity/linear_layer_sparsity": 0.9056142868508292, "compression/movement_sparsity/model_sparsity": 0.8745036501020896, "compression_loss": 101.6800308227539, "distillation_loss": 4.1866302490234375, "epoch": 3.97, "learning_rate": 4.08515967438948e-05, "loss": 105.6728, "step": 4697, "task_loss": 2.758531332015991 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9594808225295237, "compression/movement_sparsity/importance_threshold": -0.0002629753321461443, "compression/movement_sparsity/linear_layer_sparsity": 0.9056509536663095, "compression/movement_sparsity/model_sparsity": 0.8745390572996579, "compression_loss": 101.68984985351562, "distillation_loss": 4.511089324951172, "epoch": 3.97, "learning_rate": 4.0848465873512835e-05, "loss": 105.7044, "step": 4698, "task_loss": 2.151195764541626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9595803782035508, "compression/movement_sparsity/importance_threshold": -0.00026232920139821966, "compression/movement_sparsity/linear_layer_sparsity": 0.9057519513661852, "compression/movement_sparsity/model_sparsity": 0.8746365854178376, "compression_loss": 101.69966125488281, "distillation_loss": 5.833030700683594, "epoch": 3.97, "learning_rate": 4.0845335003130873e-05, "loss": 105.9798, "step": 4699, "task_loss": 2.9345176219940186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9596797706715915, "compression/movement_sparsity/importance_threshold": -0.0002616841298807909, "compression/movement_sparsity/linear_layer_sparsity": 0.9058785502539752, "compression/movement_sparsity/model_sparsity": 0.8747588352443683, "compression_loss": 101.70946502685547, "distillation_loss": 4.042781829833984, "epoch": 3.97, "learning_rate": 4.0842204132748905e-05, "loss": 106.3268, "step": 4700, "task_loss": 1.5597920417785645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9597790000675309, "compression/movement_sparsity/importance_threshold": -0.00026104011672492544, "compression/movement_sparsity/linear_layer_sparsity": 0.9058789795240101, "compression/movement_sparsity/model_sparsity": 0.8747592497676568, "compression_loss": 101.71929168701172, "distillation_loss": 4.6326751708984375, "epoch": 3.97, "learning_rate": 4.083907326236694e-05, "loss": 105.9519, "step": 4701, "task_loss": 3.3509528636932373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9598780665252543, "compression/movement_sparsity/importance_threshold": -0.0002603971610616882, "compression/movement_sparsity/linear_layer_sparsity": 0.9058841784610994, "compression/movement_sparsity/model_sparsity": 0.8747642701052633, "compression_loss": 101.72908782958984, "distillation_loss": 4.735093116760254, "epoch": 3.97, "learning_rate": 4.083594239198497e-05, "loss": 106.3131, "step": 4702, "task_loss": 3.3809025287628174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9599769701786466, "compression/movement_sparsity/importance_threshold": -0.00025975526202214744, "compression/movement_sparsity/linear_layer_sparsity": 0.9059279401563229, "compression/movement_sparsity/model_sparsity": 0.8748065284516292, "compression_loss": 101.73886108398438, "distillation_loss": 3.5122628211975098, "epoch": 3.97, "learning_rate": 4.083281152160301e-05, "loss": 106.3776, "step": 4703, "task_loss": 2.2108540534973145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9600757111615931, "compression/movement_sparsity/importance_threshold": -0.00025911441873736635, "compression/movement_sparsity/linear_layer_sparsity": 0.9060294982920776, "compression/movement_sparsity/model_sparsity": 0.8749045977529913, "compression_loss": 101.7486343383789, "distillation_loss": 3.41514253616333, "epoch": 3.98, "learning_rate": 4.082968065122104e-05, "loss": 106.411, "step": 4704, "task_loss": 1.7220221757888794 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9601742896079788, "compression/movement_sparsity/importance_threshold": -0.00025847463033841327, "compression/movement_sparsity/linear_layer_sparsity": 0.9060256706342665, "compression/movement_sparsity/model_sparsity": 0.8749009015870013, "compression_loss": 101.75834655761719, "distillation_loss": 3.015875816345215, "epoch": 3.98, "learning_rate": 4.082654978083907e-05, "loss": 105.6279, "step": 4705, "task_loss": 1.9595389366149902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.960272705651689, "compression/movement_sparsity/importance_threshold": -0.00025783589595635304, "compression/movement_sparsity/linear_layer_sparsity": 0.9061010790703957, "compression/movement_sparsity/model_sparsity": 0.8749737195113659, "compression_loss": 101.76811981201172, "distillation_loss": 5.043632507324219, "epoch": 3.98, "learning_rate": 4.082341891045711e-05, "loss": 106.8606, "step": 4706, "task_loss": 2.4567673206329346 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9603709594266087, "compression/movement_sparsity/importance_threshold": -0.0002571982147222523, "compression/movement_sparsity/linear_layer_sparsity": 0.9061915715785843, "compression/movement_sparsity/model_sparsity": 0.8750611033235107, "compression_loss": 101.77783203125, "distillation_loss": 4.682574272155762, "epoch": 3.98, "learning_rate": 4.082028804007514e-05, "loss": 106.5937, "step": 4707, "task_loss": 2.442762613296509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9604690510666232, "compression/movement_sparsity/importance_threshold": -0.0002565615857671767, "compression/movement_sparsity/linear_layer_sparsity": 0.9062992706606714, "compression/movement_sparsity/model_sparsity": 0.875165102610807, "compression_loss": 101.78755187988281, "distillation_loss": 4.708950042724609, "epoch": 3.98, "learning_rate": 4.081715716969318e-05, "loss": 106.2357, "step": 4708, "task_loss": 2.4931373596191406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9605669807056174, "compression/movement_sparsity/importance_threshold": -0.0002559260082221921, "compression/movement_sparsity/linear_layer_sparsity": 0.9063172046087957, "compression/movement_sparsity/model_sparsity": 0.8751824204726419, "compression_loss": 101.79730987548828, "distillation_loss": 4.2543535232543945, "epoch": 3.98, "learning_rate": 4.081402629931121e-05, "loss": 106.1018, "step": 4709, "task_loss": 1.865181803703308 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9606647484774767, "compression/movement_sparsity/importance_threshold": -0.000255291481218365, "compression/movement_sparsity/linear_layer_sparsity": 0.9063857804968696, "compression/movement_sparsity/model_sparsity": 0.8752486405679963, "compression_loss": 101.8070068359375, "distillation_loss": 4.638772487640381, "epoch": 3.98, "learning_rate": 4.081089542892924e-05, "loss": 106.341, "step": 4710, "task_loss": 2.0091705322265625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9607623545160862, "compression/movement_sparsity/importance_threshold": -0.0002546580038867612, "compression/movement_sparsity/linear_layer_sparsity": 0.9064151974184272, "compression/movement_sparsity/model_sparsity": 0.8752770469278015, "compression_loss": 101.81666564941406, "distillation_loss": 4.170865535736084, "epoch": 3.98, "learning_rate": 4.080776455854728e-05, "loss": 106.1386, "step": 4711, "task_loss": 2.0149080753326416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9608597989553309, "compression/movement_sparsity/importance_threshold": -0.0002540255753584464, "compression/movement_sparsity/linear_layer_sparsity": 0.9064427660940013, "compression/movement_sparsity/model_sparsity": 0.8753036685345585, "compression_loss": 101.82630920410156, "distillation_loss": 3.6139655113220215, "epoch": 3.98, "learning_rate": 4.0804633688165314e-05, "loss": 105.5611, "step": 4712, "task_loss": 3.2421884536743164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.960957081929096, "compression/movement_sparsity/importance_threshold": -0.0002533941947644872, "compression/movement_sparsity/linear_layer_sparsity": 0.9064676995285279, "compression/movement_sparsity/model_sparsity": 0.8753277454289049, "compression_loss": 101.83583068847656, "distillation_loss": 5.066997051239014, "epoch": 3.98, "learning_rate": 4.0801502817783346e-05, "loss": 106.1238, "step": 4713, "task_loss": 2.9662957191467285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9610542035712667, "compression/movement_sparsity/importance_threshold": -0.00025276386123594934, "compression/movement_sparsity/linear_layer_sparsity": 0.9064497775045712, "compression/movement_sparsity/model_sparsity": 0.8753104390816057, "compression_loss": 101.84535217285156, "distillation_loss": 4.003117561340332, "epoch": 3.98, "learning_rate": 4.0798371947401384e-05, "loss": 106.3934, "step": 4714, "task_loss": 2.367861747741699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.961151164015728, "compression/movement_sparsity/importance_threshold": -0.00025213457390389944, "compression/movement_sparsity/linear_layer_sparsity": 0.9064926448872221, "compression/movement_sparsity/model_sparsity": 0.875351833837787, "compression_loss": 101.8548583984375, "distillation_loss": 4.4803900718688965, "epoch": 3.99, "learning_rate": 4.0795241077019416e-05, "loss": 105.821, "step": 4715, "task_loss": 2.670651912689209 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9612479633963652, "compression/movement_sparsity/importance_threshold": -0.00025150633189940236, "compression/movement_sparsity/linear_layer_sparsity": 0.9065264856749727, "compression/movement_sparsity/model_sparsity": 0.8753845120903719, "compression_loss": 101.8642807006836, "distillation_loss": 3.3339970111846924, "epoch": 3.99, "learning_rate": 4.079211020663745e-05, "loss": 105.5701, "step": 4716, "task_loss": 1.3779469728469849 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9613446018470634, "compression/movement_sparsity/importance_threshold": -0.00025087913435352556, "compression/movement_sparsity/linear_layer_sparsity": 0.9065990561592047, "compression/movement_sparsity/model_sparsity": 0.8754545895552176, "compression_loss": 101.87374114990234, "distillation_loss": 5.220317840576172, "epoch": 3.99, "learning_rate": 4.078897933625548e-05, "loss": 106.4978, "step": 4717, "task_loss": 2.2851827144622803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9614410795017077, "compression/movement_sparsity/importance_threshold": -0.0002502529803973339, "compression/movement_sparsity/linear_layer_sparsity": 0.9066647225503753, "compression/movement_sparsity/model_sparsity": 0.8755180001038381, "compression_loss": 101.88316345214844, "distillation_loss": 5.500280380249023, "epoch": 3.99, "learning_rate": 4.078584846587352e-05, "loss": 106.4076, "step": 4718, "task_loss": 3.4764060974121094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9615373964941832, "compression/movement_sparsity/importance_threshold": -0.00024962786916189403, "compression/movement_sparsity/linear_layer_sparsity": 0.9066503777767094, "compression/movement_sparsity/model_sparsity": 0.8755041481172773, "compression_loss": 101.89256286621094, "distillation_loss": 4.245757102966309, "epoch": 3.99, "learning_rate": 4.078271759549155e-05, "loss": 107.0876, "step": 4719, "task_loss": 3.123845100402832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9616335529583752, "compression/movement_sparsity/importance_threshold": -0.0002490037997782725, "compression/movement_sparsity/linear_layer_sparsity": 0.906676587097173, "compression/movement_sparsity/model_sparsity": 0.8755294570669536, "compression_loss": 101.9019546508789, "distillation_loss": 4.6381635665893555, "epoch": 3.99, "learning_rate": 4.077958672510958e-05, "loss": 106.1523, "step": 4720, "task_loss": 2.3972623348236084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9617295490281687, "compression/movement_sparsity/importance_threshold": -0.00024838077137753333, "compression/movement_sparsity/linear_layer_sparsity": 0.9067604497681561, "compression/movement_sparsity/model_sparsity": 0.8756104387971967, "compression_loss": 101.9113540649414, "distillation_loss": 4.64486026763916, "epoch": 3.99, "learning_rate": 4.077645585472762e-05, "loss": 106.9722, "step": 4721, "task_loss": 1.5838329792022705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9618253848374488, "compression/movement_sparsity/importance_threshold": -0.00024775878309074486, "compression/movement_sparsity/linear_layer_sparsity": 0.9067283141363773, "compression/movement_sparsity/model_sparsity": 0.8755794071232303, "compression_loss": 101.92062377929688, "distillation_loss": 4.569716453552246, "epoch": 3.99, "learning_rate": 4.077332498434565e-05, "loss": 106.4057, "step": 4722, "task_loss": 3.13065505027771 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9619210605201008, "compression/movement_sparsity/importance_threshold": -0.00024713783404897195, "compression/movement_sparsity/linear_layer_sparsity": 0.9068822193680535, "compression/movement_sparsity/model_sparsity": 0.8757280252367305, "compression_loss": 101.93000793457031, "distillation_loss": 4.673598766326904, "epoch": 3.99, "learning_rate": 4.0770194113963684e-05, "loss": 106.1801, "step": 4723, "task_loss": 2.191633462905884 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9620165762100098, "compression/movement_sparsity/importance_threshold": -0.0002465179233832812, "compression/movement_sparsity/linear_layer_sparsity": 0.9069083213710084, "compression/movement_sparsity/model_sparsity": 0.8757532305555847, "compression_loss": 101.93937683105469, "distillation_loss": 6.7890119552612305, "epoch": 3.99, "learning_rate": 4.0767063243581715e-05, "loss": 106.6318, "step": 4724, "task_loss": 3.5493054389953613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9621119320410608, "compression/movement_sparsity/importance_threshold": -0.0002458990502247384, "compression/movement_sparsity/linear_layer_sparsity": 0.9069389903301678, "compression/movement_sparsity/model_sparsity": 0.8757828459416483, "compression_loss": 101.94866943359375, "distillation_loss": 3.527641773223877, "epoch": 3.99, "learning_rate": 4.0763932373199754e-05, "loss": 106.9377, "step": 4725, "task_loss": 2.2168400287628174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9622071281471392, "compression/movement_sparsity/importance_threshold": -0.0002452812137044083, "compression/movement_sparsity/linear_layer_sparsity": 0.9069298921902617, "compression/movement_sparsity/model_sparsity": 0.875774060350837, "compression_loss": 101.95809936523438, "distillation_loss": 3.8685052394866943, "epoch": 3.99, "learning_rate": 4.0760801502817786e-05, "loss": 106.411, "step": 4726, "task_loss": 1.513759970664978 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9623021646621298, "compression/movement_sparsity/importance_threshold": -0.00024466441295335935, "compression/movement_sparsity/linear_layer_sparsity": 0.9069678945125171, "compression/movement_sparsity/model_sparsity": 0.8758107571764143, "compression_loss": 101.96743774414062, "distillation_loss": 5.944355487823486, "epoch": 4.0, "learning_rate": 4.075767063243582e-05, "loss": 107.6945, "step": 4727, "task_loss": 2.888057231903076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9623970417199179, "compression/movement_sparsity/importance_threshold": -0.00024404864710265723, "compression/movement_sparsity/linear_layer_sparsity": 0.9070499327858519, "compression/movement_sparsity/model_sparsity": 0.8758899771826807, "compression_loss": 101.97681427001953, "distillation_loss": 3.9286811351776123, "epoch": 4.0, "learning_rate": 4.075453976205385e-05, "loss": 106.4211, "step": 4728, "task_loss": 3.8434433937072754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9624917594543888, "compression/movement_sparsity/importance_threshold": -0.00024343391528336596, "compression/movement_sparsity/linear_layer_sparsity": 0.9071281434013755, "compression/movement_sparsity/model_sparsity": 0.8759655010229571, "compression_loss": 101.98612213134766, "distillation_loss": 4.36923885345459, "epoch": 4.0, "learning_rate": 4.075140889167189e-05, "loss": 106.2515, "step": 4729, "task_loss": 1.3009731769561768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9625863179994274, "compression/movement_sparsity/importance_threshold": -0.000242820216626553, "compression/movement_sparsity/linear_layer_sparsity": 0.9071379569913398, "compression/movement_sparsity/model_sparsity": 0.875974977485916, "compression_loss": 101.99546813964844, "distillation_loss": 5.805533409118652, "epoch": 4.0, "learning_rate": 4.074827802128992e-05, "loss": 107.2558, "step": 4730, "task_loss": 4.205356597900391 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9626807174889191, "compression/movement_sparsity/importance_threshold": -0.00024220755026328322, "compression/movement_sparsity/linear_layer_sparsity": 0.9071365380153912, "compression/movement_sparsity/model_sparsity": 0.8759736072561565, "compression_loss": 102.00474548339844, "distillation_loss": 3.9001264572143555, "epoch": 4.0, "learning_rate": 4.074514715090795e-05, "loss": 105.4999, "step": 4731, "task_loss": 2.223771333694458 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9627749580567488, "compression/movement_sparsity/importance_threshold": -0.00024159591532462495, "compression/movement_sparsity/linear_layer_sparsity": 0.9071510378032364, "compression/movement_sparsity/model_sparsity": 0.8759876089316826, "compression_loss": 102.0140609741211, "distillation_loss": 3.3805928230285645, "epoch": 4.0, "learning_rate": 4.074201628052599e-05, "loss": 106.0391, "step": 4732, "task_loss": 2.718658208847046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9629629629629629, "compression/movement_sparsity/importance_threshold": -0.00024037573624540332, "compression/movement_sparsity/linear_layer_sparsity": 0.9072069502252809, "compression/movement_sparsity/model_sparsity": 0.8760416005900232, "compression_loss": 102.03318786621094, "distillation_loss": 2.874843120574951, "epoch": 4.0, "learning_rate": 4.073888541014402e-05, "loss": 184.8975, "step": 4733, "task_loss": 1.763904333114624 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9630567275691178, "compression/movement_sparsity/importance_threshold": -0.00023976719036697056, "compression/movement_sparsity/linear_layer_sparsity": 0.9072741309857414, "compression/movement_sparsity/model_sparsity": 0.8761064734846897, "compression_loss": 102.04239654541016, "distillation_loss": 4.776275157928467, "epoch": 4.0, "learning_rate": 4.0735754539762054e-05, "loss": 106.1724, "step": 4734, "task_loss": 3.0733437538146973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9631503337891512, "compression/movement_sparsity/importance_threshold": -0.00023915967243741312, "compression/movement_sparsity/linear_layer_sparsity": 0.9073134807389397, "compression/movement_sparsity/model_sparsity": 0.8761444714528117, "compression_loss": 102.05150604248047, "distillation_loss": 4.8445143699646, "epoch": 4.0, "learning_rate": 4.0732623669380085e-05, "loss": 106.0648, "step": 4735, "task_loss": 3.791348695755005 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9632437817569484, "compression/movement_sparsity/importance_threshold": -0.00023855318158779674, "compression/movement_sparsity/linear_layer_sparsity": 0.9073618570870384, "compression/movement_sparsity/model_sparsity": 0.8761911859245302, "compression_loss": 102.06065368652344, "distillation_loss": 5.363153457641602, "epoch": 4.0, "learning_rate": 4.0729492798998124e-05, "loss": 106.3207, "step": 4736, "task_loss": 4.430835247039795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9633370716063946, "compression/movement_sparsity/importance_threshold": -0.00023794771694918542, "compression/movement_sparsity/linear_layer_sparsity": 0.90735145921286, "compression/movement_sparsity/model_sparsity": 0.8761811452493172, "compression_loss": 102.06975555419922, "distillation_loss": 4.721225738525391, "epoch": 4.0, "learning_rate": 4.0726361928616156e-05, "loss": 106.4588, "step": 4737, "task_loss": 2.8120925426483154 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9634302034713748, "compression/movement_sparsity/importance_threshold": -0.0002373432776526475, "compression/movement_sparsity/linear_layer_sparsity": 0.9072600485437634, "compression/movement_sparsity/model_sparsity": 0.8760928748179163, "compression_loss": 102.07886505126953, "distillation_loss": 3.2372875213623047, "epoch": 4.01, "learning_rate": 4.072323105823419e-05, "loss": 105.8973, "step": 4738, "task_loss": 2.0505712032318115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9635231774857742, "compression/movement_sparsity/importance_threshold": -0.00023673986282924699, "compression/movement_sparsity/linear_layer_sparsity": 0.9073137669189629, "compression/movement_sparsity/model_sparsity": 0.8761447478016707, "compression_loss": 102.08795166015625, "distillation_loss": 5.051644802093506, "epoch": 4.01, "learning_rate": 4.072010018785222e-05, "loss": 106.9286, "step": 4739, "task_loss": 2.7814853191375732 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9636159937834781, "compression/movement_sparsity/importance_threshold": -0.00023613747161005134, "compression/movement_sparsity/linear_layer_sparsity": 0.9072799619037153, "compression/movement_sparsity/model_sparsity": 0.8761121040926932, "compression_loss": 102.09706115722656, "distillation_loss": 3.7983548641204834, "epoch": 4.01, "learning_rate": 4.071696931747026e-05, "loss": 106.2101, "step": 4740, "task_loss": 2.5259625911712646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9637086524983712, "compression/movement_sparsity/importance_threshold": -0.00023553610312612803, "compression/movement_sparsity/linear_layer_sparsity": 0.9073895211559536, "compression/movement_sparsity/model_sparsity": 0.8762178996475735, "compression_loss": 102.10615539550781, "distillation_loss": 5.121649742126465, "epoch": 4.01, "learning_rate": 4.071383844708829e-05, "loss": 106.2127, "step": 4741, "task_loss": 2.024514675140381 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9638011537643392, "compression/movement_sparsity/importance_threshold": -0.00023493575650853932, "compression/movement_sparsity/linear_layer_sparsity": 0.9074158854905965, "compression/movement_sparsity/model_sparsity": 0.8762433582862151, "compression_loss": 102.1152572631836, "distillation_loss": 4.735377788543701, "epoch": 4.01, "learning_rate": 4.071070757670633e-05, "loss": 106.9508, "step": 4742, "task_loss": 2.420778751373291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9638934977152669, "compression/movement_sparsity/importance_threshold": -0.00023433643088835356, "compression/movement_sparsity/linear_layer_sparsity": 0.9073821162478517, "compression/movement_sparsity/model_sparsity": 0.8762107491208451, "compression_loss": 102.12427520751953, "distillation_loss": 2.62825870513916, "epoch": 4.01, "learning_rate": 4.070757670632436e-05, "loss": 105.3587, "step": 4743, "task_loss": 1.0795023441314697 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9639856844850394, "compression/movement_sparsity/importance_threshold": -0.00023373812539663735, "compression/movement_sparsity/linear_layer_sparsity": 0.9073810788452674, "compression/movement_sparsity/model_sparsity": 0.8762097473562309, "compression_loss": 102.13330078125, "distillation_loss": 3.961312770843506, "epoch": 4.01, "learning_rate": 4.07044458359424e-05, "loss": 105.9915, "step": 4744, "task_loss": 2.192967414855957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9640777142075421, "compression/movement_sparsity/importance_threshold": -0.00023314083916445554, "compression/movement_sparsity/linear_layer_sparsity": 0.9074420113418867, "compression/movement_sparsity/model_sparsity": 0.876268586634141, "compression_loss": 102.14239501953125, "distillation_loss": 3.751586437225342, "epoch": 4.01, "learning_rate": 4.070131496556043e-05, "loss": 105.9125, "step": 4745, "task_loss": 2.422184944152832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9641695870166599, "compression/movement_sparsity/importance_threshold": -0.00023254457132287475, "compression/movement_sparsity/linear_layer_sparsity": 0.9075440941410173, "compression/movement_sparsity/model_sparsity": 0.8763671625750781, "compression_loss": 102.15130615234375, "distillation_loss": 3.3452649116516113, "epoch": 4.01, "learning_rate": 4.069818409517846e-05, "loss": 106.0835, "step": 4746, "task_loss": 2.490675449371338 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9642613030462781, "compression/movement_sparsity/importance_threshold": -0.00023194932100295985, "compression/movement_sparsity/linear_layer_sparsity": 0.9075618373024594, "compression/movement_sparsity/model_sparsity": 0.8763842962043403, "compression_loss": 102.16027069091797, "distillation_loss": 6.149139881134033, "epoch": 4.01, "learning_rate": 4.06950532247965e-05, "loss": 107.0028, "step": 4747, "task_loss": 3.5166282653808594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9643528624302817, "compression/movement_sparsity/importance_threshold": -0.00023135508733577916, "compression/movement_sparsity/linear_layer_sparsity": 0.9075771240853686, "compression/movement_sparsity/model_sparsity": 0.876399057839229, "compression_loss": 102.16920471191406, "distillation_loss": 4.035058498382568, "epoch": 4.01, "learning_rate": 4.069192235441453e-05, "loss": 106.1605, "step": 4748, "task_loss": 2.7094407081604004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.964444265302556, "compression/movement_sparsity/importance_threshold": -0.0002307618694523967, "compression/movement_sparsity/linear_layer_sparsity": 0.9076518647681102, "compression/movement_sparsity/model_sparsity": 0.8764712309495891, "compression_loss": 102.17813873291016, "distillation_loss": 3.5685768127441406, "epoch": 4.01, "learning_rate": 4.0688791484032564e-05, "loss": 105.9053, "step": 4749, "task_loss": 2.195991277694702 }, { "compression/movement_sparsity/importance_regularization_factor": 0.964535511796986, "compression/movement_sparsity/importance_threshold": -0.0002301696664838782, "compression/movement_sparsity/linear_layer_sparsity": 0.9077283225309913, "compression/movement_sparsity/model_sparsity": 0.8765450621531038, "compression_loss": 102.18708038330078, "distillation_loss": 4.508792877197266, "epoch": 4.02, "learning_rate": 4.0685660613650596e-05, "loss": 106.4583, "step": 4750, "task_loss": 2.4835891723632812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.964626602047457, "compression/movement_sparsity/importance_threshold": -0.00022957847756129197, "compression/movement_sparsity/linear_layer_sparsity": 0.9077418803095934, "compression/movement_sparsity/model_sparsity": 0.8765581541803021, "compression_loss": 102.19597625732422, "distillation_loss": 5.926754951477051, "epoch": 4.02, "learning_rate": 4.0682529743268635e-05, "loss": 106.3283, "step": 4751, "task_loss": 2.18298077583313 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9647175361878539, "compression/movement_sparsity/importance_threshold": -0.00022898830181570205, "compression/movement_sparsity/linear_layer_sparsity": 0.9077644408347604, "compression/movement_sparsity/model_sparsity": 0.8765799396820254, "compression_loss": 102.20492553710938, "distillation_loss": 4.225768089294434, "epoch": 4.02, "learning_rate": 4.0679398872886666e-05, "loss": 106.7032, "step": 4752, "task_loss": 1.9945875406265259 }, { "compression/movement_sparsity/importance_regularization_factor": 0.964808314352062, "compression/movement_sparsity/importance_threshold": -0.0002283991383781759, "compression/movement_sparsity/linear_layer_sparsity": 0.9077887422884023, "compression/movement_sparsity/model_sparsity": 0.8766034063059747, "compression_loss": 102.2137680053711, "distillation_loss": 3.4410862922668457, "epoch": 4.02, "learning_rate": 4.06762680025047e-05, "loss": 106.8158, "step": 4753, "task_loss": 2.2282521724700928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9648989366739665, "compression/movement_sparsity/importance_threshold": -0.00022781098637977837, "compression/movement_sparsity/linear_layer_sparsity": 0.9078505571734267, "compression/movement_sparsity/model_sparsity": 0.8766630976595335, "compression_loss": 102.22260284423828, "distillation_loss": 3.0488502979278564, "epoch": 4.02, "learning_rate": 4.067313713212273e-05, "loss": 106.6297, "step": 4754, "task_loss": 1.8006576299667358 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9649894032874524, "compression/movement_sparsity/importance_threshold": -0.00022722384495157608, "compression/movement_sparsity/linear_layer_sparsity": 0.9079082582206166, "compression/movement_sparsity/model_sparsity": 0.8767188164982433, "compression_loss": 102.23145294189453, "distillation_loss": 6.169253349304199, "epoch": 4.02, "learning_rate": 4.067000626174077e-05, "loss": 106.3496, "step": 4755, "task_loss": 3.5211446285247803 }, { "compression/movement_sparsity/importance_regularization_factor": 0.965079714326405, "compression/movement_sparsity/importance_threshold": -0.00022663771322463475, "compression/movement_sparsity/linear_layer_sparsity": 0.9079507201815679, "compression/movement_sparsity/model_sparsity": 0.8767598197602077, "compression_loss": 102.24027252197266, "distillation_loss": 5.21381950378418, "epoch": 4.02, "learning_rate": 4.06668753913588e-05, "loss": 106.7939, "step": 4756, "task_loss": 2.7255361080169678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9651698699247092, "compression/movement_sparsity/importance_threshold": -0.000226052590330021, "compression/movement_sparsity/linear_layer_sparsity": 0.9080693537253771, "compression/movement_sparsity/model_sparsity": 0.8768743778768276, "compression_loss": 102.24918365478516, "distillation_loss": 4.492337226867676, "epoch": 4.02, "learning_rate": 4.066374452097683e-05, "loss": 106.6864, "step": 4757, "task_loss": 2.5605671405792236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9652598702162504, "compression/movement_sparsity/importance_threshold": -0.00022546847539880056, "compression/movement_sparsity/linear_layer_sparsity": 0.9081508434870006, "compression/movement_sparsity/model_sparsity": 0.8769530682144475, "compression_loss": 102.25791931152344, "distillation_loss": 3.2782206535339355, "epoch": 4.02, "learning_rate": 4.066061365059487e-05, "loss": 107.0034, "step": 4758, "task_loss": 1.8193538188934326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9653497153349135, "compression/movement_sparsity/importance_threshold": -0.00022488536756203915, "compression/movement_sparsity/linear_layer_sparsity": 0.9082272654773788, "compression/movement_sparsity/model_sparsity": 0.8770268648743547, "compression_loss": 102.26667785644531, "distillation_loss": 2.972283363342285, "epoch": 4.02, "learning_rate": 4.06574827802129e-05, "loss": 106.2226, "step": 4759, "task_loss": 1.8632569313049316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9654394054145838, "compression/movement_sparsity/importance_threshold": -0.0002243032659508034, "compression/movement_sparsity/linear_layer_sparsity": 0.9083182230281052, "compression/movement_sparsity/model_sparsity": 0.8771146977533955, "compression_loss": 102.27548217773438, "distillation_loss": 2.9562759399414062, "epoch": 4.02, "learning_rate": 4.0654351909830934e-05, "loss": 105.9176, "step": 4760, "task_loss": 1.5847476720809937 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9655289405891465, "compression/movement_sparsity/importance_threshold": -0.00022372216969615902, "compression/movement_sparsity/linear_layer_sparsity": 0.9083499413140166, "compression/movement_sparsity/model_sparsity": 0.877145326418609, "compression_loss": 102.28416442871094, "distillation_loss": 3.8236136436462402, "epoch": 4.02, "learning_rate": 4.0651221039448966e-05, "loss": 106.839, "step": 4761, "task_loss": 2.673823833465576 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9656183209924866, "compression/movement_sparsity/importance_threshold": -0.00022314207792917175, "compression/movement_sparsity/linear_layer_sparsity": 0.908426947588609, "compression/movement_sparsity/model_sparsity": 0.87721968729077, "compression_loss": 102.2928695678711, "distillation_loss": 5.33426570892334, "epoch": 4.03, "learning_rate": 4.0648090169067004e-05, "loss": 107.3694, "step": 4762, "task_loss": 2.5103399753570557 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9657075467584892, "compression/movement_sparsity/importance_threshold": -0.0002225629897809082, "compression/movement_sparsity/linear_layer_sparsity": 0.9084943429840868, "compression/movement_sparsity/model_sparsity": 0.8772847674470808, "compression_loss": 102.30152130126953, "distillation_loss": 3.2841296195983887, "epoch": 4.03, "learning_rate": 4.0644959298685036e-05, "loss": 105.9625, "step": 4763, "task_loss": 1.6197351217269897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9657966180210396, "compression/movement_sparsity/importance_threshold": -0.00022198490438243412, "compression/movement_sparsity/linear_layer_sparsity": 0.9085635985497159, "compression/movement_sparsity/model_sparsity": 0.8773516438709755, "compression_loss": 102.3101806640625, "distillation_loss": 4.329470634460449, "epoch": 4.03, "learning_rate": 4.064182842830307e-05, "loss": 107.0394, "step": 4764, "task_loss": 1.833674430847168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9658855349140228, "compression/movement_sparsity/importance_threshold": -0.00022140782086481608, "compression/movement_sparsity/linear_layer_sparsity": 0.9086367413939943, "compression/movement_sparsity/model_sparsity": 0.8774222740335392, "compression_loss": 102.31887817382812, "distillation_loss": 4.805695056915283, "epoch": 4.03, "learning_rate": 4.06386975579211e-05, "loss": 107.0686, "step": 4765, "task_loss": 2.7547264099121094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9659742975713241, "compression/movement_sparsity/importance_threshold": -0.0002208317383591181, "compression/movement_sparsity/linear_layer_sparsity": 0.9086537929537136, "compression/movement_sparsity/model_sparsity": 0.8774387398197254, "compression_loss": 102.32748413085938, "distillation_loss": 4.294532775878906, "epoch": 4.03, "learning_rate": 4.063556668753914e-05, "loss": 106.8014, "step": 4766, "task_loss": 2.0070948600769043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9660629061268285, "compression/movement_sparsity/importance_threshold": -0.00022025665599640937, "compression/movement_sparsity/linear_layer_sparsity": 0.9086745648537353, "compression/movement_sparsity/model_sparsity": 0.8774587981410795, "compression_loss": 102.33613586425781, "distillation_loss": 5.629484176635742, "epoch": 4.03, "learning_rate": 4.063243581715717e-05, "loss": 106.4458, "step": 4767, "task_loss": 2.523003101348877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9661513607144212, "compression/movement_sparsity/importance_threshold": -0.0002196825729077539, "compression/movement_sparsity/linear_layer_sparsity": 0.9087417098416929, "compression/movement_sparsity/model_sparsity": 0.8775236364921386, "compression_loss": 102.34471130371094, "distillation_loss": 5.136823654174805, "epoch": 4.03, "learning_rate": 4.06293049467752e-05, "loss": 106.2364, "step": 4768, "task_loss": 2.2159478664398193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9662396614679873, "compression/movement_sparsity/importance_threshold": -0.00021910948822421745, "compression/movement_sparsity/linear_layer_sparsity": 0.9087871170720502, "compression/movement_sparsity/model_sparsity": 0.8775674838444442, "compression_loss": 102.35332489013672, "distillation_loss": 4.786014556884766, "epoch": 4.03, "learning_rate": 4.062617407639324e-05, "loss": 106.1696, "step": 4769, "task_loss": 2.288083791732788 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9663278085214121, "compression/movement_sparsity/importance_threshold": -0.00021853740107686658, "compression/movement_sparsity/linear_layer_sparsity": 0.9087930791558682, "compression/movement_sparsity/model_sparsity": 0.8775732411123415, "compression_loss": 102.36185455322266, "distillation_loss": 4.844325065612793, "epoch": 4.03, "learning_rate": 4.062304320601127e-05, "loss": 107.4598, "step": 4770, "task_loss": 2.6307532787323 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9664158020085805, "compression/movement_sparsity/importance_threshold": -0.00021796631059676878, "compression/movement_sparsity/linear_layer_sparsity": 0.9088663650901582, "compression/movement_sparsity/model_sparsity": 0.8776440094493347, "compression_loss": 102.37039947509766, "distillation_loss": 5.966167449951172, "epoch": 4.03, "learning_rate": 4.0619912335629304e-05, "loss": 107.1645, "step": 4771, "task_loss": 2.768897771835327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9665036420633777, "compression/movement_sparsity/importance_threshold": -0.00021739621591498805, "compression/movement_sparsity/linear_layer_sparsity": 0.9088818784322524, "compression/movement_sparsity/model_sparsity": 0.8776589898604034, "compression_loss": 102.37899017333984, "distillation_loss": 4.714101791381836, "epoch": 4.03, "learning_rate": 4.0616781465247336e-05, "loss": 106.7347, "step": 4772, "task_loss": 2.030451774597168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9665913288196889, "compression/movement_sparsity/importance_threshold": -0.00021682711616259186, "compression/movement_sparsity/linear_layer_sparsity": 0.9089540196464494, "compression/movement_sparsity/model_sparsity": 0.8777286528019603, "compression_loss": 102.3875732421875, "distillation_loss": 5.123327255249023, "epoch": 4.03, "learning_rate": 4.0613650594865374e-05, "loss": 106.6007, "step": 4773, "task_loss": 3.2851576805114746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9666788624113993, "compression/movement_sparsity/importance_threshold": -0.0002162590104706442, "compression/movement_sparsity/linear_layer_sparsity": 0.908935990304984, "compression/movement_sparsity/model_sparsity": 0.877711242823839, "compression_loss": 102.39608764648438, "distillation_loss": 5.3328752517700195, "epoch": 4.04, "learning_rate": 4.0610519724483406e-05, "loss": 106.8239, "step": 4774, "task_loss": 2.3546202182769775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.966766242972394, "compression/movement_sparsity/importance_threshold": -0.00021569189797021344, "compression/movement_sparsity/linear_layer_sparsity": 0.9089985564125693, "compression/movement_sparsity/model_sparsity": 0.877771659593153, "compression_loss": 102.40460205078125, "distillation_loss": 4.534553050994873, "epoch": 4.04, "learning_rate": 4.0607388854101445e-05, "loss": 106.3217, "step": 4775, "task_loss": 2.3237061500549316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.966853470636558, "compression/movement_sparsity/importance_threshold": -0.00021512577779236527, "compression/movement_sparsity/linear_layer_sparsity": 0.9090352470763848, "compression/movement_sparsity/model_sparsity": 0.8778070898197928, "compression_loss": 102.41315460205078, "distillation_loss": 4.813056468963623, "epoch": 4.04, "learning_rate": 4.0604257983719476e-05, "loss": 106.86, "step": 4776, "task_loss": 2.5021116733551025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9669405455377766, "compression/movement_sparsity/importance_threshold": -0.00021456064906816372, "compression/movement_sparsity/linear_layer_sparsity": 0.9090719138918651, "compression/movement_sparsity/model_sparsity": 0.8778424970173611, "compression_loss": 102.42166900634766, "distillation_loss": 5.21835994720459, "epoch": 4.04, "learning_rate": 4.060112711333751e-05, "loss": 106.4006, "step": 4777, "task_loss": 3.1152639389038086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9670274678099349, "compression/movement_sparsity/importance_threshold": -0.000213996510928678, "compression/movement_sparsity/linear_layer_sparsity": 0.9091428865376339, "compression/movement_sparsity/model_sparsity": 0.8779110315344101, "compression_loss": 102.43013763427734, "distillation_loss": 3.961397171020508, "epoch": 4.04, "learning_rate": 4.059799624295555e-05, "loss": 106.4147, "step": 4778, "task_loss": 2.514514923095703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9671142375869181, "compression/movement_sparsity/importance_threshold": -0.0002134333625049712, "compression/movement_sparsity/linear_layer_sparsity": 0.9091958894627752, "compression/movement_sparsity/model_sparsity": 0.8779622136460169, "compression_loss": 102.43856048583984, "distillation_loss": 5.484292030334473, "epoch": 4.04, "learning_rate": 4.059486537257358e-05, "loss": 106.9094, "step": 4779, "task_loss": 2.2985546588897705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9672008550026112, "compression/movement_sparsity/importance_threshold": -0.00021287120292810997, "compression/movement_sparsity/linear_layer_sparsity": 0.9092511222072646, "compression/movement_sparsity/model_sparsity": 0.8780155489758171, "compression_loss": 102.44706726074219, "distillation_loss": 3.8057703971862793, "epoch": 4.04, "learning_rate": 4.059173450219161e-05, "loss": 106.6663, "step": 4780, "task_loss": 1.9703232049942017 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9672873201908995, "compression/movement_sparsity/importance_threshold": -0.00021231003132916176, "compression/movement_sparsity/linear_layer_sparsity": 0.90929109201718, "compression/movement_sparsity/model_sparsity": 0.8780541456998004, "compression_loss": 102.45549774169922, "distillation_loss": 4.496603965759277, "epoch": 4.04, "learning_rate": 4.058860363180965e-05, "loss": 106.6457, "step": 4781, "task_loss": 2.514622926712036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9673736332856679, "compression/movement_sparsity/importance_threshold": -0.00021174984683919144, "compression/movement_sparsity/linear_layer_sparsity": 0.9093007386687973, "compression/movement_sparsity/model_sparsity": 0.8780634609592582, "compression_loss": 102.46388244628906, "distillation_loss": 5.529583930969238, "epoch": 4.04, "learning_rate": 4.058547276142768e-05, "loss": 106.7105, "step": 4782, "task_loss": 2.671614408493042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9674597944208019, "compression/movement_sparsity/importance_threshold": -0.00021119064858926562, "compression/movement_sparsity/linear_layer_sparsity": 0.9093710435611784, "compression/movement_sparsity/model_sparsity": 0.8781313506623029, "compression_loss": 102.47227478027344, "distillation_loss": 3.6557211875915527, "epoch": 4.04, "learning_rate": 4.058234189104571e-05, "loss": 106.3312, "step": 4783, "task_loss": 2.012655735015869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9675458037301863, "compression/movement_sparsity/importance_threshold": -0.00021063243571045002, "compression/movement_sparsity/linear_layer_sparsity": 0.9093886317084413, "compression/movement_sparsity/model_sparsity": 0.8781483346025998, "compression_loss": 102.48070526123047, "distillation_loss": 3.5731394290924072, "epoch": 4.04, "learning_rate": 4.057921102066375e-05, "loss": 106.5429, "step": 4784, "task_loss": 2.142951726913452 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9676316613477064, "compression/movement_sparsity/importance_threshold": -0.00021007520733381125, "compression/movement_sparsity/linear_layer_sparsity": 0.9094970581647542, "compression/movement_sparsity/model_sparsity": 0.8782530362765796, "compression_loss": 102.48908996582031, "distillation_loss": 5.857938766479492, "epoch": 4.04, "learning_rate": 4.057608015028178e-05, "loss": 107.0811, "step": 4785, "task_loss": 3.9624807834625244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9677173674072472, "compression/movement_sparsity/importance_threshold": -0.00020951896259041505, "compression/movement_sparsity/linear_layer_sparsity": 0.9094576010940472, "compression/movement_sparsity/model_sparsity": 0.8782149346776355, "compression_loss": 102.49750518798828, "distillation_loss": 5.134703159332275, "epoch": 4.05, "learning_rate": 4.0572949279899815e-05, "loss": 107.221, "step": 4786, "task_loss": 3.6493544578552246 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9678029220426941, "compression/movement_sparsity/importance_threshold": -0.00020896370061132628, "compression/movement_sparsity/linear_layer_sparsity": 0.9094636585712061, "compression/movement_sparsity/model_sparsity": 0.878220784061819, "compression_loss": 102.50590515136719, "distillation_loss": 5.3236613273620605, "epoch": 4.05, "learning_rate": 4.0569818409517846e-05, "loss": 107.408, "step": 4787, "task_loss": 2.5965070724487305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9678883253879321, "compression/movement_sparsity/importance_threshold": -0.0002084094205276124, "compression/movement_sparsity/linear_layer_sparsity": 0.9095075395081061, "compression/movement_sparsity/model_sparsity": 0.878263157553543, "compression_loss": 102.51426696777344, "distillation_loss": 4.180915355682373, "epoch": 4.05, "learning_rate": 4.0566687539135885e-05, "loss": 107.0379, "step": 4788, "task_loss": 2.055954933166504 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9679735775768464, "compression/movement_sparsity/importance_threshold": -0.00020785612147033918, "compression/movement_sparsity/linear_layer_sparsity": 0.9095293368865445, "compression/movement_sparsity/model_sparsity": 0.8782842061249754, "compression_loss": 102.52265167236328, "distillation_loss": 6.06610107421875, "epoch": 4.05, "learning_rate": 4.056355666875392e-05, "loss": 107.4861, "step": 4789, "task_loss": 2.8245134353637695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.968058678743322, "compression/movement_sparsity/importance_threshold": -0.00020730380257057233, "compression/movement_sparsity/linear_layer_sparsity": 0.9095376957280572, "compression/movement_sparsity/model_sparsity": 0.8782922778145673, "compression_loss": 102.53105163574219, "distillation_loss": 4.640871047973633, "epoch": 4.05, "learning_rate": 4.056042579837195e-05, "loss": 106.7592, "step": 4790, "task_loss": 3.0910794734954834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9681436290212441, "compression/movement_sparsity/importance_threshold": -0.00020675246295937845, "compression/movement_sparsity/linear_layer_sparsity": 0.9095285856639834, "compression/movement_sparsity/model_sparsity": 0.8782834807092204, "compression_loss": 102.53939819335938, "distillation_loss": 4.669769763946533, "epoch": 4.05, "learning_rate": 4.055729492798998e-05, "loss": 106.6228, "step": 4791, "task_loss": 2.3145253658294678 }, { "compression/movement_sparsity/importance_regularization_factor": 0.968228428544498, "compression/movement_sparsity/importance_threshold": -0.00020620210176782242, "compression/movement_sparsity/linear_layer_sparsity": 0.909603493285072, "compression/movement_sparsity/model_sparsity": 0.8783558150230817, "compression_loss": 102.54769897460938, "distillation_loss": 4.155488967895508, "epoch": 4.05, "learning_rate": 4.055416405760802e-05, "loss": 106.9444, "step": 4792, "task_loss": 2.3427414894104004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9683130774469686, "compression/movement_sparsity/importance_threshold": -0.0002056527181269717, "compression/movement_sparsity/linear_layer_sparsity": 0.9096858773592681, "compression/movement_sparsity/model_sparsity": 0.8784353689508861, "compression_loss": 102.55604553222656, "distillation_loss": 4.8056793212890625, "epoch": 4.05, "learning_rate": 4.055103318722605e-05, "loss": 106.9131, "step": 4793, "task_loss": 3.776576280593872 }, { "compression/movement_sparsity/importance_regularization_factor": 0.968397575862541, "compression/movement_sparsity/importance_threshold": -0.00020510431116789202, "compression/movement_sparsity/linear_layer_sparsity": 0.9097250601741196, "compression/movement_sparsity/model_sparsity": 0.878473205715507, "compression_loss": 102.5643539428711, "distillation_loss": 4.079629898071289, "epoch": 4.05, "learning_rate": 4.054790231684408e-05, "loss": 107.1674, "step": 4794, "task_loss": 2.421884775161743 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9684819239251007, "compression/movement_sparsity/importance_threshold": -0.00020455688002164827, "compression/movement_sparsity/linear_layer_sparsity": 0.9098060610448699, "compression/movement_sparsity/model_sparsity": 0.8785514239571593, "compression_loss": 102.5726318359375, "distillation_loss": 3.0924665927886963, "epoch": 4.05, "learning_rate": 4.054477144646212e-05, "loss": 106.7889, "step": 4795, "task_loss": 1.2240666151046753 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9685661217685325, "compression/movement_sparsity/importance_threshold": -0.0002040104238193079, "compression/movement_sparsity/linear_layer_sparsity": 0.9098653122338526, "compression/movement_sparsity/model_sparsity": 0.8786086396855224, "compression_loss": 102.58097076416016, "distillation_loss": 4.186288356781006, "epoch": 4.05, "learning_rate": 4.054164057608015e-05, "loss": 106.6493, "step": 4796, "task_loss": 3.1425390243530273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9686501695267218, "compression/movement_sparsity/importance_threshold": -0.0002034649416919358, "compression/movement_sparsity/linear_layer_sparsity": 0.9099003215900314, "compression/movement_sparsity/model_sparsity": 0.8786424463626151, "compression_loss": 102.58924102783203, "distillation_loss": 4.488948822021484, "epoch": 4.05, "learning_rate": 4.0538509705698185e-05, "loss": 107.036, "step": 4797, "task_loss": 2.311079263687134 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9687340673335535, "compression/movement_sparsity/importance_threshold": -0.0002029204327705994, "compression/movement_sparsity/linear_layer_sparsity": 0.9099064148396934, "compression/movement_sparsity/model_sparsity": 0.8786483302904061, "compression_loss": 102.5975341796875, "distillation_loss": 5.079647541046143, "epoch": 4.06, "learning_rate": 4.0535378835316216e-05, "loss": 106.577, "step": 4798, "task_loss": 2.1009864807128906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9688178153229129, "compression/movement_sparsity/importance_threshold": -0.00020237689618636274, "compression/movement_sparsity/linear_layer_sparsity": 0.9099640324177098, "compression/movement_sparsity/model_sparsity": 0.8787039685273653, "compression_loss": 102.60585021972656, "distillation_loss": 2.9498276710510254, "epoch": 4.06, "learning_rate": 4.0532247964934255e-05, "loss": 106.8543, "step": 4799, "task_loss": 2.199976682662964 }, { "compression/movement_sparsity/importance_regularization_factor": 0.968901413628685, "compression/movement_sparsity/importance_threshold": -0.00020183433107029414, "compression/movement_sparsity/linear_layer_sparsity": 0.9099968596512115, "compression/movement_sparsity/model_sparsity": 0.8787356680444077, "compression_loss": 102.61404418945312, "distillation_loss": 4.629912376403809, "epoch": 4.06, "learning_rate": 4.052911709455229e-05, "loss": 107.1861, "step": 4800, "task_loss": 2.2776384353637695 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9689848623847551, "compression/movement_sparsity/importance_threshold": -0.00020129273655345847, "compression/movement_sparsity/linear_layer_sparsity": 0.9100737824566304, "compression/movement_sparsity/model_sparsity": 0.8788099483148183, "compression_loss": 102.62227630615234, "distillation_loss": 5.217840194702148, "epoch": 4.06, "learning_rate": 4.052598622417032e-05, "loss": 107.2902, "step": 4801, "task_loss": 3.3077914714813232 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9690681617250083, "compression/movement_sparsity/importance_threshold": -0.0002007521117669206, "compression/movement_sparsity/linear_layer_sparsity": 0.9100926345656627, "compression/movement_sparsity/model_sparsity": 0.8788281527959095, "compression_loss": 102.63042449951172, "distillation_loss": 4.774022102355957, "epoch": 4.06, "learning_rate": 4.052285535378835e-05, "loss": 107.2651, "step": 4802, "task_loss": 2.360692262649536 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9691513117833296, "compression/movement_sparsity/importance_threshold": -0.00020021245584174972, "compression/movement_sparsity/linear_layer_sparsity": 0.910082570568178, "compression/movement_sparsity/model_sparsity": 0.8788184345276989, "compression_loss": 102.63856506347656, "distillation_loss": 2.9765448570251465, "epoch": 4.06, "learning_rate": 4.051972448340639e-05, "loss": 106.8125, "step": 4803, "task_loss": 1.7794299125671387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9692343126936044, "compression/movement_sparsity/importance_threshold": -0.00019967376790900812, "compression/movement_sparsity/linear_layer_sparsity": 0.9101337490956711, "compression/movement_sparsity/model_sparsity": 0.878867854915329, "compression_loss": 102.64669799804688, "distillation_loss": 4.417428970336914, "epoch": 4.06, "learning_rate": 4.051659361302442e-05, "loss": 106.4494, "step": 4804, "task_loss": 1.8016687631607056 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9693171645897176, "compression/movement_sparsity/importance_threshold": -0.000199136047099765, "compression/movement_sparsity/linear_layer_sparsity": 0.9101401642978593, "compression/movement_sparsity/model_sparsity": 0.8788740497355866, "compression_loss": 102.65485382080078, "distillation_loss": 6.321595191955566, "epoch": 4.06, "learning_rate": 4.051346274264245e-05, "loss": 108.2758, "step": 4805, "task_loss": 3.4170355796813965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9693998676055544, "compression/movement_sparsity/importance_threshold": -0.00019859929254508434, "compression/movement_sparsity/linear_layer_sparsity": 0.9101687823001853, "compression/movement_sparsity/model_sparsity": 0.8789016846214934, "compression_loss": 102.66294860839844, "distillation_loss": 3.7921719551086426, "epoch": 4.06, "learning_rate": 4.051033187226049e-05, "loss": 107.0062, "step": 4806, "task_loss": 2.4778647422790527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.969482421875, "compression/movement_sparsity/importance_threshold": -0.00019806350337603362, "compression/movement_sparsity/linear_layer_sparsity": 0.9102196865718228, "compression/movement_sparsity/model_sparsity": 0.8789508401748003, "compression_loss": 102.6710433959961, "distillation_loss": 3.5522236824035645, "epoch": 4.06, "learning_rate": 4.050720100187852e-05, "loss": 106.2199, "step": 4807, "task_loss": 1.4650229215621948 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9695648275319395, "compression/movement_sparsity/importance_threshold": -0.00019752867872367772, "compression/movement_sparsity/linear_layer_sparsity": 0.9102887871232727, "compression/movement_sparsity/model_sparsity": 0.8790175669097297, "compression_loss": 102.67908477783203, "distillation_loss": 5.524815559387207, "epoch": 4.06, "learning_rate": 4.0504070131496554e-05, "loss": 107.2469, "step": 4808, "task_loss": 3.1743905544281006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9696470847102581, "compression/movement_sparsity/importance_threshold": -0.00019699481771908323, "compression/movement_sparsity/linear_layer_sparsity": 0.9103407645699974, "compression/movement_sparsity/model_sparsity": 0.8790677587712581, "compression_loss": 102.6871337890625, "distillation_loss": 5.033406734466553, "epoch": 4.07, "learning_rate": 4.050093926111459e-05, "loss": 107.4391, "step": 4809, "task_loss": 4.109838962554932 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9697291935438409, "compression/movement_sparsity/importance_threshold": -0.0001964619194933159, "compression/movement_sparsity/linear_layer_sparsity": 0.9103637066685287, "compression/movement_sparsity/model_sparsity": 0.8790899127381268, "compression_loss": 102.69512939453125, "distillation_loss": 6.075071334838867, "epoch": 4.07, "learning_rate": 4.0497808390732625e-05, "loss": 108.1443, "step": 4810, "task_loss": 2.9902663230895996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.969811154166573, "compression/movement_sparsity/importance_threshold": -0.0001959299831774423, "compression/movement_sparsity/linear_layer_sparsity": 0.9102781269174062, "compression/movement_sparsity/model_sparsity": 0.8790072729147294, "compression_loss": 102.70313262939453, "distillation_loss": 4.113101005554199, "epoch": 4.07, "learning_rate": 4.049467752035066e-05, "loss": 106.9826, "step": 4811, "task_loss": 1.7221742868423462 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9698929667123396, "compression/movement_sparsity/importance_threshold": -0.00019539900790252734, "compression/movement_sparsity/linear_layer_sparsity": 0.9102881670665556, "compression/movement_sparsity/model_sparsity": 0.8790169681538684, "compression_loss": 102.71116638183594, "distillation_loss": 5.085420608520508, "epoch": 4.07, "learning_rate": 4.0491546649968695e-05, "loss": 106.4105, "step": 4812, "task_loss": 3.3925178050994873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9699746313150257, "compression/movement_sparsity/importance_threshold": -0.00019486899279963932, "compression/movement_sparsity/linear_layer_sparsity": 0.9103655549145123, "compression/movement_sparsity/model_sparsity": 0.879091697491175, "compression_loss": 102.7191390991211, "distillation_loss": 4.421382904052734, "epoch": 4.07, "learning_rate": 4.048841577958673e-05, "loss": 107.1485, "step": 4813, "task_loss": 2.00091814994812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9700561481085167, "compression/movement_sparsity/importance_threshold": -0.00019433993699984227, "compression/movement_sparsity/linear_layer_sparsity": 0.9104759250101498, "compression/movement_sparsity/model_sparsity": 0.8791982760344892, "compression_loss": 102.7270278930664, "distillation_loss": 4.075606346130371, "epoch": 4.07, "learning_rate": 4.0485284909204765e-05, "loss": 106.7051, "step": 4814, "task_loss": 2.225390672683716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9701375172266976, "compression/movement_sparsity/importance_threshold": -0.0001938118396342019, "compression/movement_sparsity/linear_layer_sparsity": 0.9104857147517789, "compression/movement_sparsity/model_sparsity": 0.8792077294683766, "compression_loss": 102.73492431640625, "distillation_loss": 4.449033737182617, "epoch": 4.07, "learning_rate": 4.04821540388228e-05, "loss": 106.7153, "step": 4815, "task_loss": 1.8759071826934814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9702187388034534, "compression/movement_sparsity/importance_threshold": -0.0001932846998337857, "compression/movement_sparsity/linear_layer_sparsity": 0.9105362851467226, "compression/movement_sparsity/model_sparsity": 0.8792565626146812, "compression_loss": 102.74285125732422, "distillation_loss": 4.275077819824219, "epoch": 4.07, "learning_rate": 4.047902316844083e-05, "loss": 107.3785, "step": 4816, "task_loss": 1.3977067470550537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9702998129726695, "compression/movement_sparsity/importance_threshold": -0.0001927585167296594, "compression/movement_sparsity/linear_layer_sparsity": 0.9105789021218532, "compression/movement_sparsity/model_sparsity": 0.8792977155656109, "compression_loss": 102.75074768066406, "distillation_loss": 4.798806190490723, "epoch": 4.07, "learning_rate": 4.047589229805887e-05, "loss": 107.3084, "step": 4817, "task_loss": 2.32519793510437 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9703807398682309, "compression/movement_sparsity/importance_threshold": -0.00019223328945288873, "compression/movement_sparsity/linear_layer_sparsity": 0.9105936880897216, "compression/movement_sparsity/model_sparsity": 0.8793119935899961, "compression_loss": 102.75853729248047, "distillation_loss": 3.0738956928253174, "epoch": 4.07, "learning_rate": 4.04727614276769e-05, "loss": 106.9335, "step": 4818, "task_loss": 1.6451330184936523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9704615196240227, "compression/movement_sparsity/importance_threshold": -0.00019170901713453942, "compression/movement_sparsity/linear_layer_sparsity": 0.9105312173754774, "compression/movement_sparsity/model_sparsity": 0.8792516689369685, "compression_loss": 102.7663345336914, "distillation_loss": 2.340630531311035, "epoch": 4.07, "learning_rate": 4.046963055729493e-05, "loss": 106.1453, "step": 4819, "task_loss": 1.6985048055648804 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9705421523739303, "compression/movement_sparsity/importance_threshold": -0.0001911856989056781, "compression/movement_sparsity/linear_layer_sparsity": 0.9106661989531153, "compression/movement_sparsity/model_sparsity": 0.8793820134821627, "compression_loss": 102.77413177490234, "distillation_loss": 4.847789764404297, "epoch": 4.07, "learning_rate": 4.046649968691296e-05, "loss": 107.4584, "step": 4820, "task_loss": 3.3693039417266846 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9706226382518384, "compression/movement_sparsity/importance_threshold": -0.00019066333389737046, "compression/movement_sparsity/linear_layer_sparsity": 0.9107246512228663, "compression/movement_sparsity/model_sparsity": 0.8794384577366275, "compression_loss": 102.78185272216797, "distillation_loss": 4.841208457946777, "epoch": 4.08, "learning_rate": 4.0463368816531e-05, "loss": 107.5103, "step": 4821, "task_loss": 3.122145175933838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9707029773916325, "compression/movement_sparsity/importance_threshold": -0.00019014192124068314, "compression/movement_sparsity/linear_layer_sparsity": 0.9107830557959468, "compression/movement_sparsity/model_sparsity": 0.8794948559329492, "compression_loss": 102.78961181640625, "distillation_loss": 4.07402229309082, "epoch": 4.08, "learning_rate": 4.046023794614903e-05, "loss": 106.9386, "step": 4822, "task_loss": 2.94842267036438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9707831699271976, "compression/movement_sparsity/importance_threshold": -0.000189621460066681, "compression/movement_sparsity/linear_layer_sparsity": 0.9108056401694491, "compression/movement_sparsity/model_sparsity": 0.8795166644637441, "compression_loss": 102.79736328125, "distillation_loss": 4.205414295196533, "epoch": 4.08, "learning_rate": 4.0457107075767065e-05, "loss": 107.0387, "step": 4823, "task_loss": 2.3476462364196777 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9708632159924188, "compression/movement_sparsity/importance_threshold": -0.0001891019495064315, "compression/movement_sparsity/linear_layer_sparsity": 0.9109267897126294, "compression/movement_sparsity/model_sparsity": 0.8796336521474166, "compression_loss": 102.80502319335938, "distillation_loss": 4.694551944732666, "epoch": 4.08, "learning_rate": 4.04539762053851e-05, "loss": 107.2177, "step": 4824, "task_loss": 2.744784116744995 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9709431157211814, "compression/movement_sparsity/importance_threshold": -0.00018858338869099864, "compression/movement_sparsity/linear_layer_sparsity": 0.9110014469261976, "compression/movement_sparsity/model_sparsity": 0.8797057446560262, "compression_loss": 102.81270599365234, "distillation_loss": 6.07819938659668, "epoch": 4.08, "learning_rate": 4.0450845335003135e-05, "loss": 107.9315, "step": 4825, "task_loss": 3.659728527069092 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9710228692473704, "compression/movement_sparsity/importance_threshold": -0.00018806577675145163, "compression/movement_sparsity/linear_layer_sparsity": 0.9110894472833503, "compression/movement_sparsity/model_sparsity": 0.8797907219301899, "compression_loss": 102.82038879394531, "distillation_loss": 5.964413642883301, "epoch": 4.08, "learning_rate": 4.044771446462117e-05, "loss": 106.7412, "step": 4826, "task_loss": 3.0888419151306152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.971102476704871, "compression/movement_sparsity/importance_threshold": -0.00018754911281885275, "compression/movement_sparsity/linear_layer_sparsity": 0.9111129736660958, "compression/movement_sparsity/model_sparsity": 0.8798134401093125, "compression_loss": 102.82803344726562, "distillation_loss": 4.307492733001709, "epoch": 4.08, "learning_rate": 4.04445835942392e-05, "loss": 106.9804, "step": 4827, "task_loss": 1.547529935836792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9711819382275683, "compression/movement_sparsity/importance_threshold": -0.00018703339602427205, "compression/movement_sparsity/linear_layer_sparsity": 0.9111567472854871, "compression/movement_sparsity/model_sparsity": 0.8798557099702143, "compression_loss": 102.835693359375, "distillation_loss": 3.724393844604492, "epoch": 4.08, "learning_rate": 4.044145272385723e-05, "loss": 106.9198, "step": 4828, "task_loss": 2.17825984954834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9712612539493475, "compression/movement_sparsity/importance_threshold": -0.0001865186254987718, "compression/movement_sparsity/linear_layer_sparsity": 0.9112459162110681, "compression/movement_sparsity/model_sparsity": 0.8799418156688859, "compression_loss": 102.84330749511719, "distillation_loss": 3.9189727306365967, "epoch": 4.08, "learning_rate": 4.043832185347527e-05, "loss": 107.2554, "step": 4829, "task_loss": 2.4328033924102783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9713404240040936, "compression/movement_sparsity/importance_threshold": -0.00018600480037342122, "compression/movement_sparsity/linear_layer_sparsity": 0.9112788388379107, "compression/movement_sparsity/model_sparsity": 0.8799736073022146, "compression_loss": 102.85098266601562, "distillation_loss": 2.809699296951294, "epoch": 4.08, "learning_rate": 4.04351909830933e-05, "loss": 107.1374, "step": 4830, "task_loss": 2.165678024291992 }, { "compression/movement_sparsity/importance_regularization_factor": 0.971419448525692, "compression/movement_sparsity/importance_threshold": -0.00018549191977928342, "compression/movement_sparsity/linear_layer_sparsity": 0.9112219009374495, "compression/movement_sparsity/model_sparsity": 0.8799186253937957, "compression_loss": 102.85855102539062, "distillation_loss": 3.8819475173950195, "epoch": 4.08, "learning_rate": 4.043206011271133e-05, "loss": 106.7962, "step": 4831, "task_loss": 1.9968836307525635 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9714983276480276, "compression/movement_sparsity/importance_threshold": -0.00018497998284742675, "compression/movement_sparsity/linear_layer_sparsity": 0.9112301166889506, "compression/movement_sparsity/model_sparsity": 0.8799265589089581, "compression_loss": 102.86611938476562, "distillation_loss": 5.131187438964844, "epoch": 4.08, "learning_rate": 4.042892924232937e-05, "loss": 106.9649, "step": 4832, "task_loss": 2.927401065826416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9715770615049857, "compression/movement_sparsity/importance_threshold": -0.00018446898870891522, "compression/movement_sparsity/linear_layer_sparsity": 0.9112865656985387, "compression/movement_sparsity/model_sparsity": 0.8799810687214095, "compression_loss": 102.87371826171875, "distillation_loss": 4.387632846832275, "epoch": 4.09, "learning_rate": 4.04257983719474e-05, "loss": 107.3576, "step": 4833, "task_loss": 2.5972061157226562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9716556502304512, "compression/movement_sparsity/importance_threshold": -0.00018395893649481715, "compression/movement_sparsity/linear_layer_sparsity": 0.9112762393693661, "compression/movement_sparsity/model_sparsity": 0.8799710971334114, "compression_loss": 102.88128662109375, "distillation_loss": 6.25490665435791, "epoch": 4.09, "learning_rate": 4.0422667501565435e-05, "loss": 107.0349, "step": 4834, "task_loss": 2.669628381729126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9717340939583096, "compression/movement_sparsity/importance_threshold": -0.00018344982533619741, "compression/movement_sparsity/linear_layer_sparsity": 0.9112406338048054, "compression/movement_sparsity/model_sparsity": 0.879936714729529, "compression_loss": 102.88883972167969, "distillation_loss": 7.130356788635254, "epoch": 4.09, "learning_rate": 4.041953663118347e-05, "loss": 107.5807, "step": 4835, "task_loss": 3.482888698577881 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9718123928224458, "compression/movement_sparsity/importance_threshold": -0.00018294165436412088, "compression/movement_sparsity/linear_layer_sparsity": 0.9113155414258939, "compression/movement_sparsity/model_sparsity": 0.8800090490433902, "compression_loss": 102.89630889892578, "distillation_loss": 3.4635818004608154, "epoch": 4.09, "learning_rate": 4.0416405760801505e-05, "loss": 106.6454, "step": 4836, "task_loss": 3.1248273849487305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9718905469567449, "compression/movement_sparsity/importance_threshold": -0.0001824344227096559, "compression/movement_sparsity/linear_layer_sparsity": 0.9113442309732258, "compression/movement_sparsity/model_sparsity": 0.8800367530165119, "compression_loss": 102.90385437011719, "distillation_loss": 3.8987066745758057, "epoch": 4.09, "learning_rate": 4.041327489041954e-05, "loss": 107.1784, "step": 4837, "task_loss": 2.401874303817749 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9719685564950922, "compression/movement_sparsity/importance_threshold": -0.00018192812950386817, "compression/movement_sparsity/linear_layer_sparsity": 0.9113984024667955, "compression/movement_sparsity/model_sparsity": 0.8800890635526265, "compression_loss": 102.91133117675781, "distillation_loss": 4.577128887176514, "epoch": 4.09, "learning_rate": 4.041014402003757e-05, "loss": 107.6208, "step": 4838, "task_loss": 2.849470853805542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9720464215713728, "compression/movement_sparsity/importance_threshold": -0.00018142277387782173, "compression/movement_sparsity/linear_layer_sparsity": 0.9114014908262131, "compression/movement_sparsity/model_sparsity": 0.8800920458173973, "compression_loss": 102.91878509521484, "distillation_loss": 5.697328567504883, "epoch": 4.09, "learning_rate": 4.04070131496556e-05, "loss": 107.1888, "step": 4839, "task_loss": 3.5770835876464844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9721241423194718, "compression/movement_sparsity/importance_threshold": -0.00018091835496258403, "compression/movement_sparsity/linear_layer_sparsity": 0.9114490801792479, "compression/movement_sparsity/model_sparsity": 0.8801380003297533, "compression_loss": 102.92623901367188, "distillation_loss": 3.926116704940796, "epoch": 4.09, "learning_rate": 4.040388227927364e-05, "loss": 107.0386, "step": 4840, "task_loss": 1.9358716011047363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9722017188732743, "compression/movement_sparsity/importance_threshold": -0.0001804148718892208, "compression/movement_sparsity/linear_layer_sparsity": 0.9114663582981523, "compression/movement_sparsity/model_sparsity": 0.8801546848921196, "compression_loss": 102.93365478515625, "distillation_loss": 4.354589462280273, "epoch": 4.09, "learning_rate": 4.040075140889167e-05, "loss": 107.1706, "step": 4841, "task_loss": 2.7274999618530273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9722791513666655, "compression/movement_sparsity/importance_threshold": -0.00017991232378879868, "compression/movement_sparsity/linear_layer_sparsity": 0.9114908743868116, "compression/movement_sparsity/model_sparsity": 0.8801783587777131, "compression_loss": 102.94116973876953, "distillation_loss": 3.7643747329711914, "epoch": 4.09, "learning_rate": 4.039762053850971e-05, "loss": 107.2402, "step": 4842, "task_loss": 2.1260159015655518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9723564399335306, "compression/movement_sparsity/importance_threshold": -0.00017941070979238336, "compression/movement_sparsity/linear_layer_sparsity": 0.9114754206655555, "compression/movement_sparsity/model_sparsity": 0.8801634359393234, "compression_loss": 102.94855499267578, "distillation_loss": 4.310876846313477, "epoch": 4.09, "learning_rate": 4.039448966812774e-05, "loss": 107.0133, "step": 4843, "task_loss": 3.0646908283233643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9724335847077547, "compression/movement_sparsity/importance_threshold": -0.00017891002903103973, "compression/movement_sparsity/linear_layer_sparsity": 0.91152079212341, "compression/movement_sparsity/model_sparsity": 0.8802072487480217, "compression_loss": 102.95600891113281, "distillation_loss": 3.312499523162842, "epoch": 4.09, "learning_rate": 4.039135879774577e-05, "loss": 107.1736, "step": 4844, "task_loss": 1.5483516454696655 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9725105858232228, "compression/movement_sparsity/importance_threshold": -0.00017841028063583612, "compression/movement_sparsity/linear_layer_sparsity": 0.9115757863845466, "compression/movement_sparsity/model_sparsity": 0.8802603537871061, "compression_loss": 102.96341705322266, "distillation_loss": 3.2841696739196777, "epoch": 4.1, "learning_rate": 4.038822792736381e-05, "loss": 106.6025, "step": 4845, "task_loss": 1.9639689922332764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9725874434138202, "compression/movement_sparsity/importance_threshold": -0.00017791146373783654, "compression/movement_sparsity/linear_layer_sparsity": 0.9116601737189056, "compression/movement_sparsity/model_sparsity": 0.880341842156924, "compression_loss": 102.97078704833984, "distillation_loss": 3.412898540496826, "epoch": 4.1, "learning_rate": 4.0385097056981843e-05, "loss": 107.5788, "step": 4846, "task_loss": 3.5966975688934326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9726641576134321, "compression/movement_sparsity/importance_threshold": -0.00017741357746810758, "compression/movement_sparsity/linear_layer_sparsity": 0.9116787038754117, "compression/movement_sparsity/model_sparsity": 0.8803597357455487, "compression_loss": 102.97811126708984, "distillation_loss": 5.905839920043945, "epoch": 4.1, "learning_rate": 4.038196618659988e-05, "loss": 107.4909, "step": 4847, "task_loss": 3.6381311416625977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9727407285559434, "compression/movement_sparsity/importance_threshold": -0.00017691662095771672, "compression/movement_sparsity/linear_layer_sparsity": 0.9117464092992481, "compression/movement_sparsity/model_sparsity": 0.8804251152797902, "compression_loss": 102.98545837402344, "distillation_loss": 4.012759685516357, "epoch": 4.1, "learning_rate": 4.0378835316217914e-05, "loss": 106.9202, "step": 4848, "task_loss": 1.4443135261535645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9728171563752394, "compression/movement_sparsity/importance_threshold": -0.00017642059333772796, "compression/movement_sparsity/linear_layer_sparsity": 0.91180581550241, "compression/movement_sparsity/model_sparsity": 0.8804824806971185, "compression_loss": 102.99275970458984, "distillation_loss": 5.741436958312988, "epoch": 4.1, "learning_rate": 4.0375704445835946e-05, "loss": 107.7533, "step": 4849, "task_loss": 2.545442819595337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9728934412052053, "compression/movement_sparsity/importance_threshold": -0.0001759254937392079, "compression/movement_sparsity/linear_layer_sparsity": 0.9118821063269442, "compression/movement_sparsity/model_sparsity": 0.880556150697132, "compression_loss": 102.99999237060547, "distillation_loss": 4.678165435791016, "epoch": 4.1, "learning_rate": 4.037257357545398e-05, "loss": 107.4544, "step": 4850, "task_loss": 2.9499502182006836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.972969583179726, "compression/movement_sparsity/importance_threshold": -0.000175431321293224, "compression/movement_sparsity/linear_layer_sparsity": 0.9119891257314762, "compression/movement_sparsity/model_sparsity": 0.880659493655888, "compression_loss": 103.00725555419922, "distillation_loss": 3.210360050201416, "epoch": 4.1, "learning_rate": 4.0369442705072016e-05, "loss": 106.4823, "step": 4851, "task_loss": 1.3300392627716064 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9730455824326869, "compression/movement_sparsity/importance_threshold": -0.00017493807513084027, "compression/movement_sparsity/linear_layer_sparsity": 0.9120733461274882, "compression/movement_sparsity/model_sparsity": 0.8807408208222048, "compression_loss": 103.01446533203125, "distillation_loss": 4.075016498565674, "epoch": 4.1, "learning_rate": 4.036631183469005e-05, "loss": 107.184, "step": 4852, "task_loss": 2.6462628841400146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.973121439097973, "compression/movement_sparsity/importance_threshold": -0.00017444575438312418, "compression/movement_sparsity/linear_layer_sparsity": 0.9121003782155187, "compression/movement_sparsity/model_sparsity": 0.8807669242748511, "compression_loss": 103.0217056274414, "distillation_loss": 3.698078155517578, "epoch": 4.1, "learning_rate": 4.036318096430808e-05, "loss": 107.7997, "step": 4853, "task_loss": 1.7382259368896484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9731971533094697, "compression/movement_sparsity/importance_threshold": -0.00017395435818113973, "compression/movement_sparsity/linear_layer_sparsity": 0.9121767525092265, "compression/movement_sparsity/model_sparsity": 0.8808406748766151, "compression_loss": 103.02893829345703, "distillation_loss": 4.879194259643555, "epoch": 4.1, "learning_rate": 4.036005009392612e-05, "loss": 107.3584, "step": 4854, "task_loss": 2.227598190307617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9732727252010617, "compression/movement_sparsity/importance_threshold": -0.00017346388565595613, "compression/movement_sparsity/linear_layer_sparsity": 0.9121013559972648, "compression/movement_sparsity/model_sparsity": 0.8807678684667862, "compression_loss": 103.03616333007812, "distillation_loss": 3.75762939453125, "epoch": 4.1, "learning_rate": 4.035691922354415e-05, "loss": 107.0123, "step": 4855, "task_loss": 2.2545909881591797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9733481549066343, "compression/movement_sparsity/importance_threshold": -0.00017297433593863737, "compression/movement_sparsity/linear_layer_sparsity": 0.9121241073091141, "compression/movement_sparsity/model_sparsity": 0.8807898382010823, "compression_loss": 103.04337310791016, "distillation_loss": 6.587848663330078, "epoch": 4.1, "learning_rate": 4.035378835316218e-05, "loss": 107.4795, "step": 4856, "task_loss": 3.6715574264526367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9734234425600728, "compression/movement_sparsity/importance_threshold": -0.0001724857081602492, "compression/movement_sparsity/linear_layer_sparsity": 0.9121405030396135, "compression/movement_sparsity/model_sparsity": 0.8808056706877997, "compression_loss": 103.05056762695312, "distillation_loss": 4.162094593048096, "epoch": 4.11, "learning_rate": 4.035065748278021e-05, "loss": 107.2129, "step": 4857, "task_loss": 2.4532887935638428 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9734985882952623, "compression/movement_sparsity/importance_threshold": -0.00017199800145185906, "compression/movement_sparsity/linear_layer_sparsity": 0.9121252758775424, "compression/movement_sparsity/model_sparsity": 0.8807909666255901, "compression_loss": 103.05776977539062, "distillation_loss": 6.454315185546875, "epoch": 4.11, "learning_rate": 4.034752661239825e-05, "loss": 108.4352, "step": 4858, "task_loss": 4.206214427947998 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9735735922460879, "compression/movement_sparsity/importance_threshold": -0.00017151121494453185, "compression/movement_sparsity/linear_layer_sparsity": 0.9121405865087868, "compression/movement_sparsity/model_sparsity": 0.8808057512895503, "compression_loss": 103.06497955322266, "distillation_loss": 3.932981014251709, "epoch": 4.11, "learning_rate": 4.0344395742016284e-05, "loss": 107.1145, "step": 4859, "task_loss": 2.399012327194214 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9736484545464347, "compression/movement_sparsity/importance_threshold": -0.00017102534776933416, "compression/movement_sparsity/linear_layer_sparsity": 0.9122309478511315, "compression/movement_sparsity/model_sparsity": 0.8808930084418013, "compression_loss": 103.07212829589844, "distillation_loss": 4.696528434753418, "epoch": 4.11, "learning_rate": 4.0341264871634316e-05, "loss": 107.2135, "step": 4860, "task_loss": 2.447951555252075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9737231753301877, "compression/movement_sparsity/importance_threshold": -0.0001705403990573326, "compression/movement_sparsity/linear_layer_sparsity": 0.9122262378049153, "compression/movement_sparsity/model_sparsity": 0.8808884602001624, "compression_loss": 103.07929229736328, "distillation_loss": 3.843967914581299, "epoch": 4.11, "learning_rate": 4.033813400125235e-05, "loss": 106.5539, "step": 4861, "task_loss": 2.963442802429199 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9737977547312324, "compression/movement_sparsity/importance_threshold": -0.00017005636793959115, "compression/movement_sparsity/linear_layer_sparsity": 0.9122534964521308, "compression/movement_sparsity/model_sparsity": 0.8809147824289888, "compression_loss": 103.0864028930664, "distillation_loss": 4.53239107131958, "epoch": 4.11, "learning_rate": 4.0335003130870386e-05, "loss": 107.1592, "step": 4862, "task_loss": 3.869985818862915 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9738721928834537, "compression/movement_sparsity/importance_threshold": -0.0001695732535471773, "compression/movement_sparsity/linear_layer_sparsity": 0.9122924527077972, "compression/movement_sparsity/model_sparsity": 0.8809524004174295, "compression_loss": 103.09354400634766, "distillation_loss": 3.9507226943969727, "epoch": 4.11, "learning_rate": 4.033187226048842e-05, "loss": 106.9912, "step": 4863, "task_loss": 4.000669479370117 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9739464899207367, "compression/movement_sparsity/importance_threshold": -0.00016909105501115764, "compression/movement_sparsity/linear_layer_sparsity": 0.9123547684078622, "compression/movement_sparsity/model_sparsity": 0.8810125753814918, "compression_loss": 103.1005859375, "distillation_loss": 4.788640975952148, "epoch": 4.11, "learning_rate": 4.032874139010645e-05, "loss": 107.1196, "step": 4864, "task_loss": 3.468475818634033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9740206459769667, "compression/movement_sparsity/importance_threshold": -0.00016860977146259792, "compression/movement_sparsity/linear_layer_sparsity": 0.9124051957127943, "compression/movement_sparsity/model_sparsity": 0.8810612703533669, "compression_loss": 103.10768127441406, "distillation_loss": 4.603659629821777, "epoch": 4.11, "learning_rate": 4.032561051972448e-05, "loss": 107.6727, "step": 4865, "task_loss": 2.7617413997650146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9740946611860288, "compression/movement_sparsity/importance_threshold": -0.000168129402032563, "compression/movement_sparsity/linear_layer_sparsity": 0.9123917214033658, "compression/movement_sparsity/model_sparsity": 0.8810482589279192, "compression_loss": 103.11469268798828, "distillation_loss": 4.272139549255371, "epoch": 4.11, "learning_rate": 4.032247964934252e-05, "loss": 107.0731, "step": 4866, "task_loss": 1.6391749382019043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9741685356818082, "compression/movement_sparsity/importance_threshold": -0.00016764994585211947, "compression/movement_sparsity/linear_layer_sparsity": 0.9124452132193802, "compression/movement_sparsity/model_sparsity": 0.8810999131354934, "compression_loss": 103.12171173095703, "distillation_loss": 3.5688412189483643, "epoch": 4.11, "learning_rate": 4.031934877896055e-05, "loss": 107.2389, "step": 4867, "task_loss": 2.144543170928955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9742422695981897, "compression/movement_sparsity/importance_threshold": -0.00016717140205233481, "compression/movement_sparsity/linear_layer_sparsity": 0.9124518311324181, "compression/movement_sparsity/model_sparsity": 0.8811063037028594, "compression_loss": 103.12876892089844, "distillation_loss": 5.70465087890625, "epoch": 4.11, "learning_rate": 4.031621790857858e-05, "loss": 108.1124, "step": 4868, "task_loss": 3.167325019836426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9743158630690588, "compression/movement_sparsity/importance_threshold": -0.0001666937697642739, "compression/movement_sparsity/linear_layer_sparsity": 0.9124678929862237, "compression/movement_sparsity/model_sparsity": 0.8811218137825746, "compression_loss": 103.1358413696289, "distillation_loss": 5.836475372314453, "epoch": 4.12, "learning_rate": 4.031308703819662e-05, "loss": 107.9487, "step": 4869, "task_loss": 2.0082461833953857 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9743893162283006, "compression/movement_sparsity/importance_threshold": -0.0001662170481190016, "compression/movement_sparsity/linear_layer_sparsity": 0.9124865066119032, "compression/movement_sparsity/model_sparsity": 0.8811397879729499, "compression_loss": 103.14276885986328, "distillation_loss": 4.967957973480225, "epoch": 4.12, "learning_rate": 4.0309956167814654e-05, "loss": 107.0386, "step": 4870, "task_loss": 2.947622537612915 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9744626292098001, "compression/movement_sparsity/importance_threshold": -0.00016574123624758623, "compression/movement_sparsity/linear_layer_sparsity": 0.9124477172945838, "compression/movement_sparsity/model_sparsity": 0.8811023311880103, "compression_loss": 103.1498794555664, "distillation_loss": 4.1553263664245605, "epoch": 4.12, "learning_rate": 4.0306825297432685e-05, "loss": 107.2634, "step": 4871, "task_loss": 2.7684144973754883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9745358021474425, "compression/movement_sparsity/importance_threshold": -0.0001652663332810918, "compression/movement_sparsity/linear_layer_sparsity": 0.9124401931448055, "compression/movement_sparsity/model_sparsity": 0.8810950655159239, "compression_loss": 103.1568374633789, "distillation_loss": 7.0473785400390625, "epoch": 4.12, "learning_rate": 4.030369442705072e-05, "loss": 108.8826, "step": 4872, "task_loss": 3.4608702659606934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.974608835175113, "compression/movement_sparsity/importance_threshold": -0.0001647923383505858, "compression/movement_sparsity/linear_layer_sparsity": 0.9124589140879938, "compression/movement_sparsity/model_sparsity": 0.8811131433371213, "compression_loss": 103.163818359375, "distillation_loss": 5.356698989868164, "epoch": 4.12, "learning_rate": 4.0300563556668756e-05, "loss": 107.5184, "step": 4873, "task_loss": 2.6004228591918945 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9746817284266966, "compression/movement_sparsity/importance_threshold": -0.00016431925058713395, "compression/movement_sparsity/linear_layer_sparsity": 0.9125347398699903, "compression/movement_sparsity/model_sparsity": 0.8811863642702389, "compression_loss": 103.17076873779297, "distillation_loss": 3.710014820098877, "epoch": 4.12, "learning_rate": 4.029743268628679e-05, "loss": 107.4233, "step": 4874, "task_loss": 2.713430404663086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9747544820360787, "compression/movement_sparsity/importance_threshold": -0.00016384706912180112, "compression/movement_sparsity/linear_layer_sparsity": 0.9125774880109649, "compression/movement_sparsity/model_sparsity": 0.8812276438810622, "compression_loss": 103.177734375, "distillation_loss": 4.193863868713379, "epoch": 4.12, "learning_rate": 4.029430181590482e-05, "loss": 107.3004, "step": 4875, "task_loss": 1.6833380460739136 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9748270961371442, "compression/movement_sparsity/importance_threshold": -0.00016337579308565477, "compression/movement_sparsity/linear_layer_sparsity": 0.9126112572537096, "compression/movement_sparsity/model_sparsity": 0.8812602530464324, "compression_loss": 103.1847152709961, "distillation_loss": 3.6271731853485107, "epoch": 4.12, "learning_rate": 4.029117094552286e-05, "loss": 107.4487, "step": 4876, "task_loss": 0.8073420524597168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9748995708637783, "compression/movement_sparsity/importance_threshold": -0.00016290542160976064, "compression/movement_sparsity/linear_layer_sparsity": 0.912609802505258, "compression/movement_sparsity/model_sparsity": 0.8812588482730654, "compression_loss": 103.19166564941406, "distillation_loss": 4.907578468322754, "epoch": 4.12, "learning_rate": 4.028804007514089e-05, "loss": 107.5721, "step": 4877, "task_loss": 3.2344648838043213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9749719063498662, "compression/movement_sparsity/importance_threshold": -0.00016243595382518447, "compression/movement_sparsity/linear_layer_sparsity": 0.912605056686539, "compression/movement_sparsity/model_sparsity": 0.8812542654878193, "compression_loss": 103.1985092163086, "distillation_loss": 5.9523773193359375, "epoch": 4.12, "learning_rate": 4.028490920475893e-05, "loss": 107.4528, "step": 4878, "task_loss": 4.266650199890137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.975044102729293, "compression/movement_sparsity/importance_threshold": -0.00016196738886299286, "compression/movement_sparsity/linear_layer_sparsity": 0.9126566048632289, "compression/movement_sparsity/model_sparsity": 0.881304042826059, "compression_loss": 103.20536804199219, "distillation_loss": 4.504624366760254, "epoch": 4.12, "learning_rate": 4.028177833437696e-05, "loss": 106.7959, "step": 4879, "task_loss": 1.8314474821090698 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9751161601359437, "compression/movement_sparsity/importance_threshold": -0.00016149972585425155, "compression/movement_sparsity/linear_layer_sparsity": 0.9126723447645082, "compression/movement_sparsity/model_sparsity": 0.8813192420133078, "compression_loss": 103.21216583251953, "distillation_loss": 5.232707500457764, "epoch": 4.13, "learning_rate": 4.027864746399499e-05, "loss": 107.3584, "step": 4880, "task_loss": 2.2642223834991455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9751880787037037, "compression/movement_sparsity/importance_threshold": -0.0001610329639300254, "compression/movement_sparsity/linear_layer_sparsity": 0.9126024452938267, "compression/movement_sparsity/model_sparsity": 0.8812517438044802, "compression_loss": 103.21903991699219, "distillation_loss": 5.393105983734131, "epoch": 4.13, "learning_rate": 4.027551659361303e-05, "loss": 108.4193, "step": 4881, "task_loss": 2.069713830947876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.975259858566458, "compression/movement_sparsity/importance_threshold": -0.00016056710222138188, "compression/movement_sparsity/linear_layer_sparsity": 0.9126969443223409, "compression/movement_sparsity/model_sparsity": 0.881342996500652, "compression_loss": 103.2258529663086, "distillation_loss": 4.903548240661621, "epoch": 4.13, "learning_rate": 4.027238572323106e-05, "loss": 107.5907, "step": 4882, "task_loss": 3.4813477993011475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9753314998580918, "compression/movement_sparsity/importance_threshold": -0.0001601021398593876, "compression/movement_sparsity/linear_layer_sparsity": 0.9127048381213159, "compression/movement_sparsity/model_sparsity": 0.881350619123348, "compression_loss": 103.23270416259766, "distillation_loss": 4.186715602874756, "epoch": 4.13, "learning_rate": 4.0269254852849094e-05, "loss": 107.0065, "step": 4883, "task_loss": 3.0346081256866455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9754030027124901, "compression/movement_sparsity/importance_threshold": -0.00015963807597510656, "compression/movement_sparsity/linear_layer_sparsity": 0.9127034072211996, "compression/movement_sparsity/model_sparsity": 0.8813492373790526, "compression_loss": 103.23948669433594, "distillation_loss": 5.621318340301514, "epoch": 4.13, "learning_rate": 4.026612398246713e-05, "loss": 107.0063, "step": 4884, "task_loss": 3.539299488067627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9754743672635382, "compression/movement_sparsity/importance_threshold": -0.00015917490969960622, "compression/movement_sparsity/linear_layer_sparsity": 0.9127324544935606, "compression/movement_sparsity/model_sparsity": 0.8813772867882481, "compression_loss": 103.24628448486328, "distillation_loss": 4.240365982055664, "epoch": 4.13, "learning_rate": 4.0262993112085164e-05, "loss": 107.7461, "step": 4885, "task_loss": 1.5251898765563965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9755455936451212, "compression/movement_sparsity/importance_threshold": -0.00015871264016395319, "compression/movement_sparsity/linear_layer_sparsity": 0.9127830129643366, "compression/movement_sparsity/model_sparsity": 0.881426108420017, "compression_loss": 103.2530288696289, "distillation_loss": 4.662071228027344, "epoch": 4.13, "learning_rate": 4.0259862241703196e-05, "loss": 107.8412, "step": 4886, "task_loss": 2.657071828842163 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9756166819911242, "compression/movement_sparsity/importance_threshold": -0.00015825126649921147, "compression/movement_sparsity/linear_layer_sparsity": 0.9128061577737179, "compression/movement_sparsity/model_sparsity": 0.8814484581339942, "compression_loss": 103.25981140136719, "distillation_loss": 4.64803409576416, "epoch": 4.13, "learning_rate": 4.025673137132123e-05, "loss": 107.5654, "step": 4887, "task_loss": 2.0577316284179688 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9756876324354323, "compression/movement_sparsity/importance_threshold": -0.00015779078783644852, "compression/movement_sparsity/linear_layer_sparsity": 0.912921846048121, "compression/movement_sparsity/model_sparsity": 0.8815601721602728, "compression_loss": 103.26654052734375, "distillation_loss": 4.233486175537109, "epoch": 4.13, "learning_rate": 4.0253600500939266e-05, "loss": 107.4082, "step": 4888, "task_loss": 2.500523805618286 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9757584451119308, "compression/movement_sparsity/importance_threshold": -0.0001573312033067301, "compression/movement_sparsity/linear_layer_sparsity": 0.9128648723751568, "compression/movement_sparsity/model_sparsity": 0.8815051557082465, "compression_loss": 103.27323150634766, "distillation_loss": 4.681158065795898, "epoch": 4.13, "learning_rate": 4.02504696305573e-05, "loss": 107.8147, "step": 4889, "task_loss": 1.660998821258545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9758291201545047, "compression/movement_sparsity/importance_threshold": -0.00015687251204112192, "compression/movement_sparsity/linear_layer_sparsity": 0.9129349507083528, "compression/movement_sparsity/model_sparsity": 0.881572826635111, "compression_loss": 103.2799301147461, "distillation_loss": 4.947524070739746, "epoch": 4.13, "learning_rate": 4.024733876017533e-05, "loss": 108.0737, "step": 4890, "task_loss": 2.839801073074341 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9758996576970391, "compression/movement_sparsity/importance_threshold": -0.00015641471317069233, "compression/movement_sparsity/linear_layer_sparsity": 0.912994869650723, "compression/movement_sparsity/model_sparsity": 0.8816306871774786, "compression_loss": 103.2865982055664, "distillation_loss": 4.058843612670898, "epoch": 4.13, "learning_rate": 4.024420788979337e-05, "loss": 107.743, "step": 4891, "task_loss": 2.023789644241333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9759700578734194, "compression/movement_sparsity/importance_threshold": -0.00015595780582650359, "compression/movement_sparsity/linear_layer_sparsity": 0.9130711843235926, "compression/movement_sparsity/model_sparsity": 0.8817043802065636, "compression_loss": 103.29328918457031, "distillation_loss": 2.775730609893799, "epoch": 4.14, "learning_rate": 4.02410770194114e-05, "loss": 107.6248, "step": 4892, "task_loss": 1.7594448328018188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9760403208175304, "compression/movement_sparsity/importance_threshold": -0.00015550178913962404, "compression/movement_sparsity/linear_layer_sparsity": 0.9130622769703686, "compression/movement_sparsity/model_sparsity": 0.8816957788483251, "compression_loss": 103.29998016357422, "distillation_loss": 5.059879779815674, "epoch": 4.14, "learning_rate": 4.023794614902943e-05, "loss": 107.9662, "step": 4893, "task_loss": 3.604630947113037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9761104466632575, "compression/movement_sparsity/importance_threshold": -0.00015504666224111854, "compression/movement_sparsity/linear_layer_sparsity": 0.9130114561679045, "compression/movement_sparsity/model_sparsity": 0.8816467038967688, "compression_loss": 103.30664825439453, "distillation_loss": 3.983198642730713, "epoch": 4.14, "learning_rate": 4.0234815278647464e-05, "loss": 106.9615, "step": 4894, "task_loss": 2.8911352157592773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9761804355444857, "compression/movement_sparsity/importance_threshold": -0.00015459242426205458, "compression/movement_sparsity/linear_layer_sparsity": 0.9131513743509437, "compression/movement_sparsity/model_sparsity": 0.881781815459782, "compression_loss": 103.31327056884766, "distillation_loss": 3.8336145877838135, "epoch": 4.14, "learning_rate": 4.02316844082655e-05, "loss": 107.4055, "step": 4895, "task_loss": 2.589608907699585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9762502875951002, "compression/movement_sparsity/importance_threshold": -0.00015413907433349788, "compression/movement_sparsity/linear_layer_sparsity": 0.913147630162306, "compression/movement_sparsity/model_sparsity": 0.8817781998955424, "compression_loss": 103.31987762451172, "distillation_loss": 4.618251800537109, "epoch": 4.14, "learning_rate": 4.0228553537883534e-05, "loss": 107.215, "step": 4896, "task_loss": 2.2637813091278076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9763200029489862, "compression/movement_sparsity/importance_threshold": -0.00015368661158651244, "compression/movement_sparsity/linear_layer_sparsity": 0.9131793842207203, "compression/movement_sparsity/model_sparsity": 0.8818088631043633, "compression_loss": 103.32657623291016, "distillation_loss": 4.347321510314941, "epoch": 4.14, "learning_rate": 4.0225422667501566e-05, "loss": 107.9603, "step": 4897, "task_loss": 3.240821599960327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9763895817400288, "compression/movement_sparsity/importance_threshold": -0.0001532350351521666, "compression/movement_sparsity/linear_layer_sparsity": 0.9131645624803489, "compression/movement_sparsity/model_sparsity": 0.8817945505363707, "compression_loss": 103.33316802978516, "distillation_loss": 5.428161144256592, "epoch": 4.14, "learning_rate": 4.02222917971196e-05, "loss": 107.3597, "step": 4898, "task_loss": 3.2392184734344482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9764590241021129, "compression/movement_sparsity/importance_threshold": -0.00015278434416152696, "compression/movement_sparsity/linear_layer_sparsity": 0.9131683185931543, "compression/movement_sparsity/model_sparsity": 0.881798177615146, "compression_loss": 103.33976745605469, "distillation_loss": 5.415704250335693, "epoch": 4.14, "learning_rate": 4.0219160926737636e-05, "loss": 108.4933, "step": 4899, "task_loss": 2.4747369289398193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9765283301691241, "compression/movement_sparsity/importance_threshold": -0.00015233453774565666, "compression/movement_sparsity/linear_layer_sparsity": 0.9131514220476142, "compression/movement_sparsity/model_sparsity": 0.8817818615179251, "compression_loss": 103.34635925292969, "distillation_loss": 3.890885353088379, "epoch": 4.14, "learning_rate": 4.021603005635567e-05, "loss": 107.2023, "step": 4900, "task_loss": 2.1159677505493164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9765975000749472, "compression/movement_sparsity/importance_threshold": -0.00015188561503562403, "compression/movement_sparsity/linear_layer_sparsity": 0.913218233158878, "compression/movement_sparsity/model_sparsity": 0.8818463774619819, "compression_loss": 103.35294342041016, "distillation_loss": 4.183610916137695, "epoch": 4.14, "learning_rate": 4.02128991859737e-05, "loss": 107.2628, "step": 4901, "task_loss": 2.8122332096099854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9766665339534673, "compression/movement_sparsity/importance_threshold": -0.00015143757516249481, "compression/movement_sparsity/linear_layer_sparsity": 0.9133630998714861, "compression/movement_sparsity/model_sparsity": 0.8819862675573498, "compression_loss": 103.35945892333984, "distillation_loss": 4.3213911056518555, "epoch": 4.14, "learning_rate": 4.020976831559173e-05, "loss": 107.9445, "step": 4902, "task_loss": 2.326215982437134 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9767354319385698, "compression/movement_sparsity/importance_threshold": -0.00015099041725733387, "compression/movement_sparsity/linear_layer_sparsity": 0.9134736130571353, "compression/movement_sparsity/model_sparsity": 0.8820929842750936, "compression_loss": 103.36603546142578, "distillation_loss": 5.476161003112793, "epoch": 4.14, "learning_rate": 4.020663744520977e-05, "loss": 108.3903, "step": 4903, "task_loss": 2.8430793285369873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9768041941641397, "compression/movement_sparsity/importance_threshold": -0.00015054414045120867, "compression/movement_sparsity/linear_layer_sparsity": 0.9135750281028783, "compression/movement_sparsity/model_sparsity": 0.8821909154020262, "compression_loss": 103.37256622314453, "distillation_loss": 4.340805530548096, "epoch": 4.15, "learning_rate": 4.02035065748278e-05, "loss": 107.525, "step": 4904, "task_loss": 1.5062768459320068 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9768728207640621, "compression/movement_sparsity/importance_threshold": -0.00015009874387518496, "compression/movement_sparsity/linear_layer_sparsity": 0.9135009074768538, "compression/movement_sparsity/model_sparsity": 0.8821193410475273, "compression_loss": 103.37908172607422, "distillation_loss": 2.89214825630188, "epoch": 4.15, "learning_rate": 4.0200375704445834e-05, "loss": 106.8337, "step": 4905, "task_loss": 1.9961196184158325 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9769413118722222, "compression/movement_sparsity/importance_threshold": -0.00014965422666032846, "compression/movement_sparsity/linear_layer_sparsity": 0.9135012413535476, "compression/movement_sparsity/model_sparsity": 0.8821196634545295, "compression_loss": 103.3855972290039, "distillation_loss": 6.4654388427734375, "epoch": 4.15, "learning_rate": 4.019724483406387e-05, "loss": 108.0938, "step": 4906, "task_loss": 3.5064024925231934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9770096676225052, "compression/movement_sparsity/importance_threshold": -0.0001492105879377049, "compression/movement_sparsity/linear_layer_sparsity": 0.9135102202517774, "compression/movement_sparsity/model_sparsity": 0.8821283338999828, "compression_loss": 103.39210510253906, "distillation_loss": 3.7171525955200195, "epoch": 4.15, "learning_rate": 4.0194113963681904e-05, "loss": 106.7692, "step": 4907, "task_loss": 2.496166467666626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9770778881487961, "compression/movement_sparsity/importance_threshold": -0.00014876782683838005, "compression/movement_sparsity/linear_layer_sparsity": 0.9135678855264644, "compression/movement_sparsity/model_sparsity": 0.8821840181950852, "compression_loss": 103.39859008789062, "distillation_loss": 5.934309959411621, "epoch": 4.15, "learning_rate": 4.0190983093299936e-05, "loss": 108.282, "step": 4908, "task_loss": 3.895803451538086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9771459735849801, "compression/movement_sparsity/importance_threshold": -0.00014832594249342135, "compression/movement_sparsity/linear_layer_sparsity": 0.9135796666040886, "compression/movement_sparsity/model_sparsity": 0.8821953945564502, "compression_loss": 103.40498352050781, "distillation_loss": 4.94340705871582, "epoch": 4.15, "learning_rate": 4.018785222291797e-05, "loss": 107.4563, "step": 4909, "task_loss": 3.5095739364624023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9772139240649425, "compression/movement_sparsity/importance_threshold": -0.00014788493403389367, "compression/movement_sparsity/linear_layer_sparsity": 0.9136638154550949, "compression/movement_sparsity/model_sparsity": 0.8822766526355523, "compression_loss": 103.41143798828125, "distillation_loss": 5.9878997802734375, "epoch": 4.15, "learning_rate": 4.0184721352536006e-05, "loss": 108.6505, "step": 4910, "task_loss": 3.4021270275115967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9772817397225682, "compression/movement_sparsity/importance_threshold": -0.00014744480059086362, "compression/movement_sparsity/linear_layer_sparsity": 0.9136411833849221, "compression/movement_sparsity/model_sparsity": 0.8822547980466142, "compression_loss": 103.4178695678711, "distillation_loss": 4.713789939880371, "epoch": 4.15, "learning_rate": 4.018159048215404e-05, "loss": 108.6078, "step": 4911, "task_loss": 2.9544825553894043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9773494206917425, "compression/movement_sparsity/importance_threshold": -0.00014700554129539693, "compression/movement_sparsity/linear_layer_sparsity": 0.9135767809555208, "compression/movement_sparsity/model_sparsity": 0.882192608038788, "compression_loss": 103.4243392944336, "distillation_loss": 4.253330230712891, "epoch": 4.15, "learning_rate": 4.0178459611772077e-05, "loss": 107.5391, "step": 4912, "task_loss": 2.2982587814331055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9774169671063503, "compression/movement_sparsity/importance_threshold": -0.0001465671552785602, "compression/movement_sparsity/linear_layer_sparsity": 0.9136016786175445, "compression/movement_sparsity/model_sparsity": 0.882216650389527, "compression_loss": 103.43067932128906, "distillation_loss": 4.56576681137085, "epoch": 4.15, "learning_rate": 4.017532874139011e-05, "loss": 107.712, "step": 4913, "task_loss": 2.458975315093994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9774843791002771, "compression/movement_sparsity/importance_threshold": -0.0001461296416714183, "compression/movement_sparsity/linear_layer_sparsity": 0.9136189925089517, "compression/movement_sparsity/model_sparsity": 0.8822333694955007, "compression_loss": 103.43701934814453, "distillation_loss": 4.685150146484375, "epoch": 4.15, "learning_rate": 4.017219787100815e-05, "loss": 108.0534, "step": 4914, "task_loss": 1.9182080030441284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9775516568074077, "compression/movement_sparsity/importance_threshold": -0.00014569299960503872, "compression/movement_sparsity/linear_layer_sparsity": 0.9136454880094386, "compression/movement_sparsity/model_sparsity": 0.8822589547940362, "compression_loss": 103.44336700439453, "distillation_loss": 3.2142882347106934, "epoch": 4.15, "learning_rate": 4.016906700062618e-05, "loss": 107.7919, "step": 4915, "task_loss": 1.371606469154358 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9776188003616275, "compression/movement_sparsity/importance_threshold": -0.0001452572282104863, "compression/movement_sparsity/linear_layer_sparsity": 0.9136393470631061, "compression/movement_sparsity/model_sparsity": 0.8822530248081019, "compression_loss": 103.44965362548828, "distillation_loss": 6.138660430908203, "epoch": 4.16, "learning_rate": 4.016593613024421e-05, "loss": 107.9043, "step": 4916, "task_loss": 4.365577697753906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9776858098968216, "compression/movement_sparsity/importance_threshold": -0.00014482232661882678, "compression/movement_sparsity/linear_layer_sparsity": 0.9137038806583514, "compression/movement_sparsity/model_sparsity": 0.8823153414758219, "compression_loss": 103.45591735839844, "distillation_loss": 3.7030177116394043, "epoch": 4.16, "learning_rate": 4.016280525986225e-05, "loss": 107.0678, "step": 4917, "task_loss": 1.872741460800171 }, { "compression/movement_sparsity/importance_regularization_factor": 0.977752685546875, "compression/movement_sparsity/importance_threshold": -0.0001443882939611285, "compression/movement_sparsity/linear_layer_sparsity": 0.9137806961462617, "compression/movement_sparsity/model_sparsity": 0.8823895181154104, "compression_loss": 103.46223449707031, "distillation_loss": 4.592808723449707, "epoch": 4.16, "learning_rate": 4.015967438948028e-05, "loss": 107.8398, "step": 4918, "task_loss": 3.196143865585327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.977819427445673, "compression/movement_sparsity/importance_threshold": -0.00014395512936845548, "compression/movement_sparsity/linear_layer_sparsity": 0.9138244220689824, "compression/movement_sparsity/model_sparsity": 0.882431741918169, "compression_loss": 103.4684829711914, "distillation_loss": 3.4923319816589355, "epoch": 4.16, "learning_rate": 4.015654351909831e-05, "loss": 107.8279, "step": 4919, "task_loss": 1.210857629776001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9778860357271005, "compression/movement_sparsity/importance_threshold": -0.0001435228319718743, "compression/movement_sparsity/linear_layer_sparsity": 0.9138566769424373, "compression/movement_sparsity/model_sparsity": 0.8824628887374932, "compression_loss": 103.47472381591797, "distillation_loss": 4.646156311035156, "epoch": 4.16, "learning_rate": 4.0153412648716344e-05, "loss": 107.8, "step": 4920, "task_loss": 2.693532943725586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.977952510525043, "compression/movement_sparsity/importance_threshold": -0.00014309140090244983, "compression/movement_sparsity/linear_layer_sparsity": 0.9138098268877961, "compression/movement_sparsity/model_sparsity": 0.8824176481263565, "compression_loss": 103.48106384277344, "distillation_loss": 3.6667351722717285, "epoch": 4.16, "learning_rate": 4.015028177833438e-05, "loss": 107.391, "step": 4921, "task_loss": 3.3406569957733154 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9780188519733853, "compression/movement_sparsity/importance_threshold": -0.0001426608352912504, "compression/movement_sparsity/linear_layer_sparsity": 0.9138429403013208, "compression/movement_sparsity/model_sparsity": 0.8824496239922579, "compression_loss": 103.48724365234375, "distillation_loss": 5.576323509216309, "epoch": 4.16, "learning_rate": 4.0147150907952415e-05, "loss": 108.3693, "step": 4922, "task_loss": 2.9830191135406494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9780850602060127, "compression/movement_sparsity/importance_threshold": -0.0001422311342693409, "compression/movement_sparsity/linear_layer_sparsity": 0.9138849729922373, "compression/movement_sparsity/model_sparsity": 0.8824902127309336, "compression_loss": 103.49345397949219, "distillation_loss": 4.727092266082764, "epoch": 4.16, "learning_rate": 4.0144020037570446e-05, "loss": 107.7683, "step": 4923, "task_loss": 3.176384925842285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9781511353568103, "compression/movement_sparsity/importance_threshold": -0.00014180229696778707, "compression/movement_sparsity/linear_layer_sparsity": 0.9139491607866211, "compression/movement_sparsity/model_sparsity": 0.8825521954771157, "compression_loss": 103.4996566772461, "distillation_loss": 5.450467109680176, "epoch": 4.16, "learning_rate": 4.014088916718848e-05, "loss": 108.5112, "step": 4924, "task_loss": 3.0870888233184814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9782170775596634, "compression/movement_sparsity/importance_threshold": -0.00014137432251765548, "compression/movement_sparsity/linear_layer_sparsity": 0.913961836176818, "compression/movement_sparsity/model_sparsity": 0.8825644354286653, "compression_loss": 103.5058822631836, "distillation_loss": 4.01026725769043, "epoch": 4.16, "learning_rate": 4.013775829680652e-05, "loss": 107.9067, "step": 4925, "task_loss": 1.8630263805389404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9782828869484569, "compression/movement_sparsity/importance_threshold": -0.00014094721005001188, "compression/movement_sparsity/linear_layer_sparsity": 0.9140024498917858, "compression/movement_sparsity/model_sparsity": 0.8826036539375814, "compression_loss": 103.51206970214844, "distillation_loss": 4.55216121673584, "epoch": 4.16, "learning_rate": 4.013462742642455e-05, "loss": 107.9055, "step": 4926, "task_loss": 2.4766299724578857 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9783485636570761, "compression/movement_sparsity/importance_threshold": -0.00014052095869592114, "compression/movement_sparsity/linear_layer_sparsity": 0.9140690344438644, "compression/movement_sparsity/model_sparsity": 0.8826679511054581, "compression_loss": 103.51821899414062, "distillation_loss": 5.261106014251709, "epoch": 4.16, "learning_rate": 4.013149655604258e-05, "loss": 107.7647, "step": 4927, "task_loss": 2.000267505645752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9784141078194061, "compression/movement_sparsity/importance_threshold": -0.0001400955675864516, "compression/movement_sparsity/linear_layer_sparsity": 0.9141823736572433, "compression/movement_sparsity/model_sparsity": 0.8827773967681853, "compression_loss": 103.52444458007812, "distillation_loss": 4.078245162963867, "epoch": 4.17, "learning_rate": 4.012836568566062e-05, "loss": 107.4863, "step": 4928, "task_loss": 1.7417700290679932 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9784795195693319, "compression/movement_sparsity/importance_threshold": -0.000139671035852669, "compression/movement_sparsity/linear_layer_sparsity": 0.9143164609223086, "compression/movement_sparsity/model_sparsity": 0.8829068777231949, "compression_loss": 103.53064727783203, "distillation_loss": 3.4105663299560547, "epoch": 4.17, "learning_rate": 4.012523481527865e-05, "loss": 108.0769, "step": 4929, "task_loss": 2.265688896179199 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9785447990407389, "compression/movement_sparsity/importance_threshold": -0.0001392473626256373, "compression/movement_sparsity/linear_layer_sparsity": 0.9143264652989551, "compression/movement_sparsity/model_sparsity": 0.8829165384187265, "compression_loss": 103.53678894042969, "distillation_loss": 3.9625284671783447, "epoch": 4.17, "learning_rate": 4.012210394489668e-05, "loss": 108.4062, "step": 4930, "task_loss": 2.1866326332092285 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9786099463675121, "compression/movement_sparsity/importance_threshold": -0.00013882454703642404, "compression/movement_sparsity/linear_layer_sparsity": 0.9144020406734312, "compression/movement_sparsity/model_sparsity": 0.8829895175465924, "compression_loss": 103.54297637939453, "distillation_loss": 3.600799560546875, "epoch": 4.17, "learning_rate": 4.0118973074514714e-05, "loss": 107.8113, "step": 4931, "task_loss": 2.8793835639953613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9786749616835365, "compression/movement_sparsity/importance_threshold": -0.00013840258821609577, "compression/movement_sparsity/linear_layer_sparsity": 0.9144129274384828, "compression/movement_sparsity/model_sparsity": 0.8830000303177727, "compression_loss": 103.54908752441406, "distillation_loss": 4.692935466766357, "epoch": 4.17, "learning_rate": 4.011584220413275e-05, "loss": 108.0536, "step": 4932, "task_loss": 3.3289122581481934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9787398451226975, "compression/movement_sparsity/importance_threshold": -0.00013798148529571737, "compression/movement_sparsity/linear_layer_sparsity": 0.9144099106240708, "compression/movement_sparsity/model_sparsity": 0.8829971171402168, "compression_loss": 103.55522155761719, "distillation_loss": 5.157102584838867, "epoch": 4.17, "learning_rate": 4.0112711333750785e-05, "loss": 107.7343, "step": 4933, "task_loss": 2.7243149280548096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9788045968188802, "compression/movement_sparsity/importance_threshold": -0.00013756123740635545, "compression/movement_sparsity/linear_layer_sparsity": 0.9145124107690688, "compression/movement_sparsity/model_sparsity": 0.8830960960899066, "compression_loss": 103.5613021850586, "distillation_loss": 4.519283771514893, "epoch": 4.17, "learning_rate": 4.0109580463368816e-05, "loss": 107.2692, "step": 4934, "task_loss": 2.1239492893218994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9788692169059696, "compression/movement_sparsity/importance_threshold": -0.00013714184367907573, "compression/movement_sparsity/linear_layer_sparsity": 0.9145215804539807, "compression/movement_sparsity/model_sparsity": 0.8831049507679326, "compression_loss": 103.56739044189453, "distillation_loss": 4.871394157409668, "epoch": 4.17, "learning_rate": 4.010644959298685e-05, "loss": 108.1457, "step": 4935, "task_loss": 2.1908538341522217 }, { "compression/movement_sparsity/importance_regularization_factor": 0.978933705517851, "compression/movement_sparsity/importance_threshold": -0.00013672330324494397, "compression/movement_sparsity/linear_layer_sparsity": 0.9145230709749352, "compression/movement_sparsity/model_sparsity": 0.8831063900849069, "compression_loss": 103.57341003417969, "distillation_loss": 3.0594687461853027, "epoch": 4.17, "learning_rate": 4.010331872260489e-05, "loss": 107.3347, "step": 4936, "task_loss": 1.1005932092666626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9789980627884093, "compression/movement_sparsity/importance_threshold": -0.0001363056152350276, "compression/movement_sparsity/linear_layer_sparsity": 0.9145404683355159, "compression/movement_sparsity/model_sparsity": 0.8831231897926312, "compression_loss": 103.57939910888672, "distillation_loss": 4.168335437774658, "epoch": 4.17, "learning_rate": 4.010018785222292e-05, "loss": 107.2675, "step": 4937, "task_loss": 2.598053455352783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9790622888515299, "compression/movement_sparsity/importance_threshold": -0.00013588877878039068, "compression/movement_sparsity/linear_layer_sparsity": 0.9146141596915056, "compression/movement_sparsity/model_sparsity": 0.8831943496238415, "compression_loss": 103.58540344238281, "distillation_loss": 5.19911527633667, "epoch": 4.17, "learning_rate": 4.009705698184095e-05, "loss": 108.7044, "step": 4938, "task_loss": 2.661593437194824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9791263838410977, "compression/movement_sparsity/importance_threshold": -0.0001354727930121015, "compression/movement_sparsity/linear_layer_sparsity": 0.9146429684805139, "compression/movement_sparsity/model_sparsity": 0.8832221687423211, "compression_loss": 103.59140014648438, "distillation_loss": 3.8271543979644775, "epoch": 4.17, "learning_rate": 4.009392611145898e-05, "loss": 107.2693, "step": 4939, "task_loss": 1.9788639545440674 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9791903478909981, "compression/movement_sparsity/importance_threshold": -0.00013505765706122493, "compression/movement_sparsity/linear_layer_sparsity": 0.9147589906316108, "compression/movement_sparsity/model_sparsity": 0.8833342051756019, "compression_loss": 103.59737396240234, "distillation_loss": 4.637570381164551, "epoch": 4.18, "learning_rate": 4.009079524107702e-05, "loss": 107.9093, "step": 4940, "task_loss": 1.7599165439605713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9792541811351161, "compression/movement_sparsity/importance_threshold": -0.00013464337005882586, "compression/movement_sparsity/linear_layer_sparsity": 0.9147289894258389, "compression/movement_sparsity/model_sparsity": 0.8833052346035428, "compression_loss": 103.6033935546875, "distillation_loss": 6.147051811218262, "epoch": 4.18, "learning_rate": 4.008766437069505e-05, "loss": 107.8979, "step": 4941, "task_loss": 2.724635124206543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9793178837073369, "compression/movement_sparsity/importance_threshold": -0.0001342299311359726, "compression/movement_sparsity/linear_layer_sparsity": 0.9147467445114488, "compression/movement_sparsity/model_sparsity": 0.883322379747341, "compression_loss": 103.6093978881836, "distillation_loss": 4.143154621124268, "epoch": 4.18, "learning_rate": 4.0084533500313084e-05, "loss": 107.2416, "step": 4942, "task_loss": 2.5860044956207275 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9793814557415456, "compression/movement_sparsity/importance_threshold": -0.0001338173394237292, "compression/movement_sparsity/linear_layer_sparsity": 0.9147065123698453, "compression/movement_sparsity/model_sparsity": 0.8832835297035702, "compression_loss": 103.6153793334961, "distillation_loss": 5.226323127746582, "epoch": 4.18, "learning_rate": 4.008140262993112e-05, "loss": 108.2732, "step": 4943, "task_loss": 3.0391578674316406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9794448973716272, "compression/movement_sparsity/importance_threshold": -0.00013340559405316307, "compression/movement_sparsity/linear_layer_sparsity": 0.914689854307658, "compression/movement_sparsity/model_sparsity": 0.8832674438970651, "compression_loss": 103.62129974365234, "distillation_loss": 4.738438606262207, "epoch": 4.18, "learning_rate": 4.0078271759549155e-05, "loss": 108.218, "step": 4944, "task_loss": 2.632012128829956 }, { "compression/movement_sparsity/importance_regularization_factor": 0.979508208731467, "compression/movement_sparsity/importance_threshold": -0.0001329946941553391, "compression/movement_sparsity/linear_layer_sparsity": 0.9147160994006246, "compression/movement_sparsity/model_sparsity": 0.883292787390349, "compression_loss": 103.62714385986328, "distillation_loss": 4.697893142700195, "epoch": 4.18, "learning_rate": 4.007514088916719e-05, "loss": 107.2653, "step": 4945, "task_loss": 3.0680320262908936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9795713899549502, "compression/movement_sparsity/importance_threshold": -0.00013258463886132478, "compression/movement_sparsity/linear_layer_sparsity": 0.9148257182737011, "compression/movement_sparsity/model_sparsity": 0.8833986405179082, "compression_loss": 103.63297271728516, "distillation_loss": 5.813137054443359, "epoch": 4.18, "learning_rate": 4.0072010018785225e-05, "loss": 108.3145, "step": 4946, "task_loss": 2.376680374145508 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9796344411759619, "compression/movement_sparsity/importance_threshold": -0.00013217542730218496, "compression/movement_sparsity/linear_layer_sparsity": 0.9148985987862914, "compression/movement_sparsity/model_sparsity": 0.8834690173606844, "compression_loss": 103.63887786865234, "distillation_loss": 7.650238990783691, "epoch": 4.18, "learning_rate": 4.006887914840326e-05, "loss": 108.2981, "step": 4947, "task_loss": 4.499307155609131 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9796973625283871, "compression/movement_sparsity/importance_threshold": -0.00013176705860898624, "compression/movement_sparsity/linear_layer_sparsity": 0.9148679417512996, "compression/movement_sparsity/model_sparsity": 0.8834394134891567, "compression_loss": 103.64472961425781, "distillation_loss": 4.16168212890625, "epoch": 4.18, "learning_rate": 4.0065748278021295e-05, "loss": 108.1785, "step": 4948, "task_loss": 2.176762819290161 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9797601541461111, "compression/movement_sparsity/importance_threshold": -0.00013135953191279436, "compression/movement_sparsity/linear_layer_sparsity": 0.9148624327858519, "compression/movement_sparsity/model_sparsity": 0.8834340937736196, "compression_loss": 103.65058898925781, "distillation_loss": 4.6738739013671875, "epoch": 4.18, "learning_rate": 4.006261740763933e-05, "loss": 107.944, "step": 4949, "task_loss": 2.4931294918060303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9798228161630189, "compression/movement_sparsity/importance_threshold": -0.00013095284634467505, "compression/movement_sparsity/linear_layer_sparsity": 0.9149844885657725, "compression/movement_sparsity/model_sparsity": 0.8835519565620125, "compression_loss": 103.65642547607422, "distillation_loss": 5.454440116882324, "epoch": 4.18, "learning_rate": 4.005948653725736e-05, "loss": 108.2923, "step": 4950, "task_loss": 2.8382556438446045 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9798853487129957, "compression/movement_sparsity/importance_threshold": -0.0001305470010356958, "compression/movement_sparsity/linear_layer_sparsity": 0.9150222881771782, "compression/movement_sparsity/model_sparsity": 0.8835884576404812, "compression_loss": 103.66227722167969, "distillation_loss": 3.1105918884277344, "epoch": 4.19, "learning_rate": 4.00563556668754e-05, "loss": 107.8441, "step": 4951, "task_loss": 2.1158368587493896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9799477519299268, "compression/movement_sparsity/importance_threshold": -0.00013014199511692144, "compression/movement_sparsity/linear_layer_sparsity": 0.9150545549748009, "compression/movement_sparsity/model_sparsity": 0.8836196159743412, "compression_loss": 103.66809844970703, "distillation_loss": 4.62607479095459, "epoch": 4.19, "learning_rate": 4.005322479649343e-05, "loss": 107.8947, "step": 4952, "task_loss": 2.1516647338867188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9800100259476971, "compression/movement_sparsity/importance_threshold": -0.00012973782771941774, "compression/movement_sparsity/linear_layer_sparsity": 0.9150815512903285, "compression/movement_sparsity/model_sparsity": 0.88364568488338, "compression_loss": 103.67388153076172, "distillation_loss": 2.979776382446289, "epoch": 4.19, "learning_rate": 4.005009392611146e-05, "loss": 107.4927, "step": 4953, "task_loss": 1.3795605897903442 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9800721709001919, "compression/movement_sparsity/importance_threshold": -0.00012933449797425042, "compression/movement_sparsity/linear_layer_sparsity": 0.91505491269983, "compression/movement_sparsity/model_sparsity": 0.883619961410415, "compression_loss": 103.67961883544922, "distillation_loss": 4.7320404052734375, "epoch": 4.19, "learning_rate": 4.00469630557295e-05, "loss": 108.1684, "step": 4954, "task_loss": 2.212449789047241 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9801341869212963, "compression/movement_sparsity/importance_threshold": -0.00012893200501248783, "compression/movement_sparsity/linear_layer_sparsity": 0.9151999344266173, "compression/movement_sparsity/model_sparsity": 0.8837600011947482, "compression_loss": 103.68535614013672, "distillation_loss": 5.41484260559082, "epoch": 4.19, "learning_rate": 4.004383218534753e-05, "loss": 108.5436, "step": 4955, "task_loss": 3.7184066772460938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9801960741448954, "compression/movement_sparsity/importance_threshold": -0.00012853034796519395, "compression/movement_sparsity/linear_layer_sparsity": 0.915241108577464, "compression/movement_sparsity/model_sparsity": 0.8837997608868468, "compression_loss": 103.69103240966797, "distillation_loss": 3.7373833656311035, "epoch": 4.19, "learning_rate": 4.004070131496556e-05, "loss": 107.7886, "step": 4956, "task_loss": 2.0862343311309814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9802578327048743, "compression/movement_sparsity/importance_threshold": -0.0001281295259634354, "compression/movement_sparsity/linear_layer_sparsity": 0.9152820203966227, "compression/movement_sparsity/model_sparsity": 0.8838392672591578, "compression_loss": 103.6966781616211, "distillation_loss": 6.642441749572754, "epoch": 4.19, "learning_rate": 4.0037570444583595e-05, "loss": 108.7807, "step": 4957, "task_loss": 2.3959405422210693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9803194627351183, "compression/movement_sparsity/importance_threshold": -0.0001277295381382779, "compression/movement_sparsity/linear_layer_sparsity": 0.9153594917137529, "compression/movement_sparsity/model_sparsity": 0.883914077198215, "compression_loss": 103.70235443115234, "distillation_loss": 5.843071937561035, "epoch": 4.19, "learning_rate": 4.003443957420163e-05, "loss": 108.8933, "step": 4958, "task_loss": 2.9057092666625977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9803809643695124, "compression/movement_sparsity/importance_threshold": -0.00012733038362078807, "compression/movement_sparsity/linear_layer_sparsity": 0.9153708077488393, "compression/movement_sparsity/model_sparsity": 0.883925004492684, "compression_loss": 103.70796966552734, "distillation_loss": 4.775547027587891, "epoch": 4.19, "learning_rate": 4.0031308703819665e-05, "loss": 108.1988, "step": 4959, "task_loss": 2.249633550643921 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9804423377419418, "compression/movement_sparsity/importance_threshold": -0.00012693206154203076, "compression/movement_sparsity/linear_layer_sparsity": 0.9153920804639016, "compression/movement_sparsity/model_sparsity": 0.8839455464245415, "compression_loss": 103.71365356445312, "distillation_loss": 4.6187825202941895, "epoch": 4.19, "learning_rate": 4.00281778334377e-05, "loss": 108.0454, "step": 4960, "task_loss": 1.8897982835769653 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9805035829862916, "compression/movement_sparsity/importance_threshold": -0.00012653457103307432, "compression/movement_sparsity/linear_layer_sparsity": 0.9153806690354741, "compression/movement_sparsity/model_sparsity": 0.883934527013786, "compression_loss": 103.71931457519531, "distillation_loss": 4.770403861999512, "epoch": 4.19, "learning_rate": 4.002504696305573e-05, "loss": 108.0988, "step": 4961, "task_loss": 2.467261791229248 }, { "compression/movement_sparsity/importance_regularization_factor": 0.980564700236447, "compression/movement_sparsity/importance_threshold": -0.00012613791122498188, "compression/movement_sparsity/linear_layer_sparsity": 0.915379560087884, "compression/movement_sparsity/model_sparsity": 0.8839334561619572, "compression_loss": 103.7249526977539, "distillation_loss": 4.761317253112793, "epoch": 4.19, "learning_rate": 4.002191609267377e-05, "loss": 107.7423, "step": 4962, "task_loss": 3.055222749710083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9806256896262932, "compression/movement_sparsity/importance_threshold": -0.00012574208124882177, "compression/movement_sparsity/linear_layer_sparsity": 0.9155125980261973, "compression/movement_sparsity/model_sparsity": 0.8840619238378169, "compression_loss": 103.73051452636719, "distillation_loss": 4.94692850112915, "epoch": 4.2, "learning_rate": 4.00187852222918e-05, "loss": 108.2952, "step": 4963, "task_loss": 2.9296822547912598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.980686551289715, "compression/movement_sparsity/importance_threshold": -0.00012534708023565887, "compression/movement_sparsity/linear_layer_sparsity": 0.9155526155327832, "compression/movement_sparsity/model_sparsity": 0.8841005666199434, "compression_loss": 103.73614501953125, "distillation_loss": 3.46132493019104, "epoch": 4.2, "learning_rate": 4.001565435190983e-05, "loss": 108.1061, "step": 4964, "task_loss": 2.1058313846588135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.980747285360598, "compression/movement_sparsity/importance_threshold": -0.00012495290731656065, "compression/movement_sparsity/linear_layer_sparsity": 0.9155925972668663, "compression/movement_sparsity/model_sparsity": 0.8841391748584625, "compression_loss": 103.74172973632812, "distillation_loss": 3.667452096939087, "epoch": 4.2, "learning_rate": 4.001252348152787e-05, "loss": 107.3647, "step": 4965, "task_loss": 1.827082872390747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.980807891972827, "compression/movement_sparsity/importance_threshold": -0.0001245595616225911, "compression/movement_sparsity/linear_layer_sparsity": 0.9156002764308238, "compression/movement_sparsity/model_sparsity": 0.8841465902195141, "compression_loss": 103.74726867675781, "distillation_loss": 4.737611293792725, "epoch": 4.2, "learning_rate": 4.00093926111459e-05, "loss": 108.2347, "step": 4966, "task_loss": 2.6571435928344727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9808683712602874, "compression/movement_sparsity/importance_threshold": -0.0001241670422848177, "compression/movement_sparsity/linear_layer_sparsity": 0.9155896758457955, "compression/movement_sparsity/model_sparsity": 0.8841363537971928, "compression_loss": 103.75284576416016, "distillation_loss": 4.120554447174072, "epoch": 4.2, "learning_rate": 4.000626174076393e-05, "loss": 108.396, "step": 4967, "task_loss": 2.454707145690918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9809287233568642, "compression/movement_sparsity/importance_threshold": -0.0001237753484343053, "compression/movement_sparsity/linear_layer_sparsity": 0.9155832606436074, "compression/movement_sparsity/model_sparsity": 0.8841301589769354, "compression_loss": 103.75833129882812, "distillation_loss": 5.389728546142578, "epoch": 4.2, "learning_rate": 4.0003130870381965e-05, "loss": 108.501, "step": 4968, "task_loss": 2.711047887802124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9809889483964425, "compression/movement_sparsity/importance_threshold": -0.00012338447920212138, "compression/movement_sparsity/linear_layer_sparsity": 0.9155300907801192, "compression/movement_sparsity/model_sparsity": 0.8840788156618274, "compression_loss": 103.7637710571289, "distillation_loss": 4.9808454513549805, "epoch": 4.2, "learning_rate": 4e-05, "loss": 108.301, "step": 4969, "task_loss": 2.5470998287200928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9810490465129075, "compression/movement_sparsity/importance_threshold": -0.00012299443371933082, "compression/movement_sparsity/linear_layer_sparsity": 0.9155816985776472, "compression/movement_sparsity/model_sparsity": 0.8841286505727463, "compression_loss": 103.76925659179688, "distillation_loss": 4.676291465759277, "epoch": 4.2, "learning_rate": 3.9996869129618035e-05, "loss": 108.8184, "step": 4970, "task_loss": 3.29413104057312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9811090178401443, "compression/movement_sparsity/importance_threshold": -0.0001226052111170002, "compression/movement_sparsity/linear_layer_sparsity": 0.915639137293149, "compression/movement_sparsity/model_sparsity": 0.8841841160916686, "compression_loss": 103.77472686767578, "distillation_loss": 4.489619731903076, "epoch": 4.2, "learning_rate": 3.999373825923607e-05, "loss": 107.9161, "step": 4971, "task_loss": 2.770730972290039 }, { "compression/movement_sparsity/importance_regularization_factor": 0.981168862512038, "compression/movement_sparsity/importance_threshold": -0.00012221681052619615, "compression/movement_sparsity/linear_layer_sparsity": 0.9156397692740338, "compression/movement_sparsity/model_sparsity": 0.8841847263620657, "compression_loss": 103.78028106689453, "distillation_loss": 4.138166904449463, "epoch": 4.2, "learning_rate": 3.99906073888541e-05, "loss": 108.5964, "step": 4972, "task_loss": 2.790201425552368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.981228580662474, "compression/movement_sparsity/importance_threshold": -0.00012182923107798352, "compression/movement_sparsity/linear_layer_sparsity": 0.9156367882321248, "compression/movement_sparsity/model_sparsity": 0.884181847728117, "compression_loss": 103.7857437133789, "distillation_loss": 4.700735092163086, "epoch": 4.2, "learning_rate": 3.998747651847214e-05, "loss": 108.9152, "step": 4973, "task_loss": 2.515467643737793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9812881724253372, "compression/movement_sparsity/importance_threshold": -0.00012144247190342804, "compression/movement_sparsity/linear_layer_sparsity": 0.9157154519660186, "compression/movement_sparsity/model_sparsity": 0.8842578091207536, "compression_loss": 103.79126739501953, "distillation_loss": 4.725978851318359, "epoch": 4.2, "learning_rate": 3.998434564809017e-05, "loss": 108.2143, "step": 4974, "task_loss": 3.0130670070648193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9813476379345127, "compression/movement_sparsity/importance_threshold": -0.0001210565321335972, "compression/movement_sparsity/linear_layer_sparsity": 0.9157204481922581, "compression/movement_sparsity/model_sparsity": 0.8842626337112516, "compression_loss": 103.79668426513672, "distillation_loss": 4.4905595779418945, "epoch": 4.21, "learning_rate": 3.99812147777082e-05, "loss": 107.8745, "step": 4975, "task_loss": 2.6924655437469482 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9814069773238858, "compression/movement_sparsity/importance_threshold": -0.00012067141089955671, "compression/movement_sparsity/linear_layer_sparsity": 0.915772747591509, "compression/movement_sparsity/model_sparsity": 0.8843131364652463, "compression_loss": 103.80213928222656, "distillation_loss": 5.120213508605957, "epoch": 4.21, "learning_rate": 3.997808390732623e-05, "loss": 108.2474, "step": 4976, "task_loss": 3.1782617568969727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9814661907273416, "compression/movement_sparsity/importance_threshold": -0.00012028710733237233, "compression/movement_sparsity/linear_layer_sparsity": 0.9158479056201178, "compression/movement_sparsity/model_sparsity": 0.8843857125843594, "compression_loss": 103.80752563476562, "distillation_loss": 3.3218088150024414, "epoch": 4.21, "learning_rate": 3.997495303694427e-05, "loss": 107.3101, "step": 4977, "task_loss": 1.9380847215652466 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9815252782787652, "compression/movement_sparsity/importance_threshold": -0.00011990362056310978, "compression/movement_sparsity/linear_layer_sparsity": 0.9158126100839156, "compression/movement_sparsity/model_sparsity": 0.8843516295584075, "compression_loss": 103.81293487548828, "distillation_loss": 3.3660597801208496, "epoch": 4.21, "learning_rate": 3.99718221665623e-05, "loss": 108.1238, "step": 4978, "task_loss": 1.379673957824707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9815842401120418, "compression/movement_sparsity/importance_threshold": -0.00011952094972283566, "compression/movement_sparsity/linear_layer_sparsity": 0.9157686099053393, "compression/movement_sparsity/model_sparsity": 0.8843091409213257, "compression_loss": 103.81830596923828, "distillation_loss": 4.208396911621094, "epoch": 4.21, "learning_rate": 3.996869129618034e-05, "loss": 107.5468, "step": 4979, "task_loss": 1.991870641708374 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9816430763610565, "compression/movement_sparsity/importance_threshold": -0.00011913909394261485, "compression/movement_sparsity/linear_layer_sparsity": 0.9157727952881795, "compression/movement_sparsity/model_sparsity": 0.8843131825233895, "compression_loss": 103.82365417480469, "distillation_loss": 6.107856750488281, "epoch": 4.21, "learning_rate": 3.996556042579837e-05, "loss": 108.2807, "step": 4980, "task_loss": 2.892411231994629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9817017871596944, "compression/movement_sparsity/importance_threshold": -0.00011875805235351568, "compression/movement_sparsity/linear_layer_sparsity": 0.9158596509252391, "compression/movement_sparsity/model_sparsity": 0.884397054402117, "compression_loss": 103.82904052734375, "distillation_loss": 6.41664981842041, "epoch": 4.21, "learning_rate": 3.996242955541641e-05, "loss": 108.7472, "step": 4981, "task_loss": 3.0750980377197266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9817603726418407, "compression/movement_sparsity/importance_threshold": -0.00011837782408660215, "compression/movement_sparsity/linear_layer_sparsity": 0.9158793973468441, "compression/movement_sparsity/model_sparsity": 0.8844161224733927, "compression_loss": 103.83440399169922, "distillation_loss": 4.836499214172363, "epoch": 4.21, "learning_rate": 3.9959298685034444e-05, "loss": 107.7022, "step": 4982, "task_loss": 3.1106932163238525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9818188329413805, "compression/movement_sparsity/importance_threshold": -0.00011799840827294086, "compression/movement_sparsity/linear_layer_sparsity": 0.9158736498980437, "compression/movement_sparsity/model_sparsity": 0.8844105724671398, "compression_loss": 103.83973693847656, "distillation_loss": 5.79898738861084, "epoch": 4.21, "learning_rate": 3.9956167814652475e-05, "loss": 108.4376, "step": 4983, "task_loss": 3.161250114440918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.981877168192199, "compression/movement_sparsity/importance_threshold": -0.00011761980404359842, "compression/movement_sparsity/linear_layer_sparsity": 0.9159665749364299, "compression/movement_sparsity/model_sparsity": 0.8845003052445867, "compression_loss": 103.84501647949219, "distillation_loss": 3.6509170532226562, "epoch": 4.21, "learning_rate": 3.9953036944270514e-05, "loss": 108.4197, "step": 4984, "task_loss": 2.645055055618286 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9819353785281812, "compression/movement_sparsity/importance_threshold": -0.0001172420105296397, "compression/movement_sparsity/linear_layer_sparsity": 0.9160352104453419, "compression/movement_sparsity/model_sparsity": 0.88456658291262, "compression_loss": 103.85028076171875, "distillation_loss": 3.6058974266052246, "epoch": 4.21, "learning_rate": 3.9949906073888546e-05, "loss": 108.1965, "step": 4985, "task_loss": 2.8972392082214355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9819934640832124, "compression/movement_sparsity/importance_threshold": -0.0001168650268621313, "compression/movement_sparsity/linear_layer_sparsity": 0.9160773266054317, "compression/movement_sparsity/model_sparsity": 0.8846072522530464, "compression_loss": 103.8555908203125, "distillation_loss": 3.8345165252685547, "epoch": 4.21, "learning_rate": 3.994677520350658e-05, "loss": 108.3179, "step": 4986, "task_loss": 1.89413583278656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9820514249911777, "compression/movement_sparsity/importance_threshold": -0.00011648885217214067, "compression/movement_sparsity/linear_layer_sparsity": 0.9161454374509678, "compression/movement_sparsity/model_sparsity": 0.8846730232815048, "compression_loss": 103.86090087890625, "distillation_loss": 5.447353363037109, "epoch": 4.22, "learning_rate": 3.994364433312461e-05, "loss": 108.3103, "step": 4987, "task_loss": 2.4171149730682373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9821092613859622, "compression/movement_sparsity/importance_threshold": -0.00011611348559073097, "compression/movement_sparsity/linear_layer_sparsity": 0.9162210724462821, "compression/movement_sparsity/model_sparsity": 0.8847460599820496, "compression_loss": 103.86614990234375, "distillation_loss": 4.788578987121582, "epoch": 4.22, "learning_rate": 3.994051346274265e-05, "loss": 108.1849, "step": 4988, "task_loss": 4.36600923538208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9821669734014511, "compression/movement_sparsity/importance_threshold": -0.0001157389262489714, "compression/movement_sparsity/linear_layer_sparsity": 0.9162360134283298, "compression/movement_sparsity/model_sparsity": 0.8847604876954, "compression_loss": 103.87147521972656, "distillation_loss": 3.356156587600708, "epoch": 4.22, "learning_rate": 3.993738259236068e-05, "loss": 107.6521, "step": 4989, "task_loss": 2.2674853801727295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9822245611715296, "compression/movement_sparsity/importance_threshold": -0.00011536517327792508, "compression/movement_sparsity/linear_layer_sparsity": 0.9162767821574769, "compression/movement_sparsity/model_sparsity": 0.8847998558932816, "compression_loss": 103.87672424316406, "distillation_loss": 3.6069352626800537, "epoch": 4.22, "learning_rate": 3.993425172197871e-05, "loss": 108.1678, "step": 4990, "task_loss": 3.4576241970062256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9822820248300825, "compression/movement_sparsity/importance_threshold": -0.00011499222580866036, "compression/movement_sparsity/linear_layer_sparsity": 0.916248092610145, "compression/movement_sparsity/model_sparsity": 0.88477215192016, "compression_loss": 103.88203430175781, "distillation_loss": 3.751178741455078, "epoch": 4.22, "learning_rate": 3.993112085159675e-05, "loss": 107.8528, "step": 4991, "task_loss": 1.297133207321167 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9823393645109953, "compression/movement_sparsity/importance_threshold": -0.0001146200829722421, "compression/movement_sparsity/linear_layer_sparsity": 0.9162932017363115, "compression/movement_sparsity/model_sparsity": 0.8848157114090707, "compression_loss": 103.88727569580078, "distillation_loss": 5.279275894165039, "epoch": 4.22, "learning_rate": 3.992798998121478e-05, "loss": 108.1345, "step": 4992, "task_loss": 3.3542962074279785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9823965803481531, "compression/movement_sparsity/importance_threshold": -0.00011424874389973604, "compression/movement_sparsity/linear_layer_sparsity": 0.916348195997448, "compression/movement_sparsity/model_sparsity": 0.8848688164481552, "compression_loss": 103.89253234863281, "distillation_loss": 6.4365692138671875, "epoch": 4.22, "learning_rate": 3.9924859110832813e-05, "loss": 108.1277, "step": 4993, "task_loss": 3.56913685798645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9824536724754409, "compression/movement_sparsity/importance_threshold": -0.00011387820772220964, "compression/movement_sparsity/linear_layer_sparsity": 0.9163849224337665, "compression/movement_sparsity/model_sparsity": 0.8849042812184024, "compression_loss": 103.89773559570312, "distillation_loss": 3.897648572921753, "epoch": 4.22, "learning_rate": 3.9921728240450845e-05, "loss": 108.2212, "step": 4994, "task_loss": 2.2362706661224365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9825106410267439, "compression/movement_sparsity/importance_threshold": -0.00011350847357072778, "compression/movement_sparsity/linear_layer_sparsity": 0.9164331914643564, "compression/movement_sparsity/model_sparsity": 0.8849508920592987, "compression_loss": 103.9029541015625, "distillation_loss": 5.085212707519531, "epoch": 4.22, "learning_rate": 3.9918597370068884e-05, "loss": 107.7614, "step": 4995, "task_loss": 3.0172066688537598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9825674861359474, "compression/movement_sparsity/importance_threshold": -0.00011313954057635533, "compression/movement_sparsity/linear_layer_sparsity": 0.9164508273082899, "compression/movement_sparsity/model_sparsity": 0.8849679220577388, "compression_loss": 103.90817260742188, "distillation_loss": 4.1353759765625, "epoch": 4.22, "learning_rate": 3.9915466499686916e-05, "loss": 107.5914, "step": 4996, "task_loss": 1.713090419769287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9826242079369362, "compression/movement_sparsity/importance_threshold": -0.00011277140787016148, "compression/movement_sparsity/linear_layer_sparsity": 0.916520702930636, "compression/movement_sparsity/model_sparsity": 0.8850353972374948, "compression_loss": 103.91336059570312, "distillation_loss": 4.457376480102539, "epoch": 4.22, "learning_rate": 3.991233562930495e-05, "loss": 108.3833, "step": 4997, "task_loss": 3.1059889793395996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9826808065635957, "compression/movement_sparsity/importance_threshold": -0.00011240407458321024, "compression/movement_sparsity/linear_layer_sparsity": 0.9165310411839763, "compression/movement_sparsity/model_sparsity": 0.8850453803400287, "compression_loss": 103.9184341430664, "distillation_loss": 5.591254234313965, "epoch": 4.22, "learning_rate": 3.990920475892298e-05, "loss": 108.6697, "step": 4998, "task_loss": 2.037196159362793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9827372821498109, "compression/movement_sparsity/importance_threshold": -0.00011203753984656822, "compression/movement_sparsity/linear_layer_sparsity": 0.9165706532688627, "compression/movement_sparsity/model_sparsity": 0.8850836316279381, "compression_loss": 103.923583984375, "distillation_loss": 4.223067283630371, "epoch": 4.23, "learning_rate": 3.990607388854102e-05, "loss": 107.984, "step": 4999, "task_loss": 2.529996156692505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9827936348294671, "compression/movement_sparsity/importance_threshold": -0.00011167180279130027, "compression/movement_sparsity/linear_layer_sparsity": 0.9165533274532878, "compression/movement_sparsity/model_sparsity": 0.8850669010074286, "compression_loss": 103.9287338256836, "distillation_loss": 4.936398506164551, "epoch": 4.23, "learning_rate": 3.990294301815905e-05, "loss": 107.7701, "step": 5000, "task_loss": 3.2203876972198486 }, { "epoch": 4.23, "eval_accuracy": 0.5398019801980198, "eval_loss": 108.07545471191406, "eval_runtime": 209.8513, "eval_samples_per_second": 120.323, "eval_steps_per_second": 0.944, "step": 5000 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9828498647364494, "compression/movement_sparsity/importance_threshold": -0.00011130686254847387, "compression/movement_sparsity/linear_layer_sparsity": 0.9165707725105391, "compression/movement_sparsity/model_sparsity": 0.885083746773296, "compression_loss": 103.93382263183594, "distillation_loss": 4.227768421173096, "epoch": 4.23, "learning_rate": 3.989981214777708e-05, "loss": 108.4991, "step": 5001, "task_loss": 1.7458736896514893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9829059720046428, "compression/movement_sparsity/importance_threshold": -0.00011094271824915476, "compression/movement_sparsity/linear_layer_sparsity": 0.9166287955102551, "compression/movement_sparsity/model_sparsity": 0.8851397765044723, "compression_loss": 103.93889617919922, "distillation_loss": 5.8764448165893555, "epoch": 4.23, "learning_rate": 3.989668127739512e-05, "loss": 108.2658, "step": 5002, "task_loss": 3.405597448348999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9829619567679325, "compression/movement_sparsity/importance_threshold": -0.00011057936902440866, "compression/movement_sparsity/linear_layer_sparsity": 0.9165896842404095, "compression/movement_sparsity/model_sparsity": 0.8851020088270661, "compression_loss": 103.94393920898438, "distillation_loss": 4.89530611038208, "epoch": 4.23, "learning_rate": 3.989355040701315e-05, "loss": 108.0413, "step": 5003, "task_loss": 1.6211191415786743 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9830178191602038, "compression/movement_sparsity/importance_threshold": -0.00011021681400530132, "compression/movement_sparsity/linear_layer_sparsity": 0.916676706815816, "compression/movement_sparsity/model_sparsity": 0.8851860419092947, "compression_loss": 103.94903564453125, "distillation_loss": 4.2180681228637695, "epoch": 4.23, "learning_rate": 3.989041953663118e-05, "loss": 108.4965, "step": 5004, "task_loss": 1.9761581420898438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9830735593153417, "compression/movement_sparsity/importance_threshold": -0.00010985505232289933, "compression/movement_sparsity/linear_layer_sparsity": 0.916728875049223, "compression/movement_sparsity/model_sparsity": 0.8852364180033958, "compression_loss": 103.95413208007812, "distillation_loss": 5.20725154876709, "epoch": 4.23, "learning_rate": 3.9887288666249215e-05, "loss": 108.7808, "step": 5005, "task_loss": 2.0161399841308594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9831291773672313, "compression/movement_sparsity/importance_threshold": -0.00010949408310826931, "compression/movement_sparsity/linear_layer_sparsity": 0.9167800774250514, "compression/movement_sparsity/model_sparsity": 0.8852858614200976, "compression_loss": 103.95917510986328, "distillation_loss": 4.589644432067871, "epoch": 4.23, "learning_rate": 3.9884157795867254e-05, "loss": 108.0629, "step": 5006, "task_loss": 3.0137059688568115 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9831846734497577, "compression/movement_sparsity/importance_threshold": -0.00010913390549247698, "compression/movement_sparsity/linear_layer_sparsity": 0.9168093154840945, "compression/movement_sparsity/model_sparsity": 0.8853140950618659, "compression_loss": 103.96422576904297, "distillation_loss": 3.661501884460449, "epoch": 4.23, "learning_rate": 3.9881026925485285e-05, "loss": 107.7047, "step": 5007, "task_loss": 1.3542741537094116 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9832400476968063, "compression/movement_sparsity/importance_threshold": -0.00010877451860658634, "compression/movement_sparsity/linear_layer_sparsity": 0.9168352505487025, "compression/movement_sparsity/model_sparsity": 0.885339139177219, "compression_loss": 103.96916198730469, "distillation_loss": 3.885293483734131, "epoch": 4.23, "learning_rate": 3.987789605510332e-05, "loss": 108.1216, "step": 5008, "task_loss": 1.9020805358886719 }, { "compression/movement_sparsity/importance_regularization_factor": 0.983295300242262, "compression/movement_sparsity/importance_threshold": -0.00010841592158166661, "compression/movement_sparsity/linear_layer_sparsity": 0.9168307789858391, "compression/movement_sparsity/model_sparsity": 0.885334821226296, "compression_loss": 103.97418975830078, "distillation_loss": 5.124056816101074, "epoch": 4.23, "learning_rate": 3.987476518472135e-05, "loss": 108.2864, "step": 5009, "task_loss": 2.263322353363037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.98335043122001, "compression/movement_sparsity/importance_threshold": -0.00010805811354878177, "compression/movement_sparsity/linear_layer_sparsity": 0.9169002849589885, "compression/movement_sparsity/model_sparsity": 0.8854019394554423, "compression_loss": 103.9791488647461, "distillation_loss": 4.618778228759766, "epoch": 4.23, "learning_rate": 3.987163431433939e-05, "loss": 108.118, "step": 5010, "task_loss": 2.697439432144165 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9834054407639355, "compression/movement_sparsity/importance_threshold": -0.00010770109363899757, "compression/movement_sparsity/linear_layer_sparsity": 0.9169141408417814, "compression/movement_sparsity/model_sparsity": 0.8854153193460357, "compression_loss": 103.9841079711914, "distillation_loss": 3.3176939487457275, "epoch": 4.24, "learning_rate": 3.986850344395742e-05, "loss": 107.6461, "step": 5011, "task_loss": 2.554422378540039 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9834603290079236, "compression/movement_sparsity/importance_threshold": -0.00010734486098338148, "compression/movement_sparsity/linear_layer_sparsity": 0.9169410417639678, "compression/movement_sparsity/model_sparsity": 0.8854412961387881, "compression_loss": 103.989013671875, "distillation_loss": 4.895946025848389, "epoch": 4.24, "learning_rate": 3.986537257357546e-05, "loss": 107.8552, "step": 5012, "task_loss": 1.8799270391464233 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9835150960858594, "compression/movement_sparsity/importance_threshold": -0.00010698941471299923, "compression/movement_sparsity/linear_layer_sparsity": 0.9169195782622234, "compression/movement_sparsity/model_sparsity": 0.8854205699743579, "compression_loss": 103.99386596679688, "distillation_loss": 4.393520355224609, "epoch": 4.24, "learning_rate": 3.986224170319349e-05, "loss": 108.4448, "step": 5013, "task_loss": 2.076082229614258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9835697421316282, "compression/movement_sparsity/importance_threshold": -0.00010663475395891568, "compression/movement_sparsity/linear_layer_sparsity": 0.9168951575669051, "compression/movement_sparsity/model_sparsity": 0.8853969882050508, "compression_loss": 103.99878692626953, "distillation_loss": 3.6827821731567383, "epoch": 4.24, "learning_rate": 3.985911083281152e-05, "loss": 107.239, "step": 5014, "task_loss": 2.2017860412597656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9836242672791149, "compression/movement_sparsity/importance_threshold": -0.00010628087785219831, "compression/movement_sparsity/linear_layer_sparsity": 0.9169780305319744, "compression/movement_sparsity/model_sparsity": 0.8854770142288227, "compression_loss": 104.003662109375, "distillation_loss": 5.0807695388793945, "epoch": 4.24, "learning_rate": 3.985597996242956e-05, "loss": 108.4338, "step": 5015, "task_loss": 2.5333874225616455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9836786716622047, "compression/movement_sparsity/importance_threshold": -0.00010592778552391373, "compression/movement_sparsity/linear_layer_sparsity": 0.9169399566647131, "compression/movement_sparsity/model_sparsity": 0.8854402483160309, "compression_loss": 104.00849151611328, "distillation_loss": 5.466082572937012, "epoch": 4.24, "learning_rate": 3.985284909204759e-05, "loss": 108.0936, "step": 5016, "task_loss": 2.7008936405181885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9837329554147829, "compression/movement_sparsity/importance_threshold": -0.00010557547610512506, "compression/movement_sparsity/linear_layer_sparsity": 0.9169859481792846, "compression/movement_sparsity/model_sparsity": 0.8854846598805903, "compression_loss": 104.01339721679688, "distillation_loss": 5.272060394287109, "epoch": 4.24, "learning_rate": 3.984971822166563e-05, "loss": 108.5074, "step": 5017, "task_loss": 2.441253662109375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9837871186707345, "compression/movement_sparsity/importance_threshold": -0.0001052239487269015, "compression/movement_sparsity/linear_layer_sparsity": 0.91705466715737, "compression/movement_sparsity/model_sparsity": 0.8855510181503743, "compression_loss": 104.01817321777344, "distillation_loss": 4.43195104598999, "epoch": 4.24, "learning_rate": 3.984658735128366e-05, "loss": 108.1, "step": 5018, "task_loss": 3.2911996841430664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9838411615639447, "compression/movement_sparsity/importance_threshold": -0.00010487320252030707, "compression/movement_sparsity/linear_layer_sparsity": 0.917031975466359, "compression/movement_sparsity/model_sparsity": 0.8855291059887572, "compression_loss": 104.02308654785156, "distillation_loss": 4.403872489929199, "epoch": 4.24, "learning_rate": 3.9843456480901694e-05, "loss": 108.5147, "step": 5019, "task_loss": 3.073835849761963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9838950842282987, "compression/movement_sparsity/importance_threshold": -0.00010452323661640749, "compression/movement_sparsity/linear_layer_sparsity": 0.9171436452962688, "compression/movement_sparsity/model_sparsity": 0.8856369396164732, "compression_loss": 104.02796936035156, "distillation_loss": 5.214191913604736, "epoch": 4.24, "learning_rate": 3.9840325610519726e-05, "loss": 108.2093, "step": 5020, "task_loss": 3.1473560333251953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9839488867976816, "compression/movement_sparsity/importance_threshold": -0.0001041740501462711, "compression/movement_sparsity/linear_layer_sparsity": 0.9171572865440443, "compression/movement_sparsity/model_sparsity": 0.8856501122454221, "compression_loss": 104.03280639648438, "distillation_loss": 4.2972941398620605, "epoch": 4.24, "learning_rate": 3.9837194740137764e-05, "loss": 107.9438, "step": 5021, "task_loss": 2.3091886043548584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9840025694059783, "compression/movement_sparsity/importance_threshold": -0.00010382564224096277, "compression/movement_sparsity/linear_layer_sparsity": 0.9172569487371448, "compression/movement_sparsity/model_sparsity": 0.8857463507355928, "compression_loss": 104.03767395019531, "distillation_loss": 3.515420913696289, "epoch": 4.24, "learning_rate": 3.9834063869755796e-05, "loss": 108.5451, "step": 5022, "task_loss": 1.6604052782058716 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9840561321870743, "compression/movement_sparsity/importance_threshold": -0.00010347801203154736, "compression/movement_sparsity/linear_layer_sparsity": 0.9172771125046171, "compression/movement_sparsity/model_sparsity": 0.8857658218156215, "compression_loss": 104.04249572753906, "distillation_loss": 3.835801601409912, "epoch": 4.25, "learning_rate": 3.983093299937383e-05, "loss": 107.7302, "step": 5023, "task_loss": 1.392683506011963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9841095752748545, "compression/movement_sparsity/importance_threshold": -0.00010313115864909236, "compression/movement_sparsity/linear_layer_sparsity": 0.91728022471237, "compression/movement_sparsity/model_sparsity": 0.8857688271094638, "compression_loss": 104.04731750488281, "distillation_loss": 4.611047744750977, "epoch": 4.25, "learning_rate": 3.982780212899186e-05, "loss": 108.0716, "step": 5024, "task_loss": 2.53863525390625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9841628988032042, "compression/movement_sparsity/importance_threshold": -0.00010278508122466435, "compression/movement_sparsity/linear_layer_sparsity": 0.917294271381845, "compression/movement_sparsity/model_sparsity": 0.8857823912326298, "compression_loss": 104.05213165283203, "distillation_loss": 5.072144031524658, "epoch": 4.25, "learning_rate": 3.98246712586099e-05, "loss": 108.5776, "step": 5025, "task_loss": 3.0825793743133545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9842161029060085, "compression/movement_sparsity/importance_threshold": -0.00010243977888932648, "compression/movement_sparsity/linear_layer_sparsity": 0.9172901217715078, "compression/movement_sparsity/model_sparsity": 0.8857783841741733, "compression_loss": 104.0569076538086, "distillation_loss": 4.703003883361816, "epoch": 4.25, "learning_rate": 3.982154038822793e-05, "loss": 108.384, "step": 5026, "task_loss": 2.5511415004730225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9842691877171524, "compression/movement_sparsity/importance_threshold": -0.00010209525077414794, "compression/movement_sparsity/linear_layer_sparsity": 0.9172911710982597, "compression/movement_sparsity/model_sparsity": 0.8857793974533232, "compression_loss": 104.06168365478516, "distillation_loss": 3.736344337463379, "epoch": 4.25, "learning_rate": 3.981840951784596e-05, "loss": 107.3874, "step": 5027, "task_loss": 2.4716603755950928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9843221533705212, "compression/movement_sparsity/importance_threshold": -0.0001017514960101936, "compression/movement_sparsity/linear_layer_sparsity": 0.9173552873476377, "compression/movement_sparsity/model_sparsity": 0.8858413111122905, "compression_loss": 104.06636810302734, "distillation_loss": 6.079329490661621, "epoch": 4.25, "learning_rate": 3.9815278647464e-05, "loss": 108.9459, "step": 5028, "task_loss": 2.815108299255371 }, { "compression/movement_sparsity/importance_regularization_factor": 0.984375, "compression/movement_sparsity/importance_threshold": -0.00010140851372852921, "compression/movement_sparsity/linear_layer_sparsity": 0.9174228735297978, "compression/movement_sparsity/model_sparsity": 0.8859065755011739, "compression_loss": 104.07112121582031, "distillation_loss": 6.458298683166504, "epoch": 4.25, "learning_rate": 3.981214777708203e-05, "loss": 109.1093, "step": 5029, "task_loss": 3.5163261890411377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.984427727739474, "compression/movement_sparsity/importance_threshold": -0.0001010663030602205, "compression/movement_sparsity/linear_layer_sparsity": 0.9174192605070042, "compression/movement_sparsity/model_sparsity": 0.8859030865968281, "compression_loss": 104.07576751708984, "distillation_loss": 4.109910011291504, "epoch": 4.25, "learning_rate": 3.9809016906700064e-05, "loss": 108.4735, "step": 5030, "task_loss": 1.9316366910934448 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9844803367228281, "compression/movement_sparsity/importance_threshold": -0.00010072486313633493, "compression/movement_sparsity/linear_layer_sparsity": 0.9174607327620417, "compression/movement_sparsity/model_sparsity": 0.8859431341523216, "compression_loss": 104.08047485351562, "distillation_loss": 3.532179594039917, "epoch": 4.25, "learning_rate": 3.9805886036318096e-05, "loss": 107.8277, "step": 5031, "task_loss": 1.658061146736145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9845328270839477, "compression/movement_sparsity/importance_threshold": -0.00010038419308793737, "compression/movement_sparsity/linear_layer_sparsity": 0.9174726330813423, "compression/movement_sparsity/model_sparsity": 0.8859546256590445, "compression_loss": 104.0851821899414, "distillation_loss": 5.00424861907959, "epoch": 4.25, "learning_rate": 3.9802755165936134e-05, "loss": 108.7641, "step": 5032, "task_loss": 2.905059576034546 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9845851989567178, "compression/movement_sparsity/importance_threshold": -0.00010004429204609443, "compression/movement_sparsity/linear_layer_sparsity": 0.917447329997619, "compression/movement_sparsity/model_sparsity": 0.8859301918140885, "compression_loss": 104.08989715576172, "distillation_loss": 3.9594838619232178, "epoch": 4.25, "learning_rate": 3.9799624295554166e-05, "loss": 107.4876, "step": 5033, "task_loss": 2.8640480041503906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9846374524750237, "compression/movement_sparsity/importance_threshold": -9.970515914187184e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174743501614818, "compression/movement_sparsity/model_sparsity": 0.8859562837521989, "compression_loss": 104.09455108642578, "distillation_loss": 4.6003313064575195, "epoch": 4.26, "learning_rate": 3.97964934251722e-05, "loss": 108.484, "step": 5034, "task_loss": 2.2449417114257812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9846895877727504, "compression/movement_sparsity/importance_threshold": -9.936679350633533e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175219037420137, "compression/movement_sparsity/model_sparsity": 0.8860022037209476, "compression_loss": 104.0992202758789, "distillation_loss": 5.187588691711426, "epoch": 4.26, "learning_rate": 3.979336255479023e-05, "loss": 107.7905, "step": 5035, "task_loss": 2.2710602283477783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.984741604983783, "compression/movement_sparsity/importance_threshold": -9.902919427055065e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174951459098388, "compression/movement_sparsity/model_sparsity": 0.8859763651026247, "compression_loss": 104.10395812988281, "distillation_loss": 4.3893632888793945, "epoch": 4.26, "learning_rate": 3.979023168440827e-05, "loss": 108.8009, "step": 5036, "task_loss": 2.8516039848327637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9847935042420068, "compression/movement_sparsity/importance_threshold": -9.869236056558525e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175476837924424, "compression/movement_sparsity/model_sparsity": 0.8860270981473354, "compression_loss": 104.10865783691406, "distillation_loss": 4.754683494567871, "epoch": 4.26, "learning_rate": 3.97871008140263e-05, "loss": 108.6595, "step": 5037, "task_loss": 2.6667327880859375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9848452856813068, "compression/movement_sparsity/importance_threshold": -9.835629152250489e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175121140003846, "compression/movement_sparsity/model_sparsity": 0.8859927502870603, "compression_loss": 104.1132583618164, "distillation_loss": 4.137458801269531, "epoch": 4.26, "learning_rate": 3.978396994364433e-05, "loss": 108.7001, "step": 5038, "task_loss": 1.9446316957473755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9848969494355683, "compression/movement_sparsity/importance_threshold": -9.802098627237441e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174635468656037, "compression/movement_sparsity/model_sparsity": 0.8859458515827691, "compression_loss": 104.11792755126953, "distillation_loss": 4.700922012329102, "epoch": 4.26, "learning_rate": 3.978083907326237e-05, "loss": 108.2979, "step": 5039, "task_loss": 2.640299081802368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9849484956386761, "compression/movement_sparsity/importance_threshold": -9.76864439462613e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174635587897714, "compression/movement_sparsity/model_sparsity": 0.8859458630973048, "compression_loss": 104.12250518798828, "distillation_loss": 4.564892768859863, "epoch": 4.26, "learning_rate": 3.97777082028804e-05, "loss": 108.3729, "step": 5040, "task_loss": 1.5986037254333496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9849999244245158, "compression/movement_sparsity/importance_threshold": -9.735266367522956e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174880391059278, "compression/movement_sparsity/model_sparsity": 0.8859695024392911, "compression_loss": 104.12715148925781, "distillation_loss": 4.59499454498291, "epoch": 4.26, "learning_rate": 3.9774577332498434e-05, "loss": 108.4984, "step": 5041, "task_loss": 2.0734610557556152 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9850512359269723, "compression/movement_sparsity/importance_threshold": -9.701964459034752e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175072012433186, "compression/movement_sparsity/model_sparsity": 0.8859880062983129, "compression_loss": 104.13180541992188, "distillation_loss": 3.9584884643554688, "epoch": 4.26, "learning_rate": 3.9771446462116466e-05, "loss": 107.8022, "step": 5042, "task_loss": 1.5910801887512207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9851024302799307, "compression/movement_sparsity/importance_threshold": -9.668738582267918e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174326632714269, "compression/movement_sparsity/model_sparsity": 0.8859160289350613, "compression_loss": 104.13646697998047, "distillation_loss": 3.403926372528076, "epoch": 4.26, "learning_rate": 3.9768315591734504e-05, "loss": 108.2899, "step": 5043, "task_loss": 1.6858471632003784 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9851535076172762, "compression/movement_sparsity/importance_threshold": -9.635588650329287e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174275120310081, "compression/movement_sparsity/model_sparsity": 0.8859110546555979, "compression_loss": 104.14115905761719, "distillation_loss": 5.166037559509277, "epoch": 4.26, "learning_rate": 3.9765184721352536e-05, "loss": 107.9853, "step": 5044, "task_loss": 1.783414602279663 }, { "compression/movement_sparsity/importance_regularization_factor": 0.985204468072894, "compression/movement_sparsity/importance_threshold": -9.602514576325261e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174837941022495, "compression/movement_sparsity/model_sparsity": 0.8859654032645482, "compression_loss": 104.14583587646484, "distillation_loss": 4.999764442443848, "epoch": 4.26, "learning_rate": 3.976205385097057e-05, "loss": 107.8902, "step": 5045, "task_loss": 2.5062742233276367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.985255311780669, "compression/movement_sparsity/importance_threshold": -9.569516273362586e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175154646914903, "compression/movement_sparsity/model_sparsity": 0.8859959858716185, "compression_loss": 104.15040588378906, "distillation_loss": 5.315682411193848, "epoch": 4.27, "learning_rate": 3.9758922980588606e-05, "loss": 107.8609, "step": 5046, "task_loss": 3.308912754058838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9853060388744866, "compression/movement_sparsity/importance_threshold": -9.536593654547836e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9174842710689549, "compression/movement_sparsity/model_sparsity": 0.88596586384598, "compression_loss": 104.15495300292969, "distillation_loss": 3.9123940467834473, "epoch": 4.27, "learning_rate": 3.975579211020664e-05, "loss": 107.6444, "step": 5047, "task_loss": 2.040376663208008 }, { "compression/movement_sparsity/importance_regularization_factor": 0.985356649488232, "compression/movement_sparsity/importance_threshold": -9.503746632987496e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175080120867178, "compression/movement_sparsity/model_sparsity": 0.885988789286747, "compression_loss": 104.15955352783203, "distillation_loss": 4.720516204833984, "epoch": 4.27, "learning_rate": 3.975266123982468e-05, "loss": 108.6369, "step": 5048, "task_loss": 3.330561399459839 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9854071437557902, "compression/movement_sparsity/importance_threshold": -9.470975121788142e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9175350560989161, "compression/movement_sparsity/model_sparsity": 0.8860149042539289, "compression_loss": 104.16403198242188, "distillation_loss": 4.9318132400512695, "epoch": 4.27, "learning_rate": 3.974953036944271e-05, "loss": 108.7753, "step": 5049, "task_loss": 1.9558342695236206 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9854575218110462, "compression/movement_sparsity/importance_threshold": -9.438279034056692e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9176708127474503, "compression/movement_sparsity/model_sparsity": 0.8861459972439498, "compression_loss": 104.16854858398438, "distillation_loss": 5.18118143081665, "epoch": 4.27, "learning_rate": 3.974639949906074e-05, "loss": 108.2573, "step": 5050, "task_loss": 1.6404415369033813 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9855077837878853, "compression/movement_sparsity/importance_threshold": -9.405658282899548e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.917701112057413, "compression/movement_sparsity/model_sparsity": 0.8861752556794037, "compression_loss": 104.17300415039062, "distillation_loss": 3.054823160171509, "epoch": 4.27, "learning_rate": 3.974326862867878e-05, "loss": 107.8925, "step": 5051, "task_loss": 1.5822772979736328 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9855579298201926, "compression/movement_sparsity/importance_threshold": -9.373112781423196e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9176668181512924, "compression/movement_sparsity/model_sparsity": 0.8861421398744587, "compression_loss": 104.17745971679688, "distillation_loss": 5.718256950378418, "epoch": 4.27, "learning_rate": 3.974013775829681e-05, "loss": 108.5828, "step": 5052, "task_loss": 3.199431896209717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9856079600418534, "compression/movement_sparsity/importance_threshold": -9.34064244273447e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9177359306269097, "compression/movement_sparsity/model_sparsity": 0.8862088781239238, "compression_loss": 104.18195343017578, "distillation_loss": 5.491251468658447, "epoch": 4.27, "learning_rate": 3.973700688791484e-05, "loss": 108.7505, "step": 5053, "task_loss": 2.5815393924713135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9856578745867526, "compression/movement_sparsity/importance_threshold": -9.30824717993977e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9177675773678153, "compression/movement_sparsity/model_sparsity": 0.8862394377019225, "compression_loss": 104.18641662597656, "distillation_loss": 3.4561750888824463, "epoch": 4.27, "learning_rate": 3.973387601753288e-05, "loss": 108.1118, "step": 5054, "task_loss": 1.6874794960021973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9857076735887755, "compression/movement_sparsity/importance_threshold": -9.275926906145756e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9177634277574781, "compression/movement_sparsity/model_sparsity": 0.886235430643466, "compression_loss": 104.19083404541016, "distillation_loss": 4.959413528442383, "epoch": 4.27, "learning_rate": 3.973074514715091e-05, "loss": 108.8327, "step": 5055, "task_loss": 2.0234594345092773 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9857573571818071, "compression/movement_sparsity/importance_threshold": -9.243681534459175e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9178489121152595, "compression/movement_sparsity/model_sparsity": 0.886317978350577, "compression_loss": 104.19525146484375, "distillation_loss": 3.9742801189422607, "epoch": 4.27, "learning_rate": 3.9727614276768944e-05, "loss": 108.5353, "step": 5056, "task_loss": 2.481740951538086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9858069254997327, "compression/movement_sparsity/importance_threshold": -9.211510977986514e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9178362844217332, "compression/movement_sparsity/model_sparsity": 0.8863057844571707, "compression_loss": 104.19966125488281, "distillation_loss": 5.121835231781006, "epoch": 4.27, "learning_rate": 3.9724483406386976e-05, "loss": 108.5877, "step": 5057, "task_loss": 2.950498104095459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9858563786764374, "compression/movement_sparsity/importance_threshold": -9.179415149834346e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9179445320155316, "compression/movement_sparsity/model_sparsity": 0.8864103134131135, "compression_loss": 104.20401763916016, "distillation_loss": 3.589249610900879, "epoch": 4.28, "learning_rate": 3.9721352536005015e-05, "loss": 108.271, "step": 5058, "task_loss": 1.2893792390823364 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9859057168458063, "compression/movement_sparsity/importance_threshold": -9.147393963109332e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9179079248208895, "compression/movement_sparsity/model_sparsity": 0.8863749637882242, "compression_loss": 104.20845031738281, "distillation_loss": 5.425156593322754, "epoch": 4.28, "learning_rate": 3.9718221665623047e-05, "loss": 108.5011, "step": 5059, "task_loss": 3.690537452697754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9859549401417245, "compression/movement_sparsity/importance_threshold": -9.115447330918045e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9178764211699955, "compression/movement_sparsity/model_sparsity": 0.8863445423846551, "compression_loss": 104.21281433105469, "distillation_loss": 4.282351493835449, "epoch": 4.28, "learning_rate": 3.971509079524108e-05, "loss": 108.1632, "step": 5060, "task_loss": 1.8809864521026611 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9860040486980772, "compression/movement_sparsity/importance_threshold": -9.083575166367146e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9180074677723137, "compression/movement_sparsity/model_sparsity": 0.886471087133037, "compression_loss": 104.21723937988281, "distillation_loss": 4.14516019821167, "epoch": 4.28, "learning_rate": 3.971195992485911e-05, "loss": 108.6345, "step": 5061, "task_loss": 2.067247152328491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9860530426487495, "compression/movement_sparsity/importance_threshold": -9.051777382563207e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9179175595483392, "compression/movement_sparsity/model_sparsity": 0.8863842675331463, "compression_loss": 104.22159576416016, "distillation_loss": 4.084898471832275, "epoch": 4.28, "learning_rate": 3.970882905447715e-05, "loss": 108.5845, "step": 5062, "task_loss": 2.281670093536377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9861019221276266, "compression/movement_sparsity/importance_threshold": -9.020053892612803e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9179651846738769, "compression/movement_sparsity/model_sparsity": 0.8864302565891097, "compression_loss": 104.22602844238281, "distillation_loss": 3.6763787269592285, "epoch": 4.28, "learning_rate": 3.970569818409518e-05, "loss": 108.622, "step": 5063, "task_loss": 2.013502836227417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9861506872685937, "compression/movement_sparsity/importance_threshold": -8.988404609622507e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9180098168333379, "compression/movement_sparsity/model_sparsity": 0.8864733554965886, "compression_loss": 104.23035430908203, "distillation_loss": 5.560120105743408, "epoch": 4.28, "learning_rate": 3.970256731371321e-05, "loss": 108.7807, "step": 5064, "task_loss": 3.63095760345459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9861993382055357, "compression/movement_sparsity/importance_threshold": -8.95682944669898e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9180860718853693, "compression/movement_sparsity/model_sparsity": 0.8865469909529947, "compression_loss": 104.23475646972656, "distillation_loss": 5.780068874359131, "epoch": 4.28, "learning_rate": 3.969943644333125e-05, "loss": 108.8768, "step": 5065, "task_loss": 2.7239954471588135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9862478750723379, "compression/movement_sparsity/importance_threshold": -8.925328316948967e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.918116240029488, "compression/movement_sparsity/model_sparsity": 0.8865761227285549, "compression_loss": 104.23910522460938, "distillation_loss": 4.49990701675415, "epoch": 4.28, "learning_rate": 3.969630557294928e-05, "loss": 108.4306, "step": 5066, "task_loss": 3.3361992835998535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9862962980028855, "compression/movement_sparsity/importance_threshold": -8.893901133478783e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9182045504149992, "compression/movement_sparsity/model_sparsity": 0.8866613993806493, "compression_loss": 104.24349975585938, "distillation_loss": 5.493678092956543, "epoch": 4.28, "learning_rate": 3.9693174702567314e-05, "loss": 108.4799, "step": 5067, "task_loss": 2.815937042236328 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9863446071310636, "compression/movement_sparsity/importance_threshold": -8.862547809395174e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9182413006996529, "compression/movement_sparsity/model_sparsity": 0.886696887179968, "compression_loss": 104.24786376953125, "distillation_loss": 5.124921798706055, "epoch": 4.28, "learning_rate": 3.9690043832185346e-05, "loss": 108.559, "step": 5068, "task_loss": 3.1392881870269775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9863928025907573, "compression/movement_sparsity/importance_threshold": -8.831268257804715e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9182994071685425, "compression/movement_sparsity/model_sparsity": 0.8867529975128948, "compression_loss": 104.25225067138672, "distillation_loss": 3.531829357147217, "epoch": 4.28, "learning_rate": 3.9686912961803385e-05, "loss": 108.1634, "step": 5069, "task_loss": 1.6838289499282837 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9864408845158518, "compression/movement_sparsity/importance_threshold": -8.800062391814151e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9183497748526364, "compression/movement_sparsity/model_sparsity": 0.886801634912091, "compression_loss": 104.25662231445312, "distillation_loss": 4.779000282287598, "epoch": 4.29, "learning_rate": 3.9683782091421416e-05, "loss": 108.3984, "step": 5070, "task_loss": 2.8112823963165283 }, { "compression/movement_sparsity/importance_regularization_factor": 0.986488853040232, "compression/movement_sparsity/importance_threshold": -8.768930124529971e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.91837290773785, "compression/movement_sparsity/model_sparsity": 0.8868239731115324, "compression_loss": 104.26090240478516, "distillation_loss": 4.0896077156066895, "epoch": 4.29, "learning_rate": 3.968065122103945e-05, "loss": 108.7024, "step": 5071, "task_loss": 2.6036698818206787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9865367082977834, "compression/movement_sparsity/importance_threshold": -8.737871369058747e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9183422626270258, "compression/movement_sparsity/model_sparsity": 0.8867943807545404, "compression_loss": 104.26522064208984, "distillation_loss": 7.411805629730225, "epoch": 4.29, "learning_rate": 3.967752035065748e-05, "loss": 109.2056, "step": 5072, "task_loss": 3.96777081489563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.986584450422391, "compression/movement_sparsity/importance_threshold": -8.706886038507052e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9183389119359201, "compression/movement_sparsity/model_sparsity": 0.8867911451699821, "compression_loss": 104.26956176757812, "distillation_loss": 4.931956768035889, "epoch": 4.29, "learning_rate": 3.967438948027552e-05, "loss": 108.3339, "step": 5073, "task_loss": 2.742826223373413 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9866320795479399, "compression/movement_sparsity/importance_threshold": -8.675974045981635e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184445362128386, "compression/movement_sparsity/model_sparsity": 0.8868931409280502, "compression_loss": 104.27384185791016, "distillation_loss": 5.282876968383789, "epoch": 4.29, "learning_rate": 3.967125860989355e-05, "loss": 108.5676, "step": 5074, "task_loss": 3.7570106983184814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9866795958083151, "compression/movement_sparsity/importance_threshold": -8.645135304589067e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.918458439792302, "compression/movement_sparsity/model_sparsity": 0.8869065668767866, "compression_loss": 104.27810668945312, "distillation_loss": 4.307097434997559, "epoch": 4.29, "learning_rate": 3.966812773951158e-05, "loss": 109.0045, "step": 5075, "task_loss": 2.4788126945495605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.986726999337402, "compression/movement_sparsity/importance_threshold": -8.614369727435751e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184363920063433, "compression/movement_sparsity/model_sparsity": 0.8868852765001025, "compression_loss": 104.28236389160156, "distillation_loss": 4.750248432159424, "epoch": 4.29, "learning_rate": 3.966499686912962e-05, "loss": 107.8227, "step": 5076, "task_loss": 1.8645951747894287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9867742902690857, "compression/movement_sparsity/importance_threshold": -8.583677227628605e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184524180876459, "compression/movement_sparsity/model_sparsity": 0.8869007520362103, "compression_loss": 104.28660583496094, "distillation_loss": 4.815167427062988, "epoch": 4.29, "learning_rate": 3.966186599874765e-05, "loss": 108.3873, "step": 5077, "task_loss": 2.5902302265167236 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9868214687372513, "compression/movement_sparsity/importance_threshold": -8.553057718273943e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184281762548422, "compression/movement_sparsity/model_sparsity": 0.88687734298494, "compression_loss": 104.29078674316406, "distillation_loss": 3.261777877807617, "epoch": 4.29, "learning_rate": 3.9658735128365684e-05, "loss": 108.3348, "step": 5078, "task_loss": 1.3576774597167969 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9868685348757839, "compression/movement_sparsity/importance_threshold": -8.5225111124786e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184007625934473, "compression/movement_sparsity/model_sparsity": 0.8868508710671484, "compression_loss": 104.29499053955078, "distillation_loss": 3.8613195419311523, "epoch": 4.29, "learning_rate": 3.965560425798372e-05, "loss": 108.9398, "step": 5079, "task_loss": 2.4497485160827637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9869154888185686, "compression/movement_sparsity/importance_threshold": -8.492037323348973e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9183799191484199, "compression/movement_sparsity/model_sparsity": 0.8868307436585795, "compression_loss": 104.2992172241211, "distillation_loss": 3.258310317993164, "epoch": 4.29, "learning_rate": 3.9652473387601755e-05, "loss": 108.2256, "step": 5080, "task_loss": 1.3851344585418701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9869623306994906, "compression/movement_sparsity/importance_threshold": -8.4616362639919e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9183625217878391, "compression/movement_sparsity/model_sparsity": 0.8868139439508553, "compression_loss": 104.3034439086914, "distillation_loss": 5.044925689697266, "epoch": 4.29, "learning_rate": 3.9649342517219786e-05, "loss": 108.6804, "step": 5081, "task_loss": 2.8143155574798584 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9870090606524351, "compression/movement_sparsity/importance_threshold": -8.431307847513777e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9183735158703994, "compression/movement_sparsity/model_sparsity": 0.8868245603528578, "compression_loss": 104.30757904052734, "distillation_loss": 4.711147785186768, "epoch": 4.3, "learning_rate": 3.9646211646837825e-05, "loss": 108.8979, "step": 5082, "task_loss": 2.4021050930023193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9870556788112871, "compression/movement_sparsity/importance_threshold": -8.401051987021353e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184330890119081, "compression/movement_sparsity/model_sparsity": 0.8868820869736874, "compression_loss": 104.31172180175781, "distillation_loss": 5.489914894104004, "epoch": 4.3, "learning_rate": 3.964308077645586e-05, "loss": 109.1895, "step": 5083, "task_loss": 2.72745680809021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9871021853099319, "compression/movement_sparsity/importance_threshold": -8.370868595621115e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9184318846709769, "compression/movement_sparsity/model_sparsity": 0.8868809240055722, "compression_loss": 104.3159408569336, "distillation_loss": 4.344395637512207, "epoch": 4.3, "learning_rate": 3.9639949906073895e-05, "loss": 108.3105, "step": 5084, "task_loss": 1.8224296569824219 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9871485802822545, "compression/movement_sparsity/importance_threshold": -8.340757586419723e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9185307360206783, "compression/movement_sparsity/model_sparsity": 0.8869763795073089, "compression_loss": 104.32005310058594, "distillation_loss": 3.3406238555908203, "epoch": 4.3, "learning_rate": 3.963681903569193e-05, "loss": 108.0319, "step": 5085, "task_loss": 1.0029611587524414 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9871948638621402, "compression/movement_sparsity/importance_threshold": -8.310718872523836e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9185624185340867, "compression/movement_sparsity/model_sparsity": 0.887006973628915, "compression_loss": 104.32413482666016, "distillation_loss": 3.923398971557617, "epoch": 4.3, "learning_rate": 3.963368816530996e-05, "loss": 107.9987, "step": 5086, "task_loss": 2.284975051879883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9872410361834739, "compression/movement_sparsity/importance_threshold": -8.280752367039942e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9186249965658397, "compression/movement_sparsity/model_sparsity": 0.8870674019127648, "compression_loss": 104.32821655273438, "distillation_loss": 4.097253799438477, "epoch": 4.3, "learning_rate": 3.9630557294928e-05, "loss": 107.9638, "step": 5087, "task_loss": 2.8332343101501465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.987287097380141, "compression/movement_sparsity/importance_threshold": -8.250857983074702e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9185511978923414, "compression/movement_sparsity/model_sparsity": 0.8869961384507323, "compression_loss": 104.33219146728516, "distillation_loss": 5.056276798248291, "epoch": 4.3, "learning_rate": 3.962742642454603e-05, "loss": 108.021, "step": 5088, "task_loss": 2.4497382640838623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9873330475860265, "compression/movement_sparsity/importance_threshold": -8.221035633734775e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9185835839316404, "compression/movement_sparsity/model_sparsity": 0.8870274119299503, "compression_loss": 104.33619689941406, "distillation_loss": 4.4374589920043945, "epoch": 4.3, "learning_rate": 3.962429555416406e-05, "loss": 108.316, "step": 5089, "task_loss": 2.456451654434204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9873788869350155, "compression/movement_sparsity/importance_threshold": -8.191285232126734e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9186704991895381, "compression/movement_sparsity/model_sparsity": 0.8871113413813567, "compression_loss": 104.34025573730469, "distillation_loss": 5.933607578277588, "epoch": 4.3, "learning_rate": 3.962116468378209e-05, "loss": 108.5573, "step": 5090, "task_loss": 2.928389310836792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9874246155609934, "compression/movement_sparsity/importance_threshold": -8.161606691357067e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9186766997567088, "compression/movement_sparsity/model_sparsity": 0.8871173289399699, "compression_loss": 104.34426879882812, "distillation_loss": 5.020070552825928, "epoch": 4.3, "learning_rate": 3.961803381340013e-05, "loss": 108.8707, "step": 5091, "task_loss": 2.5833957195281982 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9874702335978449, "compression/movement_sparsity/importance_threshold": -8.131999924532521e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9186893036018999, "compression/movement_sparsity/model_sparsity": 0.8871294998043047, "compression_loss": 104.3482666015625, "distillation_loss": 4.102550506591797, "epoch": 4.3, "learning_rate": 3.961490294301816e-05, "loss": 108.7858, "step": 5092, "task_loss": 1.5360348224639893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9875157411794555, "compression/movement_sparsity/importance_threshold": -8.102464844759669e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9187139985530738, "compression/movement_sparsity/model_sparsity": 0.8871533464079352, "compression_loss": 104.35230255126953, "distillation_loss": 3.306535243988037, "epoch": 4.3, "learning_rate": 3.9611772072636195e-05, "loss": 108.44, "step": 5093, "task_loss": 1.9958795309066772 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9875611384397103, "compression/movement_sparsity/importance_threshold": -8.073001365145084e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9187294641984975, "compression/movement_sparsity/model_sparsity": 0.8871682807608607, "compression_loss": 104.35623168945312, "distillation_loss": 4.1696929931640625, "epoch": 4.31, "learning_rate": 3.960864120225423e-05, "loss": 107.9134, "step": 5094, "task_loss": 2.3206300735473633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9876064255124943, "compression/movement_sparsity/importance_threshold": -8.04360939879534e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9187715445860845, "compression/movement_sparsity/model_sparsity": 0.8872089155576797, "compression_loss": 104.36017608642578, "distillation_loss": 5.324941635131836, "epoch": 4.31, "learning_rate": 3.9605510331872265e-05, "loss": 108.5812, "step": 5095, "task_loss": 2.798781156539917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9876516025316927, "compression/movement_sparsity/importance_threshold": -8.014288858817099e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9188389399815624, "compression/movement_sparsity/model_sparsity": 0.8872739957139903, "compression_loss": 104.36417388916016, "distillation_loss": 4.621357440948486, "epoch": 4.31, "learning_rate": 3.96023794614903e-05, "loss": 108.5876, "step": 5096, "task_loss": 3.0784425735473633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9876966696311907, "compression/movement_sparsity/importance_threshold": -7.985039658316932e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9189412612640457, "compression/movement_sparsity/model_sparsity": 0.8873728019456434, "compression_loss": 104.36817169189453, "distillation_loss": 5.437366962432861, "epoch": 4.31, "learning_rate": 3.959924859110833e-05, "loss": 108.6834, "step": 5097, "task_loss": 2.530684232711792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9877416269448734, "compression/movement_sparsity/importance_threshold": -7.9558617104015e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9189176633362943, "compression/movement_sparsity/model_sparsity": 0.887350014679306, "compression_loss": 104.37213134765625, "distillation_loss": 3.222806692123413, "epoch": 4.31, "learning_rate": 3.959611772072636e-05, "loss": 108.8944, "step": 5098, "task_loss": 1.900678277015686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9877864746066259, "compression/movement_sparsity/importance_threshold": -7.926754928177376e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9189263799028362, "compression/movement_sparsity/model_sparsity": 0.8873584318049718, "compression_loss": 104.37610626220703, "distillation_loss": 4.31981086730957, "epoch": 4.31, "learning_rate": 3.95929868503444e-05, "loss": 108.7777, "step": 5099, "task_loss": 2.2582712173461914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9878312127503334, "compression/movement_sparsity/importance_threshold": -7.897719224751135e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.918962307419923, "compression/movement_sparsity/model_sparsity": 0.8873931251013207, "compression_loss": 104.38004302978516, "distillation_loss": 4.836058616638184, "epoch": 4.31, "learning_rate": 3.958985597996243e-05, "loss": 108.1034, "step": 5100, "task_loss": 2.931415319442749 }, { "compression/movement_sparsity/importance_regularization_factor": 0.987875841509881, "compression/movement_sparsity/importance_threshold": -7.868754513229437e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9189765329519126, "compression/movement_sparsity/model_sparsity": 0.8874068619425236, "compression_loss": 104.38396453857422, "distillation_loss": 5.496615886688232, "epoch": 4.31, "learning_rate": 3.958672510958046e-05, "loss": 109.2934, "step": 5101, "task_loss": 3.3170619010925293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9879203610191538, "compression/movement_sparsity/importance_threshold": -7.839860706718854e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9190264117451334, "compression/movement_sparsity/model_sparsity": 0.8874550272457522, "compression_loss": 104.38797760009766, "distillation_loss": 3.605288505554199, "epoch": 4.31, "learning_rate": 3.95835942391985e-05, "loss": 108.7197, "step": 5102, "task_loss": 2.4410243034362793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9879647714120371, "compression/movement_sparsity/importance_threshold": -7.811037718325874e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9190673712609626, "compression/movement_sparsity/model_sparsity": 0.8874945796762064, "compression_loss": 104.39188385009766, "distillation_loss": 3.586153030395508, "epoch": 4.31, "learning_rate": 3.958046336881653e-05, "loss": 108.8585, "step": 5103, "task_loss": 3.0945591926574707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9880090728224158, "compression/movement_sparsity/importance_threshold": -7.782285461157418e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919077053685083, "compression/movement_sparsity/model_sparsity": 0.8875039294792716, "compression_loss": 104.39578247070312, "distillation_loss": 3.349139928817749, "epoch": 4.31, "learning_rate": 3.9577332498434565e-05, "loss": 108.8024, "step": 5104, "task_loss": 2.2933759689331055 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9880532653841753, "compression/movement_sparsity/importance_threshold": -7.753603848319711e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919020235026298, "compression/movement_sparsity/model_sparsity": 0.8874490627162106, "compression_loss": 104.39974212646484, "distillation_loss": 3.6889076232910156, "epoch": 4.32, "learning_rate": 3.9574201628052597e-05, "loss": 107.8584, "step": 5105, "task_loss": 2.9783239364624023 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9880973492312005, "compression/movement_sparsity/importance_threshold": -7.724992792919588e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9190187683536788, "compression/movement_sparsity/model_sparsity": 0.8874476464283079, "compression_loss": 104.40364074707031, "distillation_loss": 4.173440456390381, "epoch": 4.32, "learning_rate": 3.9571070757670635e-05, "loss": 108.2777, "step": 5106, "task_loss": 2.7344300746917725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9881413244973768, "compression/movement_sparsity/importance_threshold": -7.696452208063621e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9189823757940542, "compression/movement_sparsity/model_sparsity": 0.8874125040650629, "compression_loss": 104.4075698852539, "distillation_loss": 3.3982009887695312, "epoch": 4.32, "learning_rate": 3.956793988728867e-05, "loss": 108.1211, "step": 5107, "task_loss": 1.5726566314697266 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9881851913165891, "compression/movement_sparsity/importance_threshold": -7.667982006858386e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9189949915634129, "compression/movement_sparsity/model_sparsity": 0.8874246864439336, "compression_loss": 104.4113540649414, "distillation_loss": 4.42470121383667, "epoch": 4.32, "learning_rate": 3.95648090169067e-05, "loss": 108.1936, "step": 5108, "task_loss": 1.865139126777649 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9882289498227227, "compression/movement_sparsity/importance_threshold": -7.639582102410454e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9190147499091855, "compression/movement_sparsity/model_sparsity": 0.8874437660297451, "compression_loss": 104.415283203125, "distillation_loss": 5.245319366455078, "epoch": 4.32, "learning_rate": 3.956167814652473e-05, "loss": 108.2546, "step": 5109, "task_loss": 2.296196460723877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9882726001496626, "compression/movement_sparsity/importance_threshold": -7.611252407826574e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9190753723774463, "compression/movement_sparsity/model_sparsity": 0.8875023059297246, "compression_loss": 104.4190673828125, "distillation_loss": 4.1461052894592285, "epoch": 4.32, "learning_rate": 3.955854727614277e-05, "loss": 108.6377, "step": 5110, "task_loss": 2.9222381114959717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9883161424312941, "compression/movement_sparsity/importance_threshold": -7.582992836213057e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9191452122272895, "compression/movement_sparsity/model_sparsity": 0.8875697465658732, "compression_loss": 104.42290496826172, "distillation_loss": 5.153985977172852, "epoch": 4.32, "learning_rate": 3.95554164057608e-05, "loss": 108.4886, "step": 5111, "task_loss": 2.4943034648895264 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9883595768015022, "compression/movement_sparsity/importance_threshold": -7.554803300676826e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919146965079932, "compression/movement_sparsity/model_sparsity": 0.8875714392026349, "compression_loss": 104.4267349243164, "distillation_loss": 4.280019760131836, "epoch": 4.32, "learning_rate": 3.955228553537883e-05, "loss": 108.5688, "step": 5112, "task_loss": 2.3057308197021484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9884029033941721, "compression/movement_sparsity/importance_threshold": -7.52668371432428e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9192289079599256, "compression/movement_sparsity/model_sparsity": 0.887650567092615, "compression_loss": 104.43049621582031, "distillation_loss": 3.253291130065918, "epoch": 4.32, "learning_rate": 3.954915466499687e-05, "loss": 107.9213, "step": 5113, "task_loss": 1.5135904550552368 }, { "compression/movement_sparsity/importance_regularization_factor": 0.988446122343189, "compression/movement_sparsity/importance_threshold": -7.498633990262165e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9192445763161992, "compression/movement_sparsity/model_sparsity": 0.8876656971926491, "compression_loss": 104.43431091308594, "distillation_loss": 3.9799773693084717, "epoch": 4.32, "learning_rate": 3.95460237946149e-05, "loss": 108.6788, "step": 5114, "task_loss": 1.6635079383850098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9884892337824379, "compression/movement_sparsity/importance_threshold": -7.470654041597056e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9193067370020849, "compression/movement_sparsity/model_sparsity": 0.8877257224677461, "compression_loss": 104.43810272216797, "distillation_loss": 3.292564868927002, "epoch": 4.32, "learning_rate": 3.954289292423294e-05, "loss": 108.6243, "step": 5115, "task_loss": 1.8897706270217896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9885322378458041, "compression/movement_sparsity/importance_threshold": -7.442743781435352e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9192945505027611, "compression/movement_sparsity/model_sparsity": 0.887713954612164, "compression_loss": 104.44187927246094, "distillation_loss": 4.925112247467041, "epoch": 4.32, "learning_rate": 3.953976205385097e-05, "loss": 109.1091, "step": 5116, "task_loss": 2.706157684326172 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9885751346671726, "compression/movement_sparsity/importance_threshold": -7.414903122883974e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9193256487319554, "compression/movement_sparsity/model_sparsity": 0.8877439845215161, "compression_loss": 104.44564056396484, "distillation_loss": 3.3297998905181885, "epoch": 4.33, "learning_rate": 3.9536631183469005e-05, "loss": 108.3936, "step": 5117, "task_loss": 2.0882773399353027 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9886179243804287, "compression/movement_sparsity/importance_threshold": -7.387131979049322e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9193684684179357, "compression/movement_sparsity/model_sparsity": 0.8877853332195543, "compression_loss": 104.44931030273438, "distillation_loss": 3.6479032039642334, "epoch": 4.33, "learning_rate": 3.9533500313087044e-05, "loss": 108.214, "step": 5118, "task_loss": 1.5543147325515747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9886606071194574, "compression/movement_sparsity/importance_threshold": -7.359430263038057e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919352251549951, "compression/movement_sparsity/model_sparsity": 0.8877696734508738, "compression_loss": 104.45307159423828, "distillation_loss": 2.8959057331085205, "epoch": 4.33, "learning_rate": 3.9530369442705075e-05, "loss": 108.0722, "step": 5119, "task_loss": 1.7567622661590576 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9887031830181439, "compression/movement_sparsity/importance_threshold": -7.331797887956751e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9194187168603533, "compression/movement_sparsity/model_sparsity": 0.8878338554733926, "compression_loss": 104.4566421508789, "distillation_loss": 2.9078142642974854, "epoch": 4.33, "learning_rate": 3.952723857232311e-05, "loss": 107.3543, "step": 5120, "task_loss": 1.899561882019043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9887456522103732, "compression/movement_sparsity/importance_threshold": -7.304234766912065e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9194233434373961, "compression/movement_sparsity/model_sparsity": 0.8878383231132808, "compression_loss": 104.46041107177734, "distillation_loss": 3.1991305351257324, "epoch": 4.33, "learning_rate": 3.9524107701941146e-05, "loss": 108.1783, "step": 5121, "task_loss": 1.9525432586669922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9887880148300308, "compression/movement_sparsity/importance_threshold": -7.276740813010486e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9194902737903361, "compression/movement_sparsity/model_sparsity": 0.8879029542026956, "compression_loss": 104.46406555175781, "distillation_loss": 4.073284149169922, "epoch": 4.33, "learning_rate": 3.952097683155918e-05, "loss": 108.8039, "step": 5122, "task_loss": 2.7161619663238525 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9888302710110014, "compression/movement_sparsity/importance_threshold": -7.249315939358847e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9195069557008587, "compression/movement_sparsity/model_sparsity": 0.8879190630382722, "compression_loss": 104.46774291992188, "distillation_loss": 3.2577919960021973, "epoch": 4.33, "learning_rate": 3.951784596117721e-05, "loss": 107.7952, "step": 5123, "task_loss": 2.5848278999328613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9888724208871704, "compression/movement_sparsity/importance_threshold": -7.221960059063549e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9195103660128026, "compression/movement_sparsity/model_sparsity": 0.8879223561955094, "compression_loss": 104.4714126586914, "distillation_loss": 3.1483242511749268, "epoch": 4.33, "learning_rate": 3.951471509079525e-05, "loss": 108.5843, "step": 5124, "task_loss": 2.686208963394165 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9889144645924229, "compression/movement_sparsity/importance_threshold": -7.194673085231165e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9195421200712168, "compression/movement_sparsity/model_sparsity": 0.8879530194043302, "compression_loss": 104.47503662109375, "distillation_loss": 4.472857475280762, "epoch": 4.33, "learning_rate": 3.951158422041328e-05, "loss": 109.2177, "step": 5125, "task_loss": 1.9991368055343628 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9889564022606441, "compression/movement_sparsity/importance_threshold": -7.167454930968528e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9196176239006871, "compression/movement_sparsity/model_sparsity": 0.8880259294449813, "compression_loss": 104.47862243652344, "distillation_loss": 5.288304328918457, "epoch": 4.33, "learning_rate": 3.950845335003131e-05, "loss": 109.5841, "step": 5126, "task_loss": 3.0233187675476074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.988998234025719, "compression/movement_sparsity/importance_threshold": -7.14030550938204e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9196337453753308, "compression/movement_sparsity/model_sparsity": 0.8880414970973756, "compression_loss": 104.4822998046875, "distillation_loss": 3.167375087738037, "epoch": 4.33, "learning_rate": 3.950532247964934e-05, "loss": 109.4557, "step": 5127, "task_loss": 1.8409682512283325 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9890399600215328, "compression/movement_sparsity/importance_threshold": -7.113224733578359e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919559076237595, "compression/movement_sparsity/model_sparsity": 0.8879693930742301, "compression_loss": 104.48591613769531, "distillation_loss": 5.013689994812012, "epoch": 4.33, "learning_rate": 3.950219160926738e-05, "loss": 108.4088, "step": 5128, "task_loss": 2.9031314849853516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9890815803819707, "compression/movement_sparsity/importance_threshold": -7.086212516664147e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919534321665583, "compression/movement_sparsity/model_sparsity": 0.8879454888979207, "compression_loss": 104.48957824707031, "distillation_loss": 4.366327285766602, "epoch": 4.34, "learning_rate": 3.9499060738885414e-05, "loss": 108.3311, "step": 5129, "task_loss": 2.360992193222046 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9891230952409178, "compression/movement_sparsity/importance_threshold": -7.05926877174589e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9195769386407135, "compression/movement_sparsity/model_sparsity": 0.8879866418488503, "compression_loss": 104.49317932128906, "distillation_loss": 3.920480251312256, "epoch": 4.34, "learning_rate": 3.9495929868503445e-05, "loss": 108.7955, "step": 5130, "task_loss": 2.3075594902038574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9891645047322591, "compression/movement_sparsity/importance_threshold": -7.032393411930335e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919655351967087, "compression/movement_sparsity/model_sparsity": 0.8880623614362352, "compression_loss": 104.4968032836914, "distillation_loss": 3.884077548980713, "epoch": 4.34, "learning_rate": 3.949279899812148e-05, "loss": 108.7307, "step": 5131, "task_loss": 2.2659761905670166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9892058089898801, "compression/movement_sparsity/importance_threshold": -7.005586350323882e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9196734767018935, "compression/movement_sparsity/model_sparsity": 0.8880798635306429, "compression_loss": 104.50041198730469, "distillation_loss": 6.435641288757324, "epoch": 4.34, "learning_rate": 3.9489668127739516e-05, "loss": 109.467, "step": 5132, "task_loss": 3.774122953414917 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9892470081476655, "compression/movement_sparsity/importance_threshold": -6.978847500033452e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919648638660708, "compression/movement_sparsity/model_sparsity": 0.8880558787525829, "compression_loss": 104.5040512084961, "distillation_loss": 4.023615837097168, "epoch": 4.34, "learning_rate": 3.948653725735755e-05, "loss": 108.0251, "step": 5133, "task_loss": 2.23866605758667 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9892881023395008, "compression/movement_sparsity/importance_threshold": -6.952176774165358e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9196604316624999, "compression/movement_sparsity/model_sparsity": 0.8880672666284837, "compression_loss": 104.50756072998047, "distillation_loss": 5.095547676086426, "epoch": 4.34, "learning_rate": 3.948340638697558e-05, "loss": 109.0302, "step": 5134, "task_loss": 2.1245813369750977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.989329091699271, "compression/movement_sparsity/importance_threshold": -6.92557408582626e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9196913271808445, "compression/movement_sparsity/model_sparsity": 0.8880971007907273, "compression_loss": 104.51107025146484, "distillation_loss": 4.028286457061768, "epoch": 4.34, "learning_rate": 3.948027551659361e-05, "loss": 108.3928, "step": 5135, "task_loss": 3.0291452407836914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9893699763608611, "compression/movement_sparsity/importance_threshold": -6.899039348122905e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9197451171010498, "compression/movement_sparsity/model_sparsity": 0.8881490428616965, "compression_loss": 104.51460266113281, "distillation_loss": 5.75105619430542, "epoch": 4.34, "learning_rate": 3.947714464621165e-05, "loss": 108.796, "step": 5136, "task_loss": 2.4988465309143066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9894107564581565, "compression/movement_sparsity/importance_threshold": -6.87257247416178e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9197718153123865, "compression/movement_sparsity/model_sparsity": 0.8881748239073405, "compression_loss": 104.51811218261719, "distillation_loss": 5.229672908782959, "epoch": 4.34, "learning_rate": 3.947401377582968e-05, "loss": 108.9153, "step": 5137, "task_loss": 2.3085880279541016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9894514321250422, "compression/movement_sparsity/importance_threshold": -6.846173377049458e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9197768950077994, "compression/movement_sparsity/model_sparsity": 0.888179729099589, "compression_loss": 104.52157592773438, "distillation_loss": 5.1273298263549805, "epoch": 4.34, "learning_rate": 3.947088290544771e-05, "loss": 109.19, "step": 5138, "task_loss": 2.937978982925415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9894920034954032, "compression/movement_sparsity/importance_threshold": -6.819841969892687e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198459120900757, "compression/movement_sparsity/model_sparsity": 0.8882463752327677, "compression_loss": 104.52507781982422, "distillation_loss": 4.203758239746094, "epoch": 4.34, "learning_rate": 3.946775203506575e-05, "loss": 108.6057, "step": 5139, "task_loss": 2.3907251358032227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.989532470703125, "compression/movement_sparsity/importance_threshold": -6.793578165797953e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198485831036262, "compression/movement_sparsity/model_sparsity": 0.8882489544887857, "compression_loss": 104.52854919433594, "distillation_loss": 4.24393892288208, "epoch": 4.34, "learning_rate": 3.9464621164683783e-05, "loss": 108.4101, "step": 5140, "task_loss": 2.6541402339935303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9895728338820925, "compression/movement_sparsity/importance_threshold": -6.76738187787183e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198680910418784, "compression/movement_sparsity/model_sparsity": 0.8882677922693456, "compression_loss": 104.53204345703125, "distillation_loss": 3.447514057159424, "epoch": 4.35, "learning_rate": 3.9461490294301815e-05, "loss": 108.1733, "step": 5141, "task_loss": 2.706134557723999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9896130931661908, "compression/movement_sparsity/importance_threshold": -6.741253019221065e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198736477039967, "compression/movement_sparsity/model_sparsity": 0.8882731580430259, "compression_loss": 104.53543090820312, "distillation_loss": 4.1322431564331055, "epoch": 4.35, "learning_rate": 3.945835942391985e-05, "loss": 108.3969, "step": 5142, "task_loss": 1.8942283391952515 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9896532486893052, "compression/movement_sparsity/importance_threshold": -6.715191502952144e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198532454531718, "compression/movement_sparsity/model_sparsity": 0.8882534566722814, "compression_loss": 104.53882598876953, "distillation_loss": 4.314369201660156, "epoch": 4.35, "learning_rate": 3.9455228553537886e-05, "loss": 108.4576, "step": 5143, "task_loss": 2.9120986461639404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9896933005853207, "compression/movement_sparsity/importance_threshold": -6.689197242171728e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198530785148249, "compression/movement_sparsity/model_sparsity": 0.8882532954687803, "compression_loss": 104.54229736328125, "distillation_loss": 4.21770715713501, "epoch": 4.35, "learning_rate": 3.945209768315592e-05, "loss": 108.5405, "step": 5144, "task_loss": 2.6078503131866455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9897332489881225, "compression/movement_sparsity/importance_threshold": -6.66327014998639e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198359673342674, "compression/movement_sparsity/model_sparsity": 0.8882367721099151, "compression_loss": 104.54568481445312, "distillation_loss": 3.819185495376587, "epoch": 4.35, "learning_rate": 3.944896681277395e-05, "loss": 108.5064, "step": 5145, "task_loss": 2.4702770709991455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9897730940315957, "compression/movement_sparsity/importance_threshold": -6.637410139502704e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198434557115428, "compression/movement_sparsity/model_sparsity": 0.8882440032383941, "compression_loss": 104.54901123046875, "distillation_loss": 4.44462776184082, "epoch": 4.35, "learning_rate": 3.944583594239199e-05, "loss": 108.7952, "step": 5146, "task_loss": 2.0530030727386475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9898128358496256, "compression/movement_sparsity/importance_threshold": -6.61161712382733e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198195954521033, "compression/movement_sparsity/model_sparsity": 0.8882209626522692, "compression_loss": 104.55240631103516, "distillation_loss": 3.332803726196289, "epoch": 4.35, "learning_rate": 3.944270507201002e-05, "loss": 108.5061, "step": 5147, "task_loss": 1.6693708896636963 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9898524745760973, "compression/movement_sparsity/importance_threshold": -6.585891016066842e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198177472061199, "compression/movement_sparsity/model_sparsity": 0.8882191778992211, "compression_loss": 104.55572509765625, "distillation_loss": 2.367044687271118, "epoch": 4.35, "learning_rate": 3.943957420162805e-05, "loss": 107.4678, "step": 5148, "task_loss": 1.3191876411437988 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9898920103448957, "compression/movement_sparsity/importance_threshold": -6.560231729327813e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9197364482311785, "compression/movement_sparsity/model_sparsity": 0.8881406717941739, "compression_loss": 104.55905151367188, "distillation_loss": 4.279050827026367, "epoch": 4.35, "learning_rate": 3.943644333124609e-05, "loss": 108.4733, "step": 5149, "task_loss": 2.1609551906585693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9899314432899061, "compression/movement_sparsity/importance_threshold": -6.53463917671699e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919787650607007, "compression/movement_sparsity/model_sparsity": 0.8881901152108757, "compression_loss": 104.56242370605469, "distillation_loss": 3.219327926635742, "epoch": 4.35, "learning_rate": 3.943331246086412e-05, "loss": 108.5386, "step": 5150, "task_loss": 2.036074638366699 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9899707735450137, "compression/movement_sparsity/importance_threshold": -6.50911327134086e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9197880918012095, "compression/movement_sparsity/model_sparsity": 0.8881905412487, "compression_loss": 104.56575775146484, "distillation_loss": 4.61373233795166, "epoch": 4.35, "learning_rate": 3.943018159048216e-05, "loss": 109.1616, "step": 5151, "task_loss": 2.4847023487091064 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9900100012441035, "compression/movement_sparsity/importance_threshold": -6.483653926305996e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198362654384583, "compression/movement_sparsity/model_sparsity": 0.88823705997331, "compression_loss": 104.56916046142578, "distillation_loss": 3.7984437942504883, "epoch": 4.35, "learning_rate": 3.942705072010019e-05, "loss": 109.0556, "step": 5152, "task_loss": 2.4796183109283447 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9900491265210609, "compression/movement_sparsity/importance_threshold": -6.458261054719059e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198448985358267, "compression/movement_sparsity/model_sparsity": 0.8882453964972252, "compression_loss": 104.57239532470703, "distillation_loss": 4.498291969299316, "epoch": 4.36, "learning_rate": 3.9423919849718224e-05, "loss": 108.6795, "step": 5153, "task_loss": 3.2730982303619385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9900881495097708, "compression/movement_sparsity/importance_threshold": -6.432934569686709e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9198779523285133, "compression/movement_sparsity/model_sparsity": 0.8882773147904477, "compression_loss": 104.57571411132812, "distillation_loss": 3.184936285018921, "epoch": 4.36, "learning_rate": 3.942078897933626e-05, "loss": 108.0844, "step": 5154, "task_loss": 2.130091667175293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9901270703441183, "compression/movement_sparsity/importance_threshold": -6.407674384315432e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9199171589917, "compression/movement_sparsity/model_sparsity": 0.8883151745841401, "compression_loss": 104.57901000976562, "distillation_loss": 3.707155704498291, "epoch": 4.36, "learning_rate": 3.9417658108954294e-05, "loss": 108.5523, "step": 5155, "task_loss": 1.3443821668624878 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9901658891579888, "compression/movement_sparsity/importance_threshold": -6.382480411711803e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.919918959541013, "compression/movement_sparsity/model_sparsity": 0.8883169132790452, "compression_loss": 104.58222961425781, "distillation_loss": 4.1894683837890625, "epoch": 4.36, "learning_rate": 3.9414527238572326e-05, "loss": 108.7367, "step": 5156, "task_loss": 2.5927040576934814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9902046060852672, "compression/movement_sparsity/importance_threshold": -6.357352564982655e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9199830280937206, "compression/movement_sparsity/model_sparsity": 0.8883787808798692, "compression_loss": 104.58544921875, "distillation_loss": 4.011154651641846, "epoch": 4.36, "learning_rate": 3.941139636819036e-05, "loss": 109.6233, "step": 5157, "task_loss": 2.371647357940674 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9902432212598387, "compression/movement_sparsity/importance_threshold": -6.332290757234387e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.920047752475648, "compression/movement_sparsity/model_sparsity": 0.888441281780162, "compression_loss": 104.588623046875, "distillation_loss": 3.227046489715576, "epoch": 4.36, "learning_rate": 3.9408265497808396e-05, "loss": 108.2147, "step": 5158, "task_loss": 1.0687627792358398 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9902817348155885, "compression/movement_sparsity/importance_threshold": -6.307294901573661e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9200514370434475, "compression/movement_sparsity/model_sparsity": 0.8884448397717225, "compression_loss": 104.59173583984375, "distillation_loss": 4.55789852142334, "epoch": 4.36, "learning_rate": 3.940513462742643e-05, "loss": 108.8373, "step": 5159, "task_loss": 2.756115674972534 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9903201468864018, "compression/movement_sparsity/importance_threshold": -6.282364911106963e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.92007755097057, "compression/movement_sparsity/model_sparsity": 0.8884700566051126, "compression_loss": 104.59489440917969, "distillation_loss": 2.762406587600708, "epoch": 4.36, "learning_rate": 3.940200375704446e-05, "loss": 108.0579, "step": 5160, "task_loss": 1.403152585029602 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9903584576061635, "compression/movement_sparsity/importance_threshold": -6.257500698941213e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9201663263986191, "compression/movement_sparsity/model_sparsity": 0.8885557823241029, "compression_loss": 104.5980453491211, "distillation_loss": 5.017704010009766, "epoch": 4.36, "learning_rate": 3.93988728866625e-05, "loss": 109.0174, "step": 5161, "task_loss": 2.3816614151000977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.990396667108759, "compression/movement_sparsity/importance_threshold": -6.232702178182725e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9201985454995711, "compression/movement_sparsity/model_sparsity": 0.8885868945998198, "compression_loss": 104.60125732421875, "distillation_loss": 7.2506608963012695, "epoch": 4.36, "learning_rate": 3.939574201628053e-05, "loss": 109.5533, "step": 5162, "task_loss": 3.221463203430176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9904347755280732, "compression/movement_sparsity/importance_threshold": -6.207969261938246e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9201462937969908, "compression/movement_sparsity/model_sparsity": 0.8885364379039681, "compression_loss": 104.60438537597656, "distillation_loss": 5.566883563995361, "epoch": 4.36, "learning_rate": 3.939261114589856e-05, "loss": 109.0527, "step": 5163, "task_loss": 2.26192307472229 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9904727829979916, "compression/movement_sparsity/importance_threshold": -6.183301863314263e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9201415718266069, "compression/movement_sparsity/model_sparsity": 0.8885318781477934, "compression_loss": 104.60755920410156, "distillation_loss": 4.019078731536865, "epoch": 4.36, "learning_rate": 3.9389480275516594e-05, "loss": 109.7608, "step": 5164, "task_loss": 2.3466291427612305 }, { "compression/movement_sparsity/importance_regularization_factor": 0.990510689652399, "compression/movement_sparsity/importance_threshold": -6.158699895417435e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9201749833443227, "compression/movement_sparsity/model_sparsity": 0.8885641418770898, "compression_loss": 104.6107406616211, "distillation_loss": 4.625138282775879, "epoch": 4.37, "learning_rate": 3.938634940513463e-05, "loss": 108.4193, "step": 5165, "task_loss": 2.1646554470062256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9905484956251805, "compression/movement_sparsity/importance_threshold": -6.134163271354423e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9202168967935627, "compression/movement_sparsity/model_sparsity": 0.8886046154704076, "compression_loss": 104.6138687133789, "distillation_loss": 4.115006446838379, "epoch": 4.37, "learning_rate": 3.9383218534752664e-05, "loss": 108.9411, "step": 5166, "task_loss": 3.180541753768921 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9905862010502215, "compression/movement_sparsity/importance_threshold": -6.109691904231801e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9202377879352608, "compression/movement_sparsity/model_sparsity": 0.8886247889371196, "compression_loss": 104.61700439453125, "distillation_loss": 5.0501604080200195, "epoch": 4.37, "learning_rate": 3.9380087664370696e-05, "loss": 109.2324, "step": 5167, "task_loss": 3.859081983566284 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9906238060614072, "compression/movement_sparsity/importance_threshold": -6.085285707156055e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9203505070919226, "compression/movement_sparsity/model_sparsity": 0.8887336358439853, "compression_loss": 104.62014770507812, "distillation_loss": 4.120265007019043, "epoch": 4.37, "learning_rate": 3.937695679398873e-05, "loss": 108.6656, "step": 5168, "task_loss": 2.4946391582489014 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9906613107926224, "compression/movement_sparsity/importance_threshold": -6.0609445932340196e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9203629916954373, "compression/movement_sparsity/model_sparsity": 0.8887456915629622, "compression_loss": 104.62321472167969, "distillation_loss": 4.550365447998047, "epoch": 4.37, "learning_rate": 3.9373825923606766e-05, "loss": 108.8907, "step": 5169, "task_loss": 2.7870941162109375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9906987153777524, "compression/movement_sparsity/importance_threshold": -6.036668475572181e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.920417425520695, "compression/movement_sparsity/model_sparsity": 0.8887982554188644, "compression_loss": 104.62638092041016, "distillation_loss": 5.107027053833008, "epoch": 4.37, "learning_rate": 3.93706950532248e-05, "loss": 109.4069, "step": 5170, "task_loss": 2.7315120697021484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9907360199506825, "compression/movement_sparsity/importance_threshold": -6.012457267277026e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9204911168766846, "compression/movement_sparsity/model_sparsity": 0.8888694152500746, "compression_loss": 104.62944030761719, "distillation_loss": 4.828217029571533, "epoch": 4.37, "learning_rate": 3.936756418284283e-05, "loss": 108.7269, "step": 5171, "task_loss": 2.7805323600769043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9907732246452976, "compression/movement_sparsity/importance_threshold": -5.988310881455301e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9204415481118223, "compression/movement_sparsity/model_sparsity": 0.8888215493247766, "compression_loss": 104.63251495361328, "distillation_loss": 5.018903732299805, "epoch": 4.37, "learning_rate": 3.936443331246086e-05, "loss": 108.6284, "step": 5172, "task_loss": 2.7396552562713623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.990810329595483, "compression/movement_sparsity/importance_threshold": -5.964229231213581e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9204558094163149, "compression/movement_sparsity/model_sparsity": 0.8888353207095869, "compression_loss": 104.63556671142578, "distillation_loss": 4.188074111938477, "epoch": 4.37, "learning_rate": 3.93613024420789e-05, "loss": 108.1926, "step": 5173, "task_loss": 2.7017264366149902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9908473349351238, "compression/movement_sparsity/importance_threshold": -5.9402122296584384e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9204972697471847, "compression/movement_sparsity/model_sparsity": 0.8888753567505445, "compression_loss": 104.63863372802734, "distillation_loss": 4.837446212768555, "epoch": 4.37, "learning_rate": 3.935817157169693e-05, "loss": 109.0056, "step": 5174, "task_loss": 2.7977612018585205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.990884240798105, "compression/movement_sparsity/importance_threshold": -5.916259789896534e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9205002627132614, "compression/movement_sparsity/model_sparsity": 0.8888782468990291, "compression_loss": 104.64171600341797, "distillation_loss": 2.960354804992676, "epoch": 4.37, "learning_rate": 3.9355040701314964e-05, "loss": 108.09, "step": 5175, "task_loss": 2.30739164352417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9909210473183119, "compression/movement_sparsity/importance_threshold": -5.8923718250344405e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.920496935870491, "compression/movement_sparsity/model_sparsity": 0.8888750343435423, "compression_loss": 104.64471435546875, "distillation_loss": 4.146744728088379, "epoch": 4.38, "learning_rate": 3.9351909830933e-05, "loss": 108.8711, "step": 5176, "task_loss": 3.1127123832702637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9909577546296297, "compression/movement_sparsity/importance_threshold": -5.868548248178732e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9205097901232024, "compression/movement_sparsity/model_sparsity": 0.8888874470131288, "compression_loss": 104.64767456054688, "distillation_loss": 4.926850318908691, "epoch": 4.38, "learning_rate": 3.9348778960551034e-05, "loss": 109.1265, "step": 5177, "task_loss": 2.743680000305176 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9909943628659433, "compression/movement_sparsity/importance_threshold": -5.844788972436069e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9204940979185936, "compression/movement_sparsity/model_sparsity": 0.8888722938840232, "compression_loss": 104.65065002441406, "distillation_loss": 3.9864513874053955, "epoch": 4.38, "learning_rate": 3.9345648090169066e-05, "loss": 109.0968, "step": 5178, "task_loss": 2.8968539237976074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.991030872161138, "compression/movement_sparsity/importance_threshold": -5.821093910913025e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9205202953148897, "compression/movement_sparsity/model_sparsity": 0.8888975913191638, "compression_loss": 104.65362548828125, "distillation_loss": 4.1019392013549805, "epoch": 4.38, "learning_rate": 3.93425172197871e-05, "loss": 109.0112, "step": 5179, "task_loss": 3.426274299621582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.991067282649099, "compression/movement_sparsity/importance_threshold": -5.79746297671626e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9205534921975879, "compression/movement_sparsity/model_sparsity": 0.8889296477868158, "compression_loss": 104.65653228759766, "distillation_loss": 4.083472728729248, "epoch": 4.38, "learning_rate": 3.9339386349405136e-05, "loss": 108.5775, "step": 5180, "task_loss": 2.424571990966797 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9911035944637113, "compression/movement_sparsity/importance_threshold": -5.77389608295226e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9205568428886935, "compression/movement_sparsity/model_sparsity": 0.8889328833713741, "compression_loss": 104.65950775146484, "distillation_loss": 3.8653903007507324, "epoch": 4.38, "learning_rate": 3.933625547902317e-05, "loss": 107.9973, "step": 5181, "task_loss": 1.8921763896942139 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9911398077388601, "compression/movement_sparsity/importance_threshold": -5.7503931427277735e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9206221277064999, "compression/movement_sparsity/model_sparsity": 0.8889959254548492, "compression_loss": 104.66242218017578, "distillation_loss": 2.7388858795166016, "epoch": 4.38, "learning_rate": 3.9333124608641206e-05, "loss": 108.8218, "step": 5182, "task_loss": 2.4657318592071533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9911759226084307, "compression/movement_sparsity/importance_threshold": -5.7269540691491996e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9206562904467767, "compression/movement_sparsity/model_sparsity": 0.8890289145999005, "compression_loss": 104.66537475585938, "distillation_loss": 4.307067394256592, "epoch": 4.38, "learning_rate": 3.932999373825924e-05, "loss": 108.749, "step": 5183, "task_loss": 2.803724765777588 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9912119392063079, "compression/movement_sparsity/importance_threshold": -5.703578775323372e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9206806992179273, "compression/movement_sparsity/model_sparsity": 0.889052484854672, "compression_loss": 104.66830444335938, "distillation_loss": 4.338745594024658, "epoch": 4.38, "learning_rate": 3.932686286787727e-05, "loss": 109.1161, "step": 5184, "task_loss": 3.310619354248047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9912478576663771, "compression/movement_sparsity/importance_threshold": -5.6802671743566915e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9207305303144775, "compression/movement_sparsity/model_sparsity": 0.8891006040997573, "compression_loss": 104.67121124267578, "distillation_loss": 5.573431491851807, "epoch": 4.38, "learning_rate": 3.932373199749531e-05, "loss": 109.63, "step": 5185, "task_loss": 2.7178380489349365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9912836781225234, "compression/movement_sparsity/importance_threshold": -5.657019179355991e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9207320804562702, "compression/movement_sparsity/model_sparsity": 0.8891021009894107, "compression_loss": 104.67415618896484, "distillation_loss": 2.93123722076416, "epoch": 4.38, "learning_rate": 3.932060112711334e-05, "loss": 109.1991, "step": 5186, "task_loss": 2.2868435382843018 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9913194007086319, "compression/movement_sparsity/importance_threshold": -5.633834703427584e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9207550702514721, "compression/movement_sparsity/model_sparsity": 0.8891243010144225, "compression_loss": 104.67710876464844, "distillation_loss": 5.564872741699219, "epoch": 4.38, "learning_rate": 3.931747025673138e-05, "loss": 108.8402, "step": 5187, "task_loss": 2.4937405586242676 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9913550255585877, "compression/movement_sparsity/importance_threshold": -5.6107136596783916e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9208043289879758, "compression/movement_sparsity/model_sparsity": 0.8891718675617898, "compression_loss": 104.67997741699219, "distillation_loss": 3.0269713401794434, "epoch": 4.39, "learning_rate": 3.931433938634941e-05, "loss": 108.3793, "step": 5188, "task_loss": 1.3585748672485352 }, { "compression/movement_sparsity/importance_regularization_factor": 0.991390552806276, "compression/movement_sparsity/importance_threshold": -5.587655961214813e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9208593113249448, "compression/movement_sparsity/model_sparsity": 0.8892249610863384, "compression_loss": 104.68290710449219, "distillation_loss": 2.844818353652954, "epoch": 4.39, "learning_rate": 3.931120851596744e-05, "loss": 108.6474, "step": 5189, "task_loss": 2.70680570602417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9914259825855819, "compression/movement_sparsity/importance_threshold": -5.564661521143509e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9208475302473206, "compression/movement_sparsity/model_sparsity": 0.8892135847249734, "compression_loss": 104.685791015625, "distillation_loss": 5.142570495605469, "epoch": 4.39, "learning_rate": 3.9308077645585474e-05, "loss": 109.5576, "step": 5190, "task_loss": 2.529120683670044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9914613150303906, "compression/movement_sparsity/importance_threshold": -5.5417302525710525e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.920941993503332, "compression/movement_sparsity/model_sparsity": 0.8893048028775378, "compression_loss": 104.68869018554688, "distillation_loss": 3.962461471557617, "epoch": 4.39, "learning_rate": 3.930494677520351e-05, "loss": 108.6722, "step": 5191, "task_loss": 1.6555118560791016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9914965502745872, "compression/movement_sparsity/importance_threshold": -5.5188620686041044e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.920930462833228, "compression/movement_sparsity/model_sparsity": 0.8892936683214244, "compression_loss": 104.69160461425781, "distillation_loss": 5.409380912780762, "epoch": 4.39, "learning_rate": 3.9301815904821544e-05, "loss": 109.1226, "step": 5192, "task_loss": 3.732964038848877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9915316884520569, "compression/movement_sparsity/importance_threshold": -5.496056882349151e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9209550743152285, "compression/movement_sparsity/model_sparsity": 0.8893174343233043, "compression_loss": 104.69451904296875, "distillation_loss": 4.060059070587158, "epoch": 4.39, "learning_rate": 3.9298685034439576e-05, "loss": 108.4297, "step": 5193, "task_loss": 2.1013882160186768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9915667296966847, "compression/movement_sparsity/importance_threshold": -5.47331460691294e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9209536434151121, "compression/movement_sparsity/model_sparsity": 0.889316052579009, "compression_loss": 104.69744873046875, "distillation_loss": 5.445862770080566, "epoch": 4.39, "learning_rate": 3.929555416405761e-05, "loss": 109.113, "step": 5194, "task_loss": 3.48335599899292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9916016741423559, "compression/movement_sparsity/importance_threshold": -5.450635155402044e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9209462265828426, "compression/movement_sparsity/model_sparsity": 0.8893088905377449, "compression_loss": 104.70033264160156, "distillation_loss": 4.436367034912109, "epoch": 4.39, "learning_rate": 3.929242329367565e-05, "loss": 109.0061, "step": 5195, "task_loss": 1.725710391998291 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9916365219229556, "compression/movement_sparsity/importance_threshold": -5.4280184409229507e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9209574710729233, "compression/movement_sparsity/model_sparsity": 0.8893197487449991, "compression_loss": 104.70327758789062, "distillation_loss": 5.064736366271973, "epoch": 4.39, "learning_rate": 3.928929242329368e-05, "loss": 108.7427, "step": 5196, "task_loss": 2.5350234508514404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9916712731723689, "compression/movement_sparsity/importance_threshold": -5.405464376582406e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9209971547028154, "compression/movement_sparsity/model_sparsity": 0.8893580691201233, "compression_loss": 104.70611572265625, "distillation_loss": 6.227995872497559, "epoch": 4.39, "learning_rate": 3.928616155291171e-05, "loss": 108.8092, "step": 5197, "task_loss": 2.9704463481903076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9917059280244809, "compression/movement_sparsity/importance_threshold": -5.382972875486985e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9211133795647621, "compression/movement_sparsity/model_sparsity": 0.8894703013005127, "compression_loss": 104.70901489257812, "distillation_loss": 2.6069774627685547, "epoch": 4.39, "learning_rate": 3.928303068252975e-05, "loss": 108.9094, "step": 5198, "task_loss": 1.7852319478988647 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9917404866131768, "compression/movement_sparsity/importance_threshold": -5.3605438507432594e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9211801787518583, "compression/movement_sparsity/model_sparsity": 0.8895348057300337, "compression_loss": 104.7118911743164, "distillation_loss": 4.4920268058776855, "epoch": 4.39, "learning_rate": 3.927989981214778e-05, "loss": 108.705, "step": 5199, "task_loss": 3.286885976791382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9917749490723418, "compression/movement_sparsity/importance_threshold": -5.338177215457804e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.921203502423754, "compression/movement_sparsity/model_sparsity": 0.8895573281620478, "compression_loss": 104.71475219726562, "distillation_loss": 2.7745168209075928, "epoch": 4.4, "learning_rate": 3.927676894176581e-05, "loss": 108.6118, "step": 5200, "task_loss": 0.9999497532844543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9918093155358609, "compression/movement_sparsity/importance_threshold": -5.3158728827372784e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9212098937776069, "compression/movement_sparsity/model_sparsity": 0.8895634999532337, "compression_loss": 104.71759796142578, "distillation_loss": 5.208150863647461, "epoch": 4.4, "learning_rate": 3.9273638071383844e-05, "loss": 109.3107, "step": 5201, "task_loss": 2.9855878353118896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9918435861376194, "compression/movement_sparsity/importance_threshold": -5.2936307656882564e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9211940465588188, "compression/movement_sparsity/model_sparsity": 0.8895481971351628, "compression_loss": 104.72045135498047, "distillation_loss": 5.0458807945251465, "epoch": 4.4, "learning_rate": 3.927050720100188e-05, "loss": 109.9479, "step": 5202, "task_loss": 2.377811908721924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9918777610115023, "compression/movement_sparsity/importance_threshold": -5.271450777417398e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9211761126106944, "compression/movement_sparsity/model_sparsity": 0.8895308792733277, "compression_loss": 104.72322082519531, "distillation_loss": 4.574875831604004, "epoch": 4.4, "learning_rate": 3.9267376330619914e-05, "loss": 108.4811, "step": 5203, "task_loss": 2.233610153198242 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9919118402913948, "compression/movement_sparsity/importance_threshold": -5.249332831031191e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9212013679977472, "compression/movement_sparsity/model_sparsity": 0.8895552670601407, "compression_loss": 104.72602844238281, "distillation_loss": 3.884023904800415, "epoch": 4.4, "learning_rate": 3.9264245460237946e-05, "loss": 108.3149, "step": 5204, "task_loss": 2.56939697265625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.991945824111182, "compression/movement_sparsity/importance_threshold": -5.227276839636381e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9212763590880091, "compression/movement_sparsity/model_sparsity": 0.8896276819757525, "compression_loss": 104.72876739501953, "distillation_loss": 3.881584882736206, "epoch": 4.4, "learning_rate": 3.926111458985598e-05, "loss": 108.7126, "step": 5205, "task_loss": 2.752305030822754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9919797126047492, "compression/movement_sparsity/importance_threshold": -5.205282716339455e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9213389251955945, "compression/movement_sparsity/model_sparsity": 0.8896880987450665, "compression_loss": 104.73153686523438, "distillation_loss": 3.49747896194458, "epoch": 4.4, "learning_rate": 3.9257983719474017e-05, "loss": 109.4335, "step": 5206, "task_loss": 1.4005225896835327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9920135059059813, "compression/movement_sparsity/importance_threshold": -5.183350374247161e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9213999769338901, "compression/movement_sparsity/model_sparsity": 0.8897470531683345, "compression_loss": 104.7342300415039, "distillation_loss": 4.840272903442383, "epoch": 4.4, "learning_rate": 3.925485284909205e-05, "loss": 108.9266, "step": 5207, "task_loss": 2.360377788543701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9920472041487636, "compression/movement_sparsity/importance_threshold": -5.161479726465898e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9214031368383137, "compression/movement_sparsity/model_sparsity": 0.88975010452032, "compression_loss": 104.73694610595703, "distillation_loss": 5.055424690246582, "epoch": 4.4, "learning_rate": 3.925172197871008e-05, "loss": 109.0133, "step": 5208, "task_loss": 3.0202043056488037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9920808074669812, "compression/movement_sparsity/importance_threshold": -5.1396706861024134e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.92150660284089, "compression/movement_sparsity/model_sparsity": 0.8898500161474092, "compression_loss": 104.73966217041016, "distillation_loss": 4.415210247039795, "epoch": 4.4, "learning_rate": 3.924859110832811e-05, "loss": 109.3065, "step": 5209, "task_loss": 2.103238344192505 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9921143159945193, "compression/movement_sparsity/importance_threshold": -5.117923166263281e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215305465695028, "compression/movement_sparsity/model_sparsity": 0.8898731373352847, "compression_loss": 104.74244689941406, "distillation_loss": 4.262058258056641, "epoch": 4.4, "learning_rate": 3.924546023794615e-05, "loss": 109.3781, "step": 5210, "task_loss": 2.6424453258514404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9921477298652629, "compression/movement_sparsity/importance_threshold": -5.096237080055074e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215182885251731, "compression/movement_sparsity/model_sparsity": 0.8898613003924879, "compression_loss": 104.74518585205078, "distillation_loss": 4.871269226074219, "epoch": 4.4, "learning_rate": 3.924232936756418e-05, "loss": 109.2351, "step": 5211, "task_loss": 2.4692862033843994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9921810492130972, "compression/movement_sparsity/importance_threshold": -5.074612340584452e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9216002075568316, "compression/movement_sparsity/model_sparsity": 0.8899404052533965, "compression_loss": 104.7479248046875, "distillation_loss": 4.387067794799805, "epoch": 4.41, "learning_rate": 3.9239198497182214e-05, "loss": 109.1859, "step": 5212, "task_loss": 2.600094795227051 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9922142741719074, "compression/movement_sparsity/importance_threshold": -5.05304886095799e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9216330586386683, "compression/movement_sparsity/model_sparsity": 0.8899721277995104, "compression_loss": 104.75072479248047, "distillation_loss": 6.0760040283203125, "epoch": 4.41, "learning_rate": 3.923606762680025e-05, "loss": 109.2107, "step": 5213, "task_loss": 2.412379264831543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9922474048755787, "compression/movement_sparsity/importance_threshold": -5.03154655428226e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215966064582055, "compression/movement_sparsity/model_sparsity": 0.8899369278635865, "compression_loss": 104.75344848632812, "distillation_loss": 4.4722161293029785, "epoch": 4.41, "learning_rate": 3.9232936756418284e-05, "loss": 108.2767, "step": 5214, "task_loss": 2.1222755908966064 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9922804414579961, "compression/movement_sparsity/importance_threshold": -5.010105333663924e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215816535519902, "compression/movement_sparsity/model_sparsity": 0.8899224886357001, "compression_loss": 104.75611114501953, "distillation_loss": 3.287637710571289, "epoch": 4.41, "learning_rate": 3.9229805886036316e-05, "loss": 108.7801, "step": 5215, "task_loss": 2.905970811843872 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9923133840530447, "compression/movement_sparsity/importance_threshold": -4.9887251122095536e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.921574224795553, "compression/movement_sparsity/model_sparsity": 0.8899153150799001, "compression_loss": 104.7588119506836, "distillation_loss": 4.728211402893066, "epoch": 4.41, "learning_rate": 3.9226675015654355e-05, "loss": 109.2139, "step": 5216, "task_loss": 2.10858416557312 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9923462327946098, "compression/movement_sparsity/importance_threshold": -4.96740580302581e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215831202246093, "compression/movement_sparsity/model_sparsity": 0.8899239049236028, "compression_loss": 104.76145935058594, "distillation_loss": 6.103675842285156, "epoch": 4.41, "learning_rate": 3.9223544145272386e-05, "loss": 109.478, "step": 5217, "task_loss": 3.7759177684783936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9923789878165764, "compression/movement_sparsity/importance_threshold": -4.94614731921918e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215951517097539, "compression/movement_sparsity/model_sparsity": 0.8899355230902195, "compression_loss": 104.76409149169922, "distillation_loss": 4.2615580558776855, "epoch": 4.41, "learning_rate": 3.9220413274890425e-05, "loss": 109.1701, "step": 5218, "task_loss": 2.7675790786743164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9924116492528298, "compression/movement_sparsity/importance_threshold": -4.924949573896411e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215743082647264, "compression/movement_sparsity/model_sparsity": 0.8899153956816507, "compression_loss": 104.76666259765625, "distillation_loss": 2.6360514163970947, "epoch": 4.41, "learning_rate": 3.921728240450846e-05, "loss": 108.2318, "step": 5219, "task_loss": 2.046319007873535 }, { "compression/movement_sparsity/importance_regularization_factor": 0.992444217237255, "compression/movement_sparsity/importance_threshold": -4.903812480163902e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215227600880366, "compression/movement_sparsity/model_sparsity": 0.8898656183434108, "compression_loss": 104.76920318603516, "distillation_loss": 4.097743511199951, "epoch": 4.41, "learning_rate": 3.921415153412649e-05, "loss": 108.6768, "step": 5220, "task_loss": 1.3211265802383423 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9924766919037372, "compression/movement_sparsity/importance_threshold": -4.882735951128401e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215104781953717, "compression/movement_sparsity/model_sparsity": 0.8898537583715425, "compression_loss": 104.77178192138672, "distillation_loss": 5.247386932373047, "epoch": 4.41, "learning_rate": 3.921102066374453e-05, "loss": 109.1712, "step": 5221, "task_loss": 3.5119855403900146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9925090733861615, "compression/movement_sparsity/importance_threshold": -4.861719899896654e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215229151022158, "compression/movement_sparsity/model_sparsity": 0.8898657680323762, "compression_loss": 104.77435302734375, "distillation_loss": 4.337973594665527, "epoch": 4.41, "learning_rate": 3.920788979336256e-05, "loss": 108.7604, "step": 5222, "task_loss": 2.097370147705078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9925413618184131, "compression/movement_sparsity/importance_threshold": -4.840764239574975e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9214157287593371, "compression/movement_sparsity/model_sparsity": 0.889762263870119, "compression_loss": 104.77689361572266, "distillation_loss": 3.8783559799194336, "epoch": 4.41, "learning_rate": 3.920475892298059e-05, "loss": 109.1513, "step": 5223, "task_loss": 1.5678237676620483 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9925735573343771, "compression/movement_sparsity/importance_threshold": -4.819868883270111e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9213961850485819, "compression/movement_sparsity/model_sparsity": 0.8897433915459518, "compression_loss": 104.77942657470703, "distillation_loss": 3.6743621826171875, "epoch": 4.42, "learning_rate": 3.920162805259863e-05, "loss": 108.3935, "step": 5224, "task_loss": 2.5676376819610596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9926056600679386, "compression/movement_sparsity/importance_threshold": -4.7990337440887224e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9213872299986874, "compression/movement_sparsity/model_sparsity": 0.8897347441295701, "compression_loss": 104.78193664550781, "distillation_loss": 5.513715744018555, "epoch": 4.42, "learning_rate": 3.919849718221666e-05, "loss": 109.6042, "step": 5225, "task_loss": 3.014604330062866 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9926376701529829, "compression/movement_sparsity/importance_threshold": -4.778258735137295e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9214073222211538, "compression/movement_sparsity/model_sparsity": 0.8897541461223839, "compression_loss": 104.78446197509766, "distillation_loss": 4.024284362792969, "epoch": 4.42, "learning_rate": 3.919536631183469e-05, "loss": 108.5432, "step": 5226, "task_loss": 2.5782055854797363 }, { "compression/movement_sparsity/importance_regularization_factor": 0.992669587723395, "compression/movement_sparsity/importance_threshold": -4.7575437695224036e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9213962923660907, "compression/movement_sparsity/model_sparsity": 0.889743495176774, "compression_loss": 104.78694915771484, "distillation_loss": 4.841449737548828, "epoch": 4.42, "learning_rate": 3.9192235441452725e-05, "loss": 108.646, "step": 5227, "task_loss": 2.3400683403015137 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99270141291306, "compression/movement_sparsity/importance_threshold": -4.736888760350881e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9214374665169373, "compression/movement_sparsity/model_sparsity": 0.8897832548688726, "compression_loss": 104.78944396972656, "distillation_loss": 2.705486297607422, "epoch": 4.42, "learning_rate": 3.918910457107076e-05, "loss": 108.2661, "step": 5228, "task_loss": 1.0299631357192993 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9927331458558631, "compression/movement_sparsity/importance_threshold": -4.716293620729128e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9214424985156796, "compression/movement_sparsity/model_sparsity": 0.8897881140029779, "compression_loss": 104.79192352294922, "distillation_loss": 4.05254602432251, "epoch": 4.42, "learning_rate": 3.9185973700688795e-05, "loss": 108.3629, "step": 5229, "task_loss": 3.1972274780273438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9927647866856896, "compression/movement_sparsity/importance_threshold": -4.695758263763717e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9215370810133673, "compression/movement_sparsity/model_sparsity": 0.8898794473009001, "compression_loss": 104.79434204101562, "distillation_loss": 4.012119293212891, "epoch": 4.42, "learning_rate": 3.918284283030683e-05, "loss": 108.3135, "step": 5230, "task_loss": 1.5457711219787598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9927963355364243, "compression/movement_sparsity/importance_threshold": -4.675282602561483e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.921671251747606, "compression/movement_sparsity/model_sparsity": 0.8900090088576603, "compression_loss": 104.79676818847656, "distillation_loss": 4.5647454261779785, "epoch": 4.42, "learning_rate": 3.917971195992486e-05, "loss": 108.9735, "step": 5231, "task_loss": 2.073485851287842 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9928277925419527, "compression/movement_sparsity/importance_threshold": -4.654866550228738e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.921733078556798, "compression/movement_sparsity/model_sparsity": 0.890068711725755, "compression_loss": 104.7991943359375, "distillation_loss": 3.914003610610962, "epoch": 4.42, "learning_rate": 3.91765810895429e-05, "loss": 108.7262, "step": 5232, "task_loss": 1.9792706966400146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9928591578361597, "compression/movement_sparsity/importance_threshold": -4.634510019872317e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217853660318812, "compression/movement_sparsity/model_sparsity": 0.8901192029652141, "compression_loss": 104.80162811279297, "distillation_loss": 5.11370849609375, "epoch": 4.42, "learning_rate": 3.917345021916093e-05, "loss": 108.2779, "step": 5233, "task_loss": 3.3140928745269775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9928904315529306, "compression/movement_sparsity/importance_threshold": -4.614212924598793e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217553052052712, "compression/movement_sparsity/model_sparsity": 0.890090174820476, "compression_loss": 104.8039779663086, "distillation_loss": 3.8267359733581543, "epoch": 4.42, "learning_rate": 3.917031934877896e-05, "loss": 108.7209, "step": 5234, "task_loss": 2.958289384841919 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9929216138261504, "compression/movement_sparsity/importance_threshold": -4.5939751775146524e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217993769288534, "compression/movement_sparsity/model_sparsity": 0.8901327325447727, "compression_loss": 104.8063735961914, "distillation_loss": 3.9199142456054688, "epoch": 4.42, "learning_rate": 3.9167188478397e-05, "loss": 109.0279, "step": 5235, "task_loss": 1.7346760034561157 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9929527047897043, "compression/movement_sparsity/importance_threshold": -4.573796691726469e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217190438114905, "compression/movement_sparsity/model_sparsity": 0.8900551591171248, "compression_loss": 104.80876922607422, "distillation_loss": 4.086296081542969, "epoch": 4.43, "learning_rate": 3.916405760801503e-05, "loss": 109.2378, "step": 5236, "task_loss": 2.672579288482666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9929837045774774, "compression/movement_sparsity/importance_threshold": -4.553677380341077e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217340086418736, "compression/movement_sparsity/model_sparsity": 0.890069609859547, "compression_loss": 104.8111343383789, "distillation_loss": 5.3857221603393555, "epoch": 4.43, "learning_rate": 3.916092673763306e-05, "loss": 109.0891, "step": 5237, "task_loss": 2.4617698192596436 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9930146133233549, "compression/movement_sparsity/importance_threshold": -4.533617156464963e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217246958669499, "compression/movement_sparsity/model_sparsity": 0.8900606170070914, "compression_loss": 104.81360626220703, "distillation_loss": 4.920108318328857, "epoch": 4.43, "learning_rate": 3.9157795867251095e-05, "loss": 108.7942, "step": 5238, "task_loss": 3.0120975971221924 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9930454311612219, "compression/movement_sparsity/importance_threshold": -4.5136159332046996e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217253755445052, "compression/movement_sparsity/model_sparsity": 0.8900612733356317, "compression_loss": 104.8160171508789, "distillation_loss": 5.516894340515137, "epoch": 4.43, "learning_rate": 3.915466499686913e-05, "loss": 109.6974, "step": 5239, "task_loss": 4.13726806640625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9930761582249636, "compression/movement_sparsity/importance_threshold": -4.4936736236667746e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9217443468952139, "compression/movement_sparsity/model_sparsity": 0.8900795929620808, "compression_loss": 104.81836700439453, "distillation_loss": 4.678866863250732, "epoch": 4.43, "learning_rate": 3.9151534126487165e-05, "loss": 108.469, "step": 5240, "task_loss": 2.009708881378174 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9931067946484651, "compression/movement_sparsity/importance_threshold": -4.473790140958108e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9218189325637761, "compression/movement_sparsity/model_sparsity": 0.8901516163834757, "compression_loss": 104.82076263427734, "distillation_loss": 2.997682571411133, "epoch": 4.43, "learning_rate": 3.91484032561052e-05, "loss": 109.0701, "step": 5241, "task_loss": 1.8908257484436035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9931373405656115, "compression/movement_sparsity/importance_threshold": -4.4539653981850134e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.921822247482379, "compression/movement_sparsity/model_sparsity": 0.8901548174244266, "compression_loss": 104.8231430053711, "distillation_loss": 3.7847766876220703, "epoch": 4.43, "learning_rate": 3.914527238572323e-05, "loss": 108.636, "step": 5242, "task_loss": 3.237757682800293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.993167796110288, "compression/movement_sparsity/importance_threshold": -4.434199308454151e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9218217585915058, "compression/movement_sparsity/model_sparsity": 0.8901543453284589, "compression_loss": 104.82555389404297, "distillation_loss": 3.939948081970215, "epoch": 4.43, "learning_rate": 3.914214151534127e-05, "loss": 108.888, "step": 5243, "task_loss": 1.2721835374832153 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9931981614163798, "compression/movement_sparsity/importance_threshold": -4.414491784872181e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9218601186387905, "compression/movement_sparsity/model_sparsity": 0.8901913875901101, "compression_loss": 104.82796478271484, "distillation_loss": 3.7275490760803223, "epoch": 4.43, "learning_rate": 3.91390106449593e-05, "loss": 108.6687, "step": 5244, "task_loss": 2.595417022705078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9932284366177719, "compression/movement_sparsity/importance_threshold": -4.394842740545763e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9218939832748763, "compression/movement_sparsity/model_sparsity": 0.8902240888717665, "compression_loss": 104.830322265625, "distillation_loss": 4.744063377380371, "epoch": 4.43, "learning_rate": 3.913587977457733e-05, "loss": 109.0521, "step": 5245, "task_loss": 2.6495957374572754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9932586218483495, "compression/movement_sparsity/importance_threshold": -4.3752520885814716e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9219308408770387, "compression/movement_sparsity/model_sparsity": 0.8902596803019075, "compression_loss": 104.83267211914062, "distillation_loss": 3.547182559967041, "epoch": 4.43, "learning_rate": 3.913274890419536e-05, "loss": 109.2431, "step": 5246, "task_loss": 1.8291900157928467 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9932887172419977, "compression/movement_sparsity/importance_threshold": -4.35571974208588e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9219749602972914, "compression/movement_sparsity/model_sparsity": 0.8903022840843472, "compression_loss": 104.83503723144531, "distillation_loss": 3.4591503143310547, "epoch": 4.44, "learning_rate": 3.91296180338134e-05, "loss": 108.5824, "step": 5247, "task_loss": 1.8545434474945068 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9933187229326018, "compression/movement_sparsity/importance_threshold": -4.3362456141654744e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9220000129734944, "compression/movement_sparsity/model_sparsity": 0.8903264761240516, "compression_loss": 104.83750915527344, "distillation_loss": 4.299471855163574, "epoch": 4.44, "learning_rate": 3.912648716343143e-05, "loss": 108.7157, "step": 5248, "task_loss": 2.62660813331604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9933486390540468, "compression/movement_sparsity/importance_threshold": -4.3168296179270024e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9220173626374045, "compression/movement_sparsity/model_sparsity": 0.8903432297736327, "compression_loss": 104.83989715576172, "distillation_loss": 5.310114860534668, "epoch": 4.44, "learning_rate": 3.912335629304947e-05, "loss": 109.2649, "step": 5249, "task_loss": 2.229037046432495 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9933784657402178, "compression/movement_sparsity/importance_threshold": -4.297471666477124e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9220296445300695, "compression/movement_sparsity/model_sparsity": 0.890355089745501, "compression_loss": 104.84227752685547, "distillation_loss": 3.8361470699310303, "epoch": 4.44, "learning_rate": 3.91202254226675e-05, "loss": 109.0718, "step": 5250, "task_loss": 2.7776265144348145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.993408203125, "compression/movement_sparsity/importance_threshold": -4.278171672922326e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9220268542748428, "compression/movement_sparsity/model_sparsity": 0.8903523953441251, "compression_loss": 104.8445816040039, "distillation_loss": 3.5003743171691895, "epoch": 4.44, "learning_rate": 3.9117094552285535e-05, "loss": 108.4146, "step": 5251, "task_loss": 2.2551348209381104 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9934378513422786, "compression/movement_sparsity/importance_threshold": -4.258929550369269e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9220867970655482, "compression/movement_sparsity/model_sparsity": 0.8904102789155642, "compression_loss": 104.846923828125, "distillation_loss": 4.123427867889404, "epoch": 4.44, "learning_rate": 3.911396368190357e-05, "loss": 108.7191, "step": 5252, "task_loss": 2.0874600410461426 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9934674105259387, "compression/movement_sparsity/importance_threshold": -4.2397452119244394e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9221232850185139, "compression/movement_sparsity/model_sparsity": 0.8904455133950956, "compression_loss": 104.84919738769531, "distillation_loss": 4.708954334259033, "epoch": 4.44, "learning_rate": 3.9110832811521605e-05, "loss": 108.6928, "step": 5253, "task_loss": 2.3759162425994873 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9934968808098653, "compression/movement_sparsity/importance_threshold": -4.220618570694584e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9221967498153185, "compression/movement_sparsity/model_sparsity": 0.8905164544501257, "compression_loss": 104.85151672363281, "distillation_loss": 3.94258975982666, "epoch": 4.44, "learning_rate": 3.9107701941139644e-05, "loss": 108.3729, "step": 5254, "task_loss": 2.492131471633911 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9935262623279438, "compression/movement_sparsity/importance_threshold": -4.201549539786277e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922191133532362, "compression/movement_sparsity/model_sparsity": 0.8905110311037665, "compression_loss": 104.85381317138672, "distillation_loss": 5.060490608215332, "epoch": 4.44, "learning_rate": 3.9104571070757675e-05, "loss": 109.6051, "step": 5255, "task_loss": 3.3539557456970215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9935555552140591, "compression/movement_sparsity/importance_threshold": -4.1825380323060914e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9222048105526404, "compression/movement_sparsity/model_sparsity": 0.8905242382763228, "compression_loss": 104.85609436035156, "distillation_loss": 5.043981552124023, "epoch": 4.44, "learning_rate": 3.910144020037571e-05, "loss": 109.3956, "step": 5256, "task_loss": 2.086989402770996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9935847596020966, "compression/movement_sparsity/importance_threshold": -4.1635839613606875e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922242514770705, "compression/movement_sparsity/model_sparsity": 0.8905606472385051, "compression_loss": 104.85843658447266, "distillation_loss": 4.870553016662598, "epoch": 4.44, "learning_rate": 3.909830932999374e-05, "loss": 109.1375, "step": 5257, "task_loss": 2.3580644130706787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9936138756259413, "compression/movement_sparsity/importance_threshold": -4.1446872400564654e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9222112615273313, "compression/movement_sparsity/model_sparsity": 0.8905304676401876, "compression_loss": 104.86073303222656, "distillation_loss": 6.055621147155762, "epoch": 4.44, "learning_rate": 3.909517845961178e-05, "loss": 109.3461, "step": 5258, "task_loss": 3.0998194217681885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9936429034194781, "compression/movement_sparsity/importance_threshold": -4.125847781500432e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9222626308415066, "compression/movement_sparsity/model_sparsity": 0.8905800722603905, "compression_loss": 104.86306762695312, "distillation_loss": 5.838710784912109, "epoch": 4.45, "learning_rate": 3.909204758922981e-05, "loss": 109.0118, "step": 5259, "task_loss": 3.215682029724121 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9936718431165926, "compression/movement_sparsity/importance_threshold": -4.107065498798728e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9223681835734193, "compression/movement_sparsity/model_sparsity": 0.8906819989312438, "compression_loss": 104.86539459228516, "distillation_loss": 3.9886727333068848, "epoch": 4.45, "learning_rate": 3.908891671884784e-05, "loss": 109.2574, "step": 5260, "task_loss": 2.0089874267578125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9937006948511696, "compression/movement_sparsity/importance_threshold": -4.088340305058273e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224144851163494, "compression/movement_sparsity/model_sparsity": 0.890726709873734, "compression_loss": 104.86768341064453, "distillation_loss": 5.58887243270874, "epoch": 4.45, "learning_rate": 3.908578584846588e-05, "loss": 109.2874, "step": 5261, "task_loss": 3.0017571449279785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9937294587570944, "compression/movement_sparsity/importance_threshold": -4.069672113385641e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224390131291763, "compression/movement_sparsity/model_sparsity": 0.8907503952738633, "compression_loss": 104.86988830566406, "distillation_loss": 2.7880280017852783, "epoch": 4.45, "learning_rate": 3.908265497808391e-05, "loss": 108.0398, "step": 5262, "task_loss": 2.175285816192627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9937581349682522, "compression/movement_sparsity/importance_threshold": -4.051060836887232e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224601189058917, "compression/movement_sparsity/model_sparsity": 0.8907707760022198, "compression_loss": 104.87216186523438, "distillation_loss": 3.7732460498809814, "epoch": 4.45, "learning_rate": 3.907952410770194e-05, "loss": 108.4247, "step": 5263, "task_loss": 1.9957847595214844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9937867236185278, "compression/movement_sparsity/importance_threshold": -4.032506388669879e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224551942246582, "compression/movement_sparsity/model_sparsity": 0.8907660204989366, "compression_loss": 104.87435150146484, "distillation_loss": 3.764341115951538, "epoch": 4.45, "learning_rate": 3.9076393237319975e-05, "loss": 109.0613, "step": 5264, "task_loss": 2.689113140106201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9938152248418067, "compression/movement_sparsity/importance_threshold": -4.01400868184007e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9225355511903562, "compression/movement_sparsity/model_sparsity": 0.890843616955656, "compression_loss": 104.87655639648438, "distillation_loss": 5.940247535705566, "epoch": 4.45, "learning_rate": 3.9073262366938014e-05, "loss": 109.9672, "step": 5265, "task_loss": 3.0225565433502197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9938436387719739, "compression/movement_sparsity/importance_threshold": -3.9955676295044644e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9225108085425119, "compression/movement_sparsity/model_sparsity": 0.8908197242938823, "compression_loss": 104.87875366210938, "distillation_loss": 4.918890953063965, "epoch": 4.45, "learning_rate": 3.9070131496556045e-05, "loss": 109.1973, "step": 5266, "task_loss": 2.9872195720672607 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9938719655429147, "compression/movement_sparsity/importance_threshold": -3.977183144769549e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224774208731314, "compression/movement_sparsity/model_sparsity": 0.8907874835936576, "compression_loss": 104.88092041015625, "distillation_loss": 4.459075927734375, "epoch": 4.45, "learning_rate": 3.906700062617408e-05, "loss": 109.2388, "step": 5267, "task_loss": 3.7838973999023438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9939002052885139, "compression/movement_sparsity/importance_threshold": -3.9588551407420715e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9225019846584613, "compression/movement_sparsity/model_sparsity": 0.8908112035373944, "compression_loss": 104.88311004638672, "distillation_loss": 4.022736549377441, "epoch": 4.45, "learning_rate": 3.906386975579211e-05, "loss": 109.0134, "step": 5268, "task_loss": 2.099761962890625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9939283581426569, "compression/movement_sparsity/importance_threshold": -3.9405835305286047e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224935900444456, "compression/movement_sparsity/model_sparsity": 0.890803097304195, "compression_loss": 104.88525390625, "distillation_loss": 7.006826877593994, "epoch": 4.45, "learning_rate": 3.906073888541015e-05, "loss": 109.5404, "step": 5269, "task_loss": 3.479982852935791 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9939564242392288, "compression/movement_sparsity/importance_threshold": -3.9223682272356356e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224816539526421, "compression/movement_sparsity/model_sparsity": 0.8907915712538647, "compression_loss": 104.88743591308594, "distillation_loss": 4.24521017074585, "epoch": 4.45, "learning_rate": 3.905760801502818e-05, "loss": 108.7445, "step": 5270, "task_loss": 1.5054097175598145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9939844037121147, "compression/movement_sparsity/importance_threshold": -3.904209143969911e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9225000529433043, "compression/movement_sparsity/model_sparsity": 0.8908093381825957, "compression_loss": 104.88961029052734, "distillation_loss": 3.2264976501464844, "epoch": 4.46, "learning_rate": 3.905447714464621e-05, "loss": 108.7914, "step": 5271, "task_loss": 2.335114002227783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9940122966951997, "compression/movement_sparsity/importance_threshold": -3.886106193838092e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9224911813625832, "compression/movement_sparsity/model_sparsity": 0.8908007713679645, "compression_loss": 104.89173126220703, "distillation_loss": 4.088120937347412, "epoch": 4.46, "learning_rate": 3.905134627426425e-05, "loss": 108.8399, "step": 5272, "task_loss": 2.5771496295928955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9940401033223691, "compression/movement_sparsity/importance_threshold": -3.8680592899464905e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9225525312050697, "compression/movement_sparsity/model_sparsity": 0.8908600136546274, "compression_loss": 104.89390563964844, "distillation_loss": 5.352330684661865, "epoch": 4.46, "learning_rate": 3.904821540388228e-05, "loss": 108.9064, "step": 5273, "task_loss": 3.81376576423645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9940678237275079, "compression/movement_sparsity/importance_threshold": -3.850068345402028e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226211667139818, "compression/movement_sparsity/model_sparsity": 0.8909262913226608, "compression_loss": 104.89605712890625, "distillation_loss": 3.9217419624328613, "epoch": 4.46, "learning_rate": 3.904508453350031e-05, "loss": 109.0913, "step": 5274, "task_loss": 2.286540985107422 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9940954580445013, "compression/movement_sparsity/importance_threshold": -3.832133273311191e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226502139863427, "compression/movement_sparsity/model_sparsity": 0.8909543407318563, "compression_loss": 104.8980712890625, "distillation_loss": 4.030981063842773, "epoch": 4.46, "learning_rate": 3.9041953663118345e-05, "loss": 108.5615, "step": 5275, "task_loss": 1.5816586017608643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9941230064072344, "compression/movement_sparsity/importance_threshold": -3.8142539867804665e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226232057466475, "compression/movement_sparsity/model_sparsity": 0.8909282603082816, "compression_loss": 104.90018463134766, "distillation_loss": 4.432637691497803, "epoch": 4.46, "learning_rate": 3.9038822792736384e-05, "loss": 109.6509, "step": 5276, "task_loss": 3.0886049270629883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9941504689495924, "compression/movement_sparsity/importance_threshold": -3.796430398916601e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226220610265544, "compression/movement_sparsity/model_sparsity": 0.8909271549128454, "compression_loss": 104.9021987915039, "distillation_loss": 3.5032286643981934, "epoch": 4.46, "learning_rate": 3.9035691922354415e-05, "loss": 108.4376, "step": 5277, "task_loss": 2.1379404067993164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9941778458054604, "compression/movement_sparsity/importance_threshold": -3.778662422826082e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226249705234576, "compression/movement_sparsity/model_sparsity": 0.8909299644595793, "compression_loss": 104.90426635742188, "distillation_loss": 4.3856964111328125, "epoch": 4.46, "learning_rate": 3.903256105197245e-05, "loss": 108.431, "step": 5278, "task_loss": 1.803727626800537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9942051371087235, "compression/movement_sparsity/importance_threshold": -3.760949971615742e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226187461079517, "compression/movement_sparsity/model_sparsity": 0.8909239538718945, "compression_loss": 104.90626525878906, "distillation_loss": 4.30812406539917, "epoch": 4.46, "learning_rate": 3.902943018159048e-05, "loss": 108.3781, "step": 5279, "task_loss": 2.271852493286133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9942323429932669, "compression/movement_sparsity/importance_threshold": -3.7432929583919826e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226271645703026, "compression/movement_sparsity/model_sparsity": 0.8909320831341655, "compression_loss": 104.90834045410156, "distillation_loss": 5.2103118896484375, "epoch": 4.46, "learning_rate": 3.902629931120852e-05, "loss": 109.2174, "step": 5280, "task_loss": 2.771564483642578 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9942594635929758, "compression/movement_sparsity/importance_threshold": -3.725691296261376e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226562356909989, "compression/movement_sparsity/model_sparsity": 0.8909601555724326, "compression_loss": 104.9103775024414, "distillation_loss": 5.607692718505859, "epoch": 4.46, "learning_rate": 3.902316844082655e-05, "loss": 109.2381, "step": 5281, "task_loss": 2.8570120334625244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9942864990417352, "compression/movement_sparsity/importance_threshold": -3.70814489833067e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226770552876911, "compression/movement_sparsity/model_sparsity": 0.8909802599519298, "compression_loss": 104.91243743896484, "distillation_loss": 5.03422212600708, "epoch": 4.46, "learning_rate": 3.902003757044458e-05, "loss": 108.4783, "step": 5282, "task_loss": 3.5717780590057373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9943134494734303, "compression/movement_sparsity/importance_threshold": -3.6906536777063514e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9226743842741406, "compression/movement_sparsity/model_sparsity": 0.8909776806959119, "compression_loss": 104.9144515991211, "distillation_loss": 3.5012094974517822, "epoch": 4.47, "learning_rate": 3.901690670006262e-05, "loss": 108.8754, "step": 5283, "task_loss": 2.726625680923462 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9943403150219462, "compression/movement_sparsity/importance_threshold": -3.673217547495167e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922692580553953, "compression/movement_sparsity/model_sparsity": 0.8909952518775343, "compression_loss": 104.91647338867188, "distillation_loss": 5.708859920501709, "epoch": 4.47, "learning_rate": 3.901377582968065e-05, "loss": 109.7948, "step": 5284, "task_loss": 2.945394992828369 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9943670958211681, "compression/movement_sparsity/importance_threshold": -3.655836420803603e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9227418512146244, "compression/movement_sparsity/model_sparsity": 0.8910428299394373, "compression_loss": 104.91847229003906, "distillation_loss": 3.705343723297119, "epoch": 4.47, "learning_rate": 3.901064495929869e-05, "loss": 109.1796, "step": 5285, "task_loss": 1.8018414974212646 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9943937920049812, "compression/movement_sparsity/importance_threshold": -3.6385102107382335e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228017940053298, "compression/movement_sparsity/model_sparsity": 0.8911007135108765, "compression_loss": 104.9204330444336, "distillation_loss": 2.894232749938965, "epoch": 4.47, "learning_rate": 3.900751408891672e-05, "loss": 108.9192, "step": 5286, "task_loss": 1.8421754837036133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9944204037072706, "compression/movement_sparsity/importance_threshold": -3.6212388304058055e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9227865429949236, "compression/movement_sparsity/model_sparsity": 0.8910859864195952, "compression_loss": 104.92242431640625, "distillation_loss": 3.6209769248962402, "epoch": 4.47, "learning_rate": 3.900438321853476e-05, "loss": 109.0368, "step": 5287, "task_loss": 2.1490585803985596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9944469310619213, "compression/movement_sparsity/importance_threshold": -3.6040221929128056e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228407860334991, "compression/movement_sparsity/model_sparsity": 0.8911383660429246, "compression_loss": 104.92440032958984, "distillation_loss": 4.708249568939209, "epoch": 4.47, "learning_rate": 3.900125234815279e-05, "loss": 109.5676, "step": 5288, "task_loss": 2.9055228233337402 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9944733742028186, "compression/movement_sparsity/importance_threshold": -3.586860211365894e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9227827034129448, "compression/movement_sparsity/model_sparsity": 0.8910822787390694, "compression_loss": 104.9263687133789, "distillation_loss": 3.7109951972961426, "epoch": 4.47, "learning_rate": 3.8998121477770824e-05, "loss": 108.6798, "step": 5289, "task_loss": 1.9437371492385864 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9944997332638476, "compression/movement_sparsity/importance_threshold": -3.569752798871645e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9227781245325726, "compression/movement_sparsity/model_sparsity": 0.8910778571573243, "compression_loss": 104.92829895019531, "distillation_loss": 4.350333213806152, "epoch": 4.47, "learning_rate": 3.8994990607388856e-05, "loss": 108.7913, "step": 5290, "task_loss": 1.6621261835098267 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9945260083788934, "compression/movement_sparsity/importance_threshold": -3.552699868536631e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9227775521725261, "compression/movement_sparsity/model_sparsity": 0.8910773044596062, "compression_loss": 104.93026733398438, "distillation_loss": 3.402581214904785, "epoch": 4.47, "learning_rate": 3.8991859737006894e-05, "loss": 109.848, "step": 5291, "task_loss": 2.1002132892608643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9945521996818412, "compression/movement_sparsity/importance_threshold": -3.535701333467513e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922789345174318, "compression/movement_sparsity/model_sparsity": 0.891088692335507, "compression_loss": 104.93221282958984, "distillation_loss": 4.1564412117004395, "epoch": 4.47, "learning_rate": 3.8988728866624926e-05, "loss": 110.3441, "step": 5292, "task_loss": 2.4135560989379883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9945783073065761, "compression/movement_sparsity/importance_threshold": -3.518757106770777e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228027121662378, "compression/movement_sparsity/model_sparsity": 0.8911016001301327, "compression_loss": 104.93416595458984, "distillation_loss": 4.276082992553711, "epoch": 4.47, "learning_rate": 3.898559799624296e-05, "loss": 108.8935, "step": 5293, "task_loss": 1.8918296098709106 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9946043313869832, "compression/movement_sparsity/importance_threshold": -3.5018671015533444e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228116552919646, "compression/movement_sparsity/model_sparsity": 0.8911102360319786, "compression_loss": 104.93617248535156, "distillation_loss": 5.882461071014404, "epoch": 4.47, "learning_rate": 3.898246712586099e-05, "loss": 109.8707, "step": 5294, "task_loss": 2.807783365249634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9946302720569479, "compression/movement_sparsity/importance_threshold": -3.485031230921441e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9227935424813258, "compression/movement_sparsity/model_sparsity": 0.8910927454521067, "compression_loss": 104.93814086914062, "distillation_loss": 3.617096185684204, "epoch": 4.48, "learning_rate": 3.897933625547903e-05, "loss": 108.878, "step": 5295, "task_loss": 2.2942380905151367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.994656129450355, "compression/movement_sparsity/importance_threshold": -3.4682494079818146e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228546776887949, "compression/movement_sparsity/model_sparsity": 0.8911517804771253, "compression_loss": 104.94014739990234, "distillation_loss": 4.526923179626465, "epoch": 4.48, "learning_rate": 3.897620538509706e-05, "loss": 108.8202, "step": 5296, "task_loss": 2.555478811264038 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9946819037010897, "compression/movement_sparsity/importance_threshold": -3.4515215458412116e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228901759358469, "compression/movement_sparsity/model_sparsity": 0.8911860592501857, "compression_loss": 104.94213104248047, "distillation_loss": 4.733516216278076, "epoch": 4.48, "learning_rate": 3.897307451471509e-05, "loss": 108.9877, "step": 5297, "task_loss": 2.7907636165618896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9947075949430373, "compression/movement_sparsity/importance_threshold": -3.434847557606032e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229329836976596, "compression/movement_sparsity/model_sparsity": 0.8912273964336881, "compression_loss": 104.94408416748047, "distillation_loss": 3.833735942840576, "epoch": 4.48, "learning_rate": 3.896994364433313e-05, "loss": 108.7655, "step": 5298, "task_loss": 2.63924503326416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9947332033100829, "compression/movement_sparsity/importance_threshold": -3.4182273563829364e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922882246364369, "compression/movement_sparsity/model_sparsity": 0.8911784020838822, "compression_loss": 104.94599914550781, "distillation_loss": 4.926333904266357, "epoch": 4.48, "learning_rate": 3.896681277395116e-05, "loss": 109.2898, "step": 5299, "task_loss": 2.8917059898376465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9947587289361115, "compression/movement_sparsity/importance_threshold": -3.401660855278585e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229009076867191, "compression/movement_sparsity/model_sparsity": 0.8911964223324007, "compression_loss": 104.94790649414062, "distillation_loss": 5.437877178192139, "epoch": 4.48, "learning_rate": 3.8963681903569194e-05, "loss": 108.9934, "step": 5300, "task_loss": 2.3221662044525146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9947841719550083, "compression/movement_sparsity/importance_threshold": -3.385147967399551e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922893836655311, "compression/movement_sparsity/model_sparsity": 0.8911895942126745, "compression_loss": 104.94984436035156, "distillation_loss": 4.193413257598877, "epoch": 4.48, "learning_rate": 3.8960551033187225e-05, "loss": 108.8129, "step": 5301, "task_loss": 2.473317861557007 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9948095325006586, "compression/movement_sparsity/importance_threshold": -3.368688605852408e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9228940393661609, "compression/movement_sparsity/model_sparsity": 0.891189789959783, "compression_loss": 104.95173645019531, "distillation_loss": 3.0134565830230713, "epoch": 4.48, "learning_rate": 3.8957420162805264e-05, "loss": 108.6144, "step": 5302, "task_loss": 1.5591151714324951 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9948348107069473, "compression/movement_sparsity/importance_threshold": -3.3522826837439033e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229326140484628, "compression/movement_sparsity/model_sparsity": 0.8912270394830784, "compression_loss": 104.9536361694336, "distillation_loss": 4.1993408203125, "epoch": 4.48, "learning_rate": 3.8954289292423296e-05, "loss": 108.8887, "step": 5303, "task_loss": 2.623748540878296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9948600067077598, "compression/movement_sparsity/importance_threshold": -3.335930114180437e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.922937979923899, "compression/movement_sparsity/model_sparsity": 0.891232221024186, "compression_loss": 104.95553588867188, "distillation_loss": 4.849981307983398, "epoch": 4.48, "learning_rate": 3.895115842204133e-05, "loss": 109.7846, "step": 5304, "task_loss": 3.043452501296997 }, { "compression/movement_sparsity/importance_regularization_factor": 0.994885120636981, "compression/movement_sparsity/importance_threshold": -3.3196308102686686e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229890749822187, "compression/movement_sparsity/model_sparsity": 0.8912815608100656, "compression_loss": 104.95742797851562, "distillation_loss": 4.00096321105957, "epoch": 4.48, "learning_rate": 3.894802755165936e-05, "loss": 108.6422, "step": 5305, "task_loss": 2.257833242416382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9949101526284962, "compression/movement_sparsity/importance_threshold": -3.303384685115259e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229958479094359, "compression/movement_sparsity/model_sparsity": 0.8912881010663969, "compression_loss": 104.95935821533203, "distillation_loss": 3.4167909622192383, "epoch": 4.48, "learning_rate": 3.89448966812774e-05, "loss": 108.3845, "step": 5306, "task_loss": 1.7278761863708496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9949351028161905, "compression/movement_sparsity/importance_threshold": -3.287191651826782e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9230142230517627, "compression/movement_sparsity/model_sparsity": 0.8913058449660562, "compression_loss": 104.961181640625, "distillation_loss": 3.8735382556915283, "epoch": 4.49, "learning_rate": 3.894176581089543e-05, "loss": 109.0669, "step": 5307, "task_loss": 1.8631408214569092 }, { "compression/movement_sparsity/importance_regularization_factor": 0.994959971333949, "compression/movement_sparsity/importance_threshold": -3.271051623509897e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229520742900446, "compression/movement_sparsity/model_sparsity": 0.891245831205495, "compression_loss": 104.96311950683594, "distillation_loss": 3.133443593978882, "epoch": 4.49, "learning_rate": 3.893863494051346e-05, "loss": 109.3277, "step": 5308, "task_loss": 1.8549857139587402 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9949847583156568, "compression/movement_sparsity/importance_threshold": -3.2549645132710914e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229521816075533, "compression/movement_sparsity/model_sparsity": 0.8912459348363172, "compression_loss": 104.96497344970703, "distillation_loss": 4.805838108062744, "epoch": 4.49, "learning_rate": 3.89355040701315e-05, "loss": 109.7262, "step": 5309, "task_loss": 2.9234650135040283 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9950094638951992, "compression/movement_sparsity/importance_threshold": -3.238930234217025e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229887411055249, "compression/movement_sparsity/model_sparsity": 0.8912812384030633, "compression_loss": 104.96681213378906, "distillation_loss": 3.557203531265259, "epoch": 4.49, "learning_rate": 3.893237319974953e-05, "loss": 109.0063, "step": 5310, "task_loss": 1.8688182830810547 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9950340882064611, "compression/movement_sparsity/importance_threshold": -3.2229486994544455e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9229970522503671, "compression/movement_sparsity/model_sparsity": 0.8912892640345121, "compression_loss": 104.96865844726562, "distillation_loss": 4.956334590911865, "epoch": 4.49, "learning_rate": 3.8929242329367564e-05, "loss": 109.5451, "step": 5311, "task_loss": 2.433436155319214 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9950586313833278, "compression/movement_sparsity/importance_threshold": -3.2070198220897524e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9230566730885464, "compression/movement_sparsity/model_sparsity": 0.8913468367134848, "compression_loss": 104.97045135498047, "distillation_loss": 4.149485111236572, "epoch": 4.49, "learning_rate": 3.8926111458985595e-05, "loss": 108.5969, "step": 5312, "task_loss": 1.9444470405578613 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9950830935596846, "compression/movement_sparsity/importance_threshold": -3.191143515229519e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9230921713355984, "compression/movement_sparsity/model_sparsity": 0.8913811154865452, "compression_loss": 104.97228240966797, "distillation_loss": 3.8159453868865967, "epoch": 4.49, "learning_rate": 3.8922980588603634e-05, "loss": 108.7004, "step": 5313, "task_loss": 1.2942448854446411 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9951074748694163, "compression/movement_sparsity/importance_threshold": -3.17531969198058e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9231203123712189, "compression/movement_sparsity/model_sparsity": 0.8914082897910203, "compression_loss": 104.97412109375, "distillation_loss": 3.9022929668426514, "epoch": 4.49, "learning_rate": 3.8919849718221666e-05, "loss": 109.4979, "step": 5314, "task_loss": 2.8320107460021973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9951317754464083, "compression/movement_sparsity/importance_threshold": -3.159548265449421e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9231519352637892, "compression/movement_sparsity/model_sparsity": 0.8914388263399474, "compression_loss": 104.97590637207031, "distillation_loss": 5.155794620513916, "epoch": 4.49, "learning_rate": 3.89167188478397e-05, "loss": 109.8148, "step": 5315, "task_loss": 3.1778371334075928 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9951559954245457, "compression/movement_sparsity/importance_threshold": -3.1438291487425295e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9231753662531937, "compression/movement_sparsity/model_sparsity": 0.8914614524027836, "compression_loss": 104.97773742675781, "distillation_loss": 5.103322982788086, "epoch": 4.49, "learning_rate": 3.8913587977457736e-05, "loss": 109.7091, "step": 5316, "task_loss": 2.435004234313965 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9951801349377135, "compression/movement_sparsity/importance_threshold": -3.128162254966652e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9232103994577079, "compression/movement_sparsity/model_sparsity": 0.891495282108948, "compression_loss": 104.97955322265625, "distillation_loss": 2.7119407653808594, "epoch": 4.49, "learning_rate": 3.891045710707577e-05, "loss": 108.6346, "step": 5317, "task_loss": 1.9715889692306519 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9952041941197969, "compression/movement_sparsity/importance_threshold": -3.1125474972284496e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9232078834583368, "compression/movement_sparsity/model_sparsity": 0.8914928525418954, "compression_loss": 104.98138427734375, "distillation_loss": 4.23311710357666, "epoch": 4.5, "learning_rate": 3.89073262366938e-05, "loss": 108.8413, "step": 5318, "task_loss": 2.291778564453125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9952281731046811, "compression/movement_sparsity/importance_threshold": -3.0969847886343216e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9232371572898828, "compression/movement_sparsity/model_sparsity": 0.891521120727271, "compression_loss": 104.98318481445312, "distillation_loss": 7.256402969360352, "epoch": 4.5, "learning_rate": 3.890419536631184e-05, "loss": 109.8205, "step": 5319, "task_loss": 3.2907369136810303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9952520720262512, "compression/movement_sparsity/importance_threshold": -3.081474042291015e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923205486700642, "compression/movement_sparsity/model_sparsity": 0.8914905381202006, "compression_loss": 104.9850082397461, "distillation_loss": 3.632678508758545, "epoch": 4.5, "learning_rate": 3.890106449592987e-05, "loss": 108.8173, "step": 5320, "task_loss": 2.4619462490081787 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9952758910183924, "compression/movement_sparsity/importance_threshold": -3.066015171305104e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923231290599406, "compression/movement_sparsity/model_sparsity": 0.89151545557566, "compression_loss": 104.9867172241211, "distillation_loss": 2.4886250495910645, "epoch": 4.5, "learning_rate": 3.889793362554791e-05, "loss": 108.9896, "step": 5321, "task_loss": 2.262223958969116 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9952996302149898, "compression/movement_sparsity/importance_threshold": -3.0506080887831612e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9232429762836891, "compression/movement_sparsity/model_sparsity": 0.8915267398207387, "compression_loss": 104.9884262084961, "distillation_loss": 4.979306221008301, "epoch": 4.5, "learning_rate": 3.889480275516594e-05, "loss": 109.1671, "step": 5322, "task_loss": 1.7315515279769897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9953232897499285, "compression/movement_sparsity/importance_threshold": -3.035252707831934e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9233317278634029, "compression/movement_sparsity/model_sparsity": 0.8916124425106575, "compression_loss": 104.9902114868164, "distillation_loss": 3.23844575881958, "epoch": 4.5, "learning_rate": 3.889167188478397e-05, "loss": 108.4679, "step": 5323, "task_loss": 1.4200379848480225 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9953468697570936, "compression/movement_sparsity/importance_threshold": -3.0199489415578228e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9233527382467772, "compression/movement_sparsity/model_sparsity": 0.8916327311227275, "compression_loss": 104.99189758300781, "distillation_loss": 4.032686233520508, "epoch": 4.5, "learning_rate": 3.888854101440201e-05, "loss": 109.1486, "step": 5324, "task_loss": 2.0730175971984863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9953703703703703, "compression/movement_sparsity/importance_threshold": -3.004696703067574e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9233582829847279, "compression/movement_sparsity/model_sparsity": 0.891638085381872, "compression_loss": 104.99366760253906, "distillation_loss": 3.8643064498901367, "epoch": 4.5, "learning_rate": 3.888541014402004e-05, "loss": 108.5947, "step": 5325, "task_loss": 1.968390703201294 }, { "compression/movement_sparsity/importance_regularization_factor": 0.995393791723644, "compression/movement_sparsity/importance_threshold": -2.989495905467588e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9233584976197453, "compression/movement_sparsity/model_sparsity": 0.8916382926435162, "compression_loss": 104.9954605102539, "distillation_loss": 4.609438896179199, "epoch": 4.5, "learning_rate": 3.8882279273638074e-05, "loss": 109.1978, "step": 5326, "task_loss": 2.3824996948242188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9954171339507993, "compression/movement_sparsity/importance_threshold": -2.9743464618647852e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9234288382846293, "compression/movement_sparsity/model_sparsity": 0.8917062168901683, "compression_loss": 104.9971923828125, "distillation_loss": 3.4963865280151367, "epoch": 4.5, "learning_rate": 3.8879148403256106e-05, "loss": 108.7941, "step": 5327, "task_loss": 1.604846715927124 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9954403971857217, "compression/movement_sparsity/importance_threshold": -2.9592482853656524e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923355433108663, "compression/movement_sparsity/model_sparsity": 0.891635333407817, "compression_loss": 104.99887084960938, "distillation_loss": 4.464405536651611, "epoch": 4.5, "learning_rate": 3.8876017532874145e-05, "loss": 109.0303, "step": 5328, "task_loss": 2.488966941833496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9954635815622964, "compression/movement_sparsity/importance_threshold": -2.9442012890765895e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9233772185629336, "compression/movement_sparsity/model_sparsity": 0.8916563704647137, "compression_loss": 105.0005874633789, "distillation_loss": 2.9950201511383057, "epoch": 4.5, "learning_rate": 3.8872886662492176e-05, "loss": 108.5445, "step": 5329, "task_loss": 1.4457182884216309 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9954866872144084, "compression/movement_sparsity/importance_threshold": -2.9292053861044304e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923389583924772, "compression/movement_sparsity/model_sparsity": 0.8916683110383326, "compression_loss": 105.002197265625, "distillation_loss": 4.265384197235107, "epoch": 4.51, "learning_rate": 3.886975579211021e-05, "loss": 109.3405, "step": 5330, "task_loss": 2.5543572902679443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9955097142759428, "compression/movement_sparsity/importance_threshold": -2.9142604895556617e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9234191797088442, "compression/movement_sparsity/model_sparsity": 0.8916968901161746, "compression_loss": 105.0038070678711, "distillation_loss": 4.415205001831055, "epoch": 4.51, "learning_rate": 3.886662492172824e-05, "loss": 109.6637, "step": 5331, "task_loss": 2.355177164077759 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9955326628807848, "compression/movement_sparsity/importance_threshold": -2.899366512536944e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235075616393613, "compression/movement_sparsity/model_sparsity": 0.8917822358554838, "compression_loss": 105.00550079345703, "distillation_loss": 4.2797465324401855, "epoch": 4.51, "learning_rate": 3.886349405134628e-05, "loss": 109.3465, "step": 5332, "task_loss": 2.350774049758911 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9955555331628195, "compression/movement_sparsity/importance_threshold": -2.884523368154937e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235748735656657, "compression/movement_sparsity/model_sparsity": 0.891847235410044, "compression_loss": 105.00715637207031, "distillation_loss": 3.9964568614959717, "epoch": 4.51, "learning_rate": 3.886036318096431e-05, "loss": 109.1354, "step": 5333, "task_loss": 2.89021372795105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9955783252559322, "compression/movement_sparsity/importance_threshold": -2.869730969516128e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235912573719974, "compression/movement_sparsity/model_sparsity": 0.8918630563822256, "compression_loss": 105.00868225097656, "distillation_loss": 3.6269025802612305, "epoch": 4.51, "learning_rate": 3.885723231058234e-05, "loss": 109.0428, "step": 5334, "task_loss": 2.161698341369629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9956010392940078, "compression/movement_sparsity/importance_threshold": -2.854989229727177e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235287628094179, "compression/movement_sparsity/model_sparsity": 0.8918027087001265, "compression_loss": 105.01029205322266, "distillation_loss": 3.8611457347869873, "epoch": 4.51, "learning_rate": 3.885410144020038e-05, "loss": 108.8666, "step": 5335, "task_loss": 2.2263081073760986 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9956236754109316, "compression/movement_sparsity/importance_threshold": -2.8402980618947443e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923499381660363, "compression/movement_sparsity/model_sparsity": 0.8917743368839287, "compression_loss": 105.0118637084961, "distillation_loss": 3.1845884323120117, "epoch": 4.51, "learning_rate": 3.885097056981841e-05, "loss": 108.6325, "step": 5336, "task_loss": 1.425430417060852 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9956462337405887, "compression/movement_sparsity/importance_threshold": -2.82565737912523e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235764356316261, "compression/movement_sparsity/model_sparsity": 0.8918487438142331, "compression_loss": 105.01348876953125, "distillation_loss": 2.824215888977051, "epoch": 4.51, "learning_rate": 3.8847839699436444e-05, "loss": 109.1219, "step": 5337, "task_loss": 1.3678439855575562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9956687144168642, "compression/movement_sparsity/importance_threshold": -2.8110670945254677e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235980422233822, "compression/movement_sparsity/model_sparsity": 0.8918696081530928, "compression_loss": 105.01496124267578, "distillation_loss": 2.8060150146484375, "epoch": 4.51, "learning_rate": 3.8844708829054476e-05, "loss": 108.2126, "step": 5338, "task_loss": 3.1663877964019775 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9956911175736434, "compression/movement_sparsity/importance_threshold": -2.7965271212019444e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.92358973107854, "compression/movement_sparsity/model_sparsity": 0.8918615825216439, "compression_loss": 105.01653289794922, "distillation_loss": 5.371906757354736, "epoch": 4.51, "learning_rate": 3.8841577958672514e-05, "loss": 109.3051, "step": 5339, "task_loss": 2.2729039192199707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9957134433448112, "compression/movement_sparsity/importance_threshold": -2.78203737226132e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235893852776785, "compression/movement_sparsity/model_sparsity": 0.891861248600106, "compression_loss": 105.01805114746094, "distillation_loss": 4.993561744689941, "epoch": 4.51, "learning_rate": 3.8838447088290546e-05, "loss": 109.1469, "step": 5340, "task_loss": 2.5543055534362793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9957356918642528, "compression/movement_sparsity/importance_threshold": -2.7675977608102552e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235663239374708, "compression/movement_sparsity/model_sparsity": 0.8918389794878793, "compression_loss": 105.01959991455078, "distillation_loss": 3.4479775428771973, "epoch": 4.51, "learning_rate": 3.883531621790858e-05, "loss": 108.8961, "step": 5341, "task_loss": 2.082054615020752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9957578632658536, "compression/movement_sparsity/importance_threshold": -2.753208199955063e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9236088812917632, "compression/movement_sparsity/model_sparsity": 0.89188007486613, "compression_loss": 105.02119445800781, "distillation_loss": 5.277400970458984, "epoch": 4.52, "learning_rate": 3.883218534752661e-05, "loss": 109.8288, "step": 5342, "task_loss": 2.5926430225372314 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9957799576834985, "compression/movement_sparsity/importance_threshold": -2.738868602802664e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235472214209182, "compression/movement_sparsity/model_sparsity": 0.8918205332015364, "compression_loss": 105.02271270751953, "distillation_loss": 3.324037551879883, "epoch": 4.52, "learning_rate": 3.882905447714465e-05, "loss": 109.018, "step": 5343, "task_loss": 1.4740777015686035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9958019752510725, "compression/movement_sparsity/importance_threshold": -2.7245788824596315e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235293232452967, "compression/movement_sparsity/model_sparsity": 0.8918032498833088, "compression_loss": 105.02433013916016, "distillation_loss": 4.5581769943237305, "epoch": 4.52, "learning_rate": 3.882592360676268e-05, "loss": 109.7447, "step": 5344, "task_loss": 3.1742050647735596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9958239161024611, "compression/movement_sparsity/importance_threshold": -2.7103389520323658e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923521226735472, "compression/movement_sparsity/model_sparsity": 0.8917954315135044, "compression_loss": 105.02586364746094, "distillation_loss": 4.124095439910889, "epoch": 4.52, "learning_rate": 3.882279273638071e-05, "loss": 109.9077, "step": 5345, "task_loss": 3.359849691390991 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9958457803715492, "compression/movement_sparsity/importance_threshold": -2.6961487246277005e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9234857404125877, "compression/movement_sparsity/model_sparsity": 0.8917611642549798, "compression_loss": 105.02743530273438, "distillation_loss": 4.745896339416504, "epoch": 4.52, "learning_rate": 3.881966186599875e-05, "loss": 109.3974, "step": 5346, "task_loss": 2.79905366897583 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9958675681922221, "compression/movement_sparsity/importance_threshold": -2.6820081133520357e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923531898865506, "compression/movement_sparsity/model_sparsity": 0.8918057370230404, "compression_loss": 105.02896881103516, "distillation_loss": 4.567929267883301, "epoch": 4.52, "learning_rate": 3.881653099561678e-05, "loss": 109.28, "step": 5347, "task_loss": 2.4269886016845703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9958892796983647, "compression/movement_sparsity/importance_threshold": -2.667917031312205e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235469113925596, "compression/movement_sparsity/model_sparsity": 0.8918202338236058, "compression_loss": 105.03050994873047, "distillation_loss": 2.949063301086426, "epoch": 4.52, "learning_rate": 3.8813400125234814e-05, "loss": 108.5808, "step": 5348, "task_loss": 1.89532470703125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9959109150238624, "compression/movement_sparsity/importance_threshold": -2.653875391614522e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235925094095991, "compression/movement_sparsity/model_sparsity": 0.8918642654084841, "compression_loss": 105.03202819824219, "distillation_loss": 6.029988765716553, "epoch": 4.52, "learning_rate": 3.8810269254852846e-05, "loss": 109.225, "step": 5349, "task_loss": 3.6671972274780273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9959324743026002, "compression/movement_sparsity/importance_threshold": -2.6398831073658202e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9235862611457579, "compression/movement_sparsity/model_sparsity": 0.8918582317917277, "compression_loss": 105.03336334228516, "distillation_loss": 3.80770206451416, "epoch": 4.52, "learning_rate": 3.8807138384470884e-05, "loss": 108.6907, "step": 5350, "task_loss": 3.520747423171997 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9959539576684633, "compression/movement_sparsity/importance_threshold": -2.625940091672673e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9236334093045901, "compression/movement_sparsity/model_sparsity": 0.8919037602662594, "compression_loss": 105.0348129272461, "distillation_loss": 3.414616823196411, "epoch": 4.52, "learning_rate": 3.8804007514088916e-05, "loss": 108.25, "step": 5351, "task_loss": 1.5243449211120605 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9959753652553368, "compression/movement_sparsity/importance_threshold": -2.6120462576415676e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9236657595713863, "compression/movement_sparsity/model_sparsity": 0.8919349992018699, "compression_loss": 105.03612518310547, "distillation_loss": 3.2055225372314453, "epoch": 4.52, "learning_rate": 3.8800876643706955e-05, "loss": 109.3166, "step": 5352, "task_loss": 1.5317103862762451 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9959966971971058, "compression/movement_sparsity/importance_threshold": -2.5982015183792506e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9236921239060292, "compression/movement_sparsity/model_sparsity": 0.8919604578405117, "compression_loss": 105.03749084472656, "distillation_loss": 3.7468931674957275, "epoch": 4.52, "learning_rate": 3.8797745773324987e-05, "loss": 109.4554, "step": 5353, "task_loss": 2.1353678703308105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9960179536276556, "compression/movement_sparsity/importance_threshold": -2.584405786992209e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9236549681996759, "compression/movement_sparsity/model_sparsity": 0.8919245785469759, "compression_loss": 105.0387954711914, "distillation_loss": 4.468011856079102, "epoch": 4.53, "learning_rate": 3.879461490294302e-05, "loss": 109.3868, "step": 5354, "task_loss": 3.340430736541748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.996039134680871, "compression/movement_sparsity/importance_threshold": -2.5706589765871896e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237054312771108, "compression/movement_sparsity/model_sparsity": 0.8919733080624583, "compression_loss": 105.04013061523438, "distillation_loss": 4.155969142913818, "epoch": 4.53, "learning_rate": 3.879148403256106e-05, "loss": 109.4963, "step": 5355, "task_loss": 2.6551907062530518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9960602404906376, "compression/movement_sparsity/importance_threshold": -2.5569610002705925e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237654575369897, "compression/movement_sparsity/model_sparsity": 0.8920312722356482, "compression_loss": 105.04146575927734, "distillation_loss": 4.151162147521973, "epoch": 4.53, "learning_rate": 3.878835316217909e-05, "loss": 109.1514, "step": 5356, "task_loss": 1.7516695261001587 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9960812711908402, "compression/movement_sparsity/importance_threshold": -2.543311771149165e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923766793043765, "compression/movement_sparsity/model_sparsity": 0.8920325618636571, "compression_loss": 105.04277801513672, "distillation_loss": 2.7479889392852783, "epoch": 4.53, "learning_rate": 3.878522229179713e-05, "loss": 108.4572, "step": 5357, "task_loss": 1.6538057327270508 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9961022269153641, "compression/movement_sparsity/importance_threshold": -2.5297112023295668e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237912733599214, "compression/movement_sparsity/model_sparsity": 0.8920562012056433, "compression_loss": 105.04412078857422, "distillation_loss": 3.905812978744507, "epoch": 4.53, "learning_rate": 3.878209142141516e-05, "loss": 109.0654, "step": 5358, "task_loss": 2.156081199645996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9961231077980944, "compression/movement_sparsity/importance_threshold": -2.5161592069181983e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237805535332168, "compression/movement_sparsity/model_sparsity": 0.892045849637964, "compression_loss": 105.04542541503906, "distillation_loss": 4.070124626159668, "epoch": 4.53, "learning_rate": 3.877896055103319e-05, "loss": 109.428, "step": 5359, "task_loss": 2.3347327709198 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9961439139729162, "compression/movement_sparsity/importance_threshold": -2.5026556980218065e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237374595913808, "compression/movement_sparsity/model_sparsity": 0.8920042361056025, "compression_loss": 105.04668426513672, "distillation_loss": 4.562074661254883, "epoch": 4.53, "learning_rate": 3.877582968065122e-05, "loss": 109.4048, "step": 5360, "task_loss": 2.9916505813598633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9961646455737146, "compression/movement_sparsity/importance_threshold": -2.4892005887470516e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237441132769215, "compression/movement_sparsity/model_sparsity": 0.8920106612165759, "compression_loss": 105.04794311523438, "distillation_loss": 2.55051851272583, "epoch": 4.53, "learning_rate": 3.877269881026926e-05, "loss": 108.7109, "step": 5361, "task_loss": 1.061310887336731 }, { "compression/movement_sparsity/importance_regularization_factor": 0.996185302734375, "compression/movement_sparsity/importance_threshold": -2.4757937922004203e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237074941581118, "compression/movement_sparsity/model_sparsity": 0.8919753000771509, "compression_loss": 105.04924011230469, "distillation_loss": 5.152840614318848, "epoch": 4.53, "learning_rate": 3.876956793988729e-05, "loss": 109.6232, "step": 5362, "task_loss": 2.7339820861816406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9962058855887823, "compression/movement_sparsity/importance_threshold": -2.4624352214884862e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237277533189251, "compression/movement_sparsity/model_sparsity": 0.8919948632734658, "compression_loss": 105.05044555664062, "distillation_loss": 4.604339122772217, "epoch": 4.53, "learning_rate": 3.8766437069505325e-05, "loss": 109.8719, "step": 5363, "task_loss": 2.188149929046631 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9962263942708216, "compression/movement_sparsity/importance_threshold": -2.449124789718083e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237746749185722, "compression/movement_sparsity/model_sparsity": 0.8920401729718173, "compression_loss": 105.05174255371094, "distillation_loss": 3.9044249057769775, "epoch": 4.53, "learning_rate": 3.8763306199123356e-05, "loss": 108.8073, "step": 5364, "task_loss": 2.0767111778259277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9962468289143782, "compression/movement_sparsity/importance_threshold": -2.435862409995524e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.92373827043478, "compression/movement_sparsity/model_sparsity": 0.8920050190940365, "compression_loss": 105.05297088623047, "distillation_loss": 3.1993820667266846, "epoch": 4.53, "learning_rate": 3.8760175328741395e-05, "loss": 108.5069, "step": 5365, "task_loss": 1.7538849115371704 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9962671896533372, "compression/movement_sparsity/importance_threshold": -2.4226479954275563e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9237709545782699, "compression/movement_sparsity/model_sparsity": 0.8920365804366494, "compression_loss": 105.05415344238281, "distillation_loss": 4.960831165313721, "epoch": 4.54, "learning_rate": 3.875704445835943e-05, "loss": 109.2954, "step": 5366, "task_loss": 2.5524208545684814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9962874766215837, "compression/movement_sparsity/importance_threshold": -2.4094814591207533e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9238542806617093, "compression/movement_sparsity/model_sparsity": 0.8921170440127817, "compression_loss": 105.05535125732422, "distillation_loss": 3.464226484298706, "epoch": 4.54, "learning_rate": 3.875391358797746e-05, "loss": 109.3566, "step": 5367, "task_loss": 2.134813070297241 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9963076899530029, "compression/movement_sparsity/importance_threshold": -2.3963627141816886e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9238882406911362, "compression/movement_sparsity/model_sparsity": 0.8921498374107245, "compression_loss": 105.05656433105469, "distillation_loss": 4.896143913269043, "epoch": 4.54, "learning_rate": 3.875078271759549e-05, "loss": 109.4426, "step": 5368, "task_loss": 2.6203103065490723 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9963278297814798, "compression/movement_sparsity/importance_threshold": -2.383291673717109e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9238818016406128, "compression/movement_sparsity/model_sparsity": 0.8921436195613954, "compression_loss": 105.05778503417969, "distillation_loss": 3.1612913608551025, "epoch": 4.54, "learning_rate": 3.874765184721353e-05, "loss": 109.2007, "step": 5369, "task_loss": 1.4873918294906616 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9963478962408997, "compression/movement_sparsity/importance_threshold": -2.3702682508335016e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9239093703161869, "compression/movement_sparsity/model_sparsity": 0.8921702411681525, "compression_loss": 105.0589599609375, "distillation_loss": 4.689936637878418, "epoch": 4.54, "learning_rate": 3.874452097683156e-05, "loss": 109.9295, "step": 5370, "task_loss": 1.9766478538513184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9963678894651476, "compression/movement_sparsity/importance_threshold": -2.3572923586375263e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9238760661159799, "compression/movement_sparsity/model_sparsity": 0.8921380810696783, "compression_loss": 105.0601577758789, "distillation_loss": 2.923607349395752, "epoch": 4.54, "learning_rate": 3.874139010644959e-05, "loss": 108.4874, "step": 5371, "task_loss": 2.089992046356201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9963878095881089, "compression/movement_sparsity/importance_threshold": -2.3443639102356702e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9238640107825001, "compression/movement_sparsity/model_sparsity": 0.89212643987399, "compression_loss": 105.06135559082031, "distillation_loss": 3.319979190826416, "epoch": 4.54, "learning_rate": 3.873825923606763e-05, "loss": 109.0253, "step": 5372, "task_loss": 1.444828748703003 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9964076567436684, "compression/movement_sparsity/importance_threshold": -2.33148281873468e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.923841486029836, "compression/movement_sparsity/model_sparsity": 0.8921046889158741, "compression_loss": 105.0625, "distillation_loss": 3.4745864868164062, "epoch": 4.54, "learning_rate": 3.873512836568566e-05, "loss": 108.9085, "step": 5373, "task_loss": 2.89908766746521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9964274310657115, "compression/movement_sparsity/importance_threshold": -2.3186489972410428e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9239579136026326, "compression/movement_sparsity/model_sparsity": 0.892217116843372, "compression_loss": 105.06367492675781, "distillation_loss": 4.011798858642578, "epoch": 4.54, "learning_rate": 3.8731997495303695e-05, "loss": 108.9356, "step": 5374, "task_loss": 2.025961399078369 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9964471326881232, "compression/movement_sparsity/importance_threshold": -2.3058623588614185e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9239522257746703, "compression/movement_sparsity/model_sparsity": 0.892211624409798, "compression_loss": 105.06478118896484, "distillation_loss": 4.176133632659912, "epoch": 4.54, "learning_rate": 3.8728866624921726e-05, "loss": 109.7125, "step": 5375, "task_loss": 1.7605586051940918 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9964667617447887, "compression/movement_sparsity/importance_threshold": -2.2931228167024677e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9239445823832156, "compression/movement_sparsity/model_sparsity": 0.8922042435923537, "compression_loss": 105.06596374511719, "distillation_loss": 5.654709815979004, "epoch": 4.54, "learning_rate": 3.8725735754539765e-05, "loss": 109.714, "step": 5376, "task_loss": 3.1003432273864746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9964863183695931, "compression/movement_sparsity/importance_threshold": -2.280430283870677e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9240040124347128, "compression/movement_sparsity/model_sparsity": 0.8922616320387536, "compression_loss": 105.06709289550781, "distillation_loss": 3.419468879699707, "epoch": 4.54, "learning_rate": 3.87226048841578e-05, "loss": 108.0293, "step": 5377, "task_loss": 2.3688979148864746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9965058026964216, "compression/movement_sparsity/importance_threshold": -2.2677846734727065e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.924091046934287, "compression/movement_sparsity/model_sparsity": 0.892345676635518, "compression_loss": 105.06820678710938, "distillation_loss": 2.858613967895508, "epoch": 4.55, "learning_rate": 3.871947401377583e-05, "loss": 108.3221, "step": 5378, "task_loss": 2.2977075576782227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9965252148591592, "compression/movement_sparsity/importance_threshold": -2.2551858986153035e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9241011705526098, "compression/movement_sparsity/model_sparsity": 0.8923554524764076, "compression_loss": 105.06927490234375, "distillation_loss": 5.770139217376709, "epoch": 4.55, "learning_rate": 3.871634314339386e-05, "loss": 109.9744, "step": 5379, "task_loss": 4.273530960083008 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9965445549916914, "compression/movement_sparsity/importance_threshold": -2.2426338724047812e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9240992507616205, "compression/movement_sparsity/model_sparsity": 0.8923535986361447, "compression_loss": 105.07042694091797, "distillation_loss": 4.4093017578125, "epoch": 4.55, "learning_rate": 3.87132122730119e-05, "loss": 109.0312, "step": 5380, "task_loss": 2.9445528984069824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9965638232279028, "compression/movement_sparsity/importance_threshold": -2.2301285079479732e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9240649807038349, "compression/movement_sparsity/model_sparsity": 0.8923205058602712, "compression_loss": 105.07157897949219, "distillation_loss": 3.26068115234375, "epoch": 4.55, "learning_rate": 3.871008140262993e-05, "loss": 109.2206, "step": 5381, "task_loss": 2.1686394214630127 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9965830197016791, "compression/movement_sparsity/importance_threshold": -2.2176697183512797e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9241093863041109, "compression/movement_sparsity/model_sparsity": 0.8923633859915701, "compression_loss": 105.07266998291016, "distillation_loss": 3.894324779510498, "epoch": 4.55, "learning_rate": 3.870695053224796e-05, "loss": 109.0356, "step": 5382, "task_loss": 2.005593776702881 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9966021445469049, "compression/movement_sparsity/importance_threshold": -2.205257416721621e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9240938968103519, "compression/movement_sparsity/model_sparsity": 0.892348428609573, "compression_loss": 105.0737533569336, "distillation_loss": 4.804529190063477, "epoch": 4.55, "learning_rate": 3.8703819661866e-05, "loss": 109.3605, "step": 5383, "task_loss": 3.1910228729248047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9966211978974657, "compression/movement_sparsity/importance_threshold": -2.1928915161653108e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9241200703583127, "compression/movement_sparsity/model_sparsity": 0.892373703015642, "compression_loss": 105.07485961914062, "distillation_loss": 4.354255676269531, "epoch": 4.55, "learning_rate": 3.870068879148403e-05, "loss": 109.2824, "step": 5384, "task_loss": 2.2885844707489014 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9966401798872466, "compression/movement_sparsity/importance_threshold": -2.180571929789009e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9241595274290197, "compression/movement_sparsity/model_sparsity": 0.8924118046145861, "compression_loss": 105.07595825195312, "distillation_loss": 2.5335981845855713, "epoch": 4.55, "learning_rate": 3.8697557921102065e-05, "loss": 108.609, "step": 5385, "task_loss": 2.231628656387329 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9966590906501327, "compression/movement_sparsity/importance_threshold": -2.1682985706994624e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9242037303184458, "compression/movement_sparsity/model_sparsity": 0.8924544889987764, "compression_loss": 105.0770263671875, "distillation_loss": 4.4565911293029785, "epoch": 4.55, "learning_rate": 3.86944270507201e-05, "loss": 109.5951, "step": 5386, "task_loss": 3.623366594314575 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9966779303200091, "compression/movement_sparsity/importance_threshold": -2.1560713520031582e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9242962976318031, "compression/movement_sparsity/model_sparsity": 0.8925438763401494, "compression_loss": 105.07804870605469, "distillation_loss": 4.1281843185424805, "epoch": 4.55, "learning_rate": 3.8691296180338135e-05, "loss": 109.5902, "step": 5387, "task_loss": 1.8519588708877563 }, { "compression/movement_sparsity/importance_regularization_factor": 0.996696699030761, "compression/movement_sparsity/importance_threshold": -2.1438901868066698e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9242684427762057, "compression/movement_sparsity/model_sparsity": 0.8925169783845335, "compression_loss": 105.07907104492188, "distillation_loss": 3.5561437606811523, "epoch": 4.55, "learning_rate": 3.8688165309956173e-05, "loss": 109.2162, "step": 5388, "task_loss": 1.629446268081665 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9967153969162734, "compression/movement_sparsity/importance_threshold": -2.1317549882166574e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9242292241888513, "compression/movement_sparsity/model_sparsity": 0.8924791070763052, "compression_loss": 105.08018493652344, "distillation_loss": 5.73874568939209, "epoch": 4.56, "learning_rate": 3.8685034439574205e-05, "loss": 109.1001, "step": 5389, "task_loss": 3.3973846435546875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9967340241104317, "compression/movement_sparsity/importance_threshold": -2.1196656693397813e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9242533944766492, "compression/movement_sparsity/model_sparsity": 0.8925024470403607, "compression_loss": 105.08118438720703, "distillation_loss": 4.768984317779541, "epoch": 4.56, "learning_rate": 3.868190356919224e-05, "loss": 109.2747, "step": 5390, "task_loss": 4.146753311157227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9967525807471208, "compression/movement_sparsity/importance_threshold": -2.1076221432825282e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9243144700632802, "compression/movement_sparsity/model_sparsity": 0.8925614244927004, "compression_loss": 105.08226013183594, "distillation_loss": 5.163151741027832, "epoch": 4.56, "learning_rate": 3.8678772698810276e-05, "loss": 109.4935, "step": 5391, "task_loss": 2.5901589393615723 }, { "compression/movement_sparsity/importance_regularization_factor": 0.996771066960226, "compression/movement_sparsity/importance_threshold": -2.0956243231514717e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9243609027720542, "compression/movement_sparsity/model_sparsity": 0.8926062620950843, "compression_loss": 105.08326721191406, "distillation_loss": 3.53347110748291, "epoch": 4.56, "learning_rate": 3.867564182842831e-05, "loss": 108.6019, "step": 5392, "task_loss": 2.0257177352905273 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9967894828836323, "compression/movement_sparsity/importance_threshold": -2.0836721220534454e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.924432853199569, "compression/movement_sparsity/model_sparsity": 0.8926757408040685, "compression_loss": 105.08433532714844, "distillation_loss": 2.8883910179138184, "epoch": 4.56, "learning_rate": 3.867251095804634e-05, "loss": 108.4581, "step": 5393, "task_loss": 1.5464166402816772 }, { "compression/movement_sparsity/importance_regularization_factor": 0.996807828651225, "compression/movement_sparsity/importance_threshold": -2.0717654530948496e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9244846756321144, "compression/movement_sparsity/model_sparsity": 0.8927257829766316, "compression_loss": 105.08536529541016, "distillation_loss": 2.789808750152588, "epoch": 4.56, "learning_rate": 3.866938008766438e-05, "loss": 108.9332, "step": 5394, "task_loss": 1.5644052028656006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9968261043968891, "compression/movement_sparsity/importance_threshold": -2.0599042293823443e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9244790831974932, "compression/movement_sparsity/model_sparsity": 0.892720382659344, "compression_loss": 105.08641815185547, "distillation_loss": 5.797066688537598, "epoch": 4.56, "learning_rate": 3.866624921728241e-05, "loss": 109.4659, "step": 5395, "task_loss": 2.8911988735198975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9968443102545098, "compression/movement_sparsity/importance_threshold": -2.048088364022503e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9244808479743033, "compression/movement_sparsity/model_sparsity": 0.8927220868106416, "compression_loss": 105.08739471435547, "distillation_loss": 3.254239559173584, "epoch": 4.56, "learning_rate": 3.866311834690044e-05, "loss": 108.7768, "step": 5396, "task_loss": 1.4735738039016724 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9968624463579723, "compression/movement_sparsity/importance_threshold": -2.0363177701219863e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9244961109088772, "compression/movement_sparsity/model_sparsity": 0.8927368254164586, "compression_loss": 105.08851623535156, "distillation_loss": 4.360401630401611, "epoch": 4.56, "learning_rate": 3.865998747651847e-05, "loss": 108.9221, "step": 5397, "task_loss": 1.9387401342391968 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9968805128411616, "compression/movement_sparsity/importance_threshold": -2.0245923607873674e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9245324915443343, "compression/movement_sparsity/model_sparsity": 0.8927719562651677, "compression_loss": 105.08960723876953, "distillation_loss": 4.3813347816467285, "epoch": 4.56, "learning_rate": 3.865685660613651e-05, "loss": 108.9444, "step": 5398, "task_loss": 1.942732572555542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9968985098379629, "compression/movement_sparsity/importance_threshold": -2.0129120491253934e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9244500955459705, "compression/movement_sparsity/model_sparsity": 0.8926923908228275, "compression_loss": 105.09059143066406, "distillation_loss": 4.57605504989624, "epoch": 4.56, "learning_rate": 3.865372573575454e-05, "loss": 108.9765, "step": 5399, "task_loss": 2.224964141845703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9969164374822614, "compression/movement_sparsity/importance_threshold": -2.0012767482423775e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9244345345072056, "compression/movement_sparsity/model_sparsity": 0.8926773643536156, "compression_loss": 105.09156799316406, "distillation_loss": 3.0222039222717285, "epoch": 4.56, "learning_rate": 3.8650594865372575e-05, "loss": 108.7984, "step": 5400, "task_loss": 1.835209846496582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9969342959079422, "compression/movement_sparsity/importance_threshold": -1.989686371245067e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9245444037878024, "compression/movement_sparsity/model_sparsity": 0.8927834592864264, "compression_loss": 105.09263610839844, "distillation_loss": 4.198319911956787, "epoch": 4.57, "learning_rate": 3.864746399499061e-05, "loss": 109.5881, "step": 5401, "task_loss": 2.307222366333008 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9969520852488906, "compression/movement_sparsity/importance_threshold": -1.9781408312401216e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9245429013426804, "compression/movement_sparsity/model_sparsity": 0.8927820084549164, "compression_loss": 105.09371948242188, "distillation_loss": 7.26679801940918, "epoch": 4.57, "learning_rate": 3.8644333124608645e-05, "loss": 110.131, "step": 5402, "task_loss": 3.252574920654297 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9969698056389914, "compression/movement_sparsity/importance_threshold": -1.9666400413341154e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9245748938844474, "compression/movement_sparsity/model_sparsity": 0.8928129019544532, "compression_loss": 105.0947494506836, "distillation_loss": 4.190859317779541, "epoch": 4.57, "learning_rate": 3.864120225422668e-05, "loss": 108.6939, "step": 5403, "task_loss": 2.3342881202697754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9969874572121299, "compression/movement_sparsity/importance_threshold": -1.9551839146336215e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9246180712954568, "compression/movement_sparsity/model_sparsity": 0.8928545960885652, "compression_loss": 105.09579467773438, "distillation_loss": 3.4124059677124023, "epoch": 4.57, "learning_rate": 3.863807138384471e-05, "loss": 108.4328, "step": 5404, "task_loss": 2.061204671859741 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9970050401021913, "compression/movement_sparsity/importance_threshold": -1.9437723642453004e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9246001731198354, "compression/movement_sparsity/model_sparsity": 0.8928373127703375, "compression_loss": 105.09676361083984, "distillation_loss": 3.9953110218048096, "epoch": 4.57, "learning_rate": 3.863494051346274e-05, "loss": 109.0441, "step": 5405, "task_loss": 2.9763576984405518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9970225544430608, "compression/movement_sparsity/importance_threshold": -1.932405303275639e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9245848028677528, "compression/movement_sparsity/model_sparsity": 0.8928224705336983, "compression_loss": 105.0977783203125, "distillation_loss": 5.719180107116699, "epoch": 4.57, "learning_rate": 3.863180964308078e-05, "loss": 109.3453, "step": 5406, "task_loss": 3.1963179111480713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9970400003686233, "compression/movement_sparsity/importance_threshold": -1.9210826448313838e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.92462405722761, "compression/movement_sparsity/model_sparsity": 0.8928603763855341, "compression_loss": 105.09880828857422, "distillation_loss": 3.210707664489746, "epoch": 4.57, "learning_rate": 3.862867877269881e-05, "loss": 109.2258, "step": 5407, "task_loss": 1.8107291460037231 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9970573780127642, "compression/movement_sparsity/importance_threshold": -1.909804302019022e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247170776593374, "compression/movement_sparsity/model_sparsity": 0.8929502012792673, "compression_loss": 105.09980010986328, "distillation_loss": 4.035536766052246, "epoch": 4.57, "learning_rate": 3.862554790231684e-05, "loss": 108.533, "step": 5408, "task_loss": 2.2668843269348145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9970746875093683, "compression/movement_sparsity/importance_threshold": -1.8985701879453004e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247460414625249, "compression/movement_sparsity/model_sparsity": 0.8929781700867122, "compression_loss": 105.10071563720703, "distillation_loss": 3.6903271675109863, "epoch": 4.57, "learning_rate": 3.862241703193488e-05, "loss": 109.2876, "step": 5409, "task_loss": 1.880687952041626 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9970919289923212, "compression/movement_sparsity/importance_threshold": -1.887380215716706e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247761738341408, "compression/movement_sparsity/model_sparsity": 0.893007267318665, "compression_loss": 105.1016845703125, "distillation_loss": 3.3730239868164062, "epoch": 4.57, "learning_rate": 3.861928616155291e-05, "loss": 108.8686, "step": 5410, "task_loss": 0.9647518992424011 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9971091025955078, "compression/movement_sparsity/importance_threshold": -1.876234298439812e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.924715944863412, "compression/movement_sparsity/model_sparsity": 0.8929491073983667, "compression_loss": 105.10261535644531, "distillation_loss": 3.067903995513916, "epoch": 4.57, "learning_rate": 3.8616155291170945e-05, "loss": 108.8011, "step": 5411, "task_loss": 1.9716331958770752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9971262084528132, "compression/movement_sparsity/importance_threshold": -1.865132349221279e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247323644422466, "compression/movement_sparsity/model_sparsity": 0.8929649629141558, "compression_loss": 105.10358428955078, "distillation_loss": 3.9024672508239746, "epoch": 4.57, "learning_rate": 3.861302442078898e-05, "loss": 108.7685, "step": 5412, "task_loss": 2.5585455894470215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9971432466981225, "compression/movement_sparsity/importance_threshold": -1.854074281167767e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247444555482294, "compression/movement_sparsity/model_sparsity": 0.8929766386534514, "compression_loss": 105.10455322265625, "distillation_loss": 4.257826328277588, "epoch": 4.58, "learning_rate": 3.8609893550407015e-05, "loss": 109.8517, "step": 5413, "task_loss": 2.326820135116577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.997160217465321, "compression/movement_sparsity/importance_threshold": -1.8430600073857627e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248090606884805, "compression/movement_sparsity/model_sparsity": 0.8930390244083863, "compression_loss": 105.10546875, "distillation_loss": 5.346103668212891, "epoch": 4.58, "learning_rate": 3.860676268002505e-05, "loss": 109.111, "step": 5414, "task_loss": 3.3305671215057373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9971771208882938, "compression/movement_sparsity/importance_threshold": -1.8320894409820133e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248871043656571, "compression/movement_sparsity/model_sparsity": 0.8931143870451616, "compression_loss": 105.10639190673828, "distillation_loss": 3.5003116130828857, "epoch": 4.58, "learning_rate": 3.860363180964308e-05, "loss": 109.019, "step": 5415, "task_loss": 1.4780279397964478 }, { "compression/movement_sparsity/importance_regularization_factor": 0.997193957100926, "compression/movement_sparsity/importance_threshold": -1.8211624950630056e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248415540452882, "compression/movement_sparsity/model_sparsity": 0.8930704015184264, "compression_loss": 105.10734558105469, "distillation_loss": 4.190576553344727, "epoch": 4.58, "learning_rate": 3.860050093926111e-05, "loss": 108.7538, "step": 5416, "task_loss": 2.012813091278076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9972107262371027, "compression/movement_sparsity/importance_threshold": -1.810279082735313e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248124233037538, "compression/movement_sparsity/model_sparsity": 0.8930422715074804, "compression_loss": 105.1082763671875, "distillation_loss": 3.7812423706054688, "epoch": 4.58, "learning_rate": 3.859737006887915e-05, "loss": 109.1974, "step": 5417, "task_loss": 2.917998790740967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9972274284307091, "compression/movement_sparsity/importance_threshold": -1.7994391171056827e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248024070029396, "compression/movement_sparsity/model_sparsity": 0.893032599297413, "compression_loss": 105.1091537475586, "distillation_loss": 3.6733474731445312, "epoch": 4.58, "learning_rate": 3.859423919849718e-05, "loss": 109.0382, "step": 5418, "task_loss": 2.1816415786743164 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9972440638156304, "compression/movement_sparsity/importance_threshold": -1.7886425112806012e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247843895856418, "compression/movement_sparsity/model_sparsity": 0.8930152008338275, "compression_loss": 105.11002349853516, "distillation_loss": 4.892642974853516, "epoch": 4.58, "learning_rate": 3.859110832811522e-05, "loss": 109.8337, "step": 5419, "task_loss": 3.4692134857177734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9972606325257516, "compression/movement_sparsity/importance_threshold": -1.777889178366729e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247656924907888, "compression/movement_sparsity/model_sparsity": 0.8929971460417016, "compression_loss": 105.11089324951172, "distillation_loss": 4.056175708770752, "epoch": 4.58, "learning_rate": 3.858797745773325e-05, "loss": 109.2173, "step": 5420, "task_loss": 2.679232597351074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9972771346949579, "compression/movement_sparsity/importance_threshold": -1.767179031470726e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247870844475276, "compression/movement_sparsity/model_sparsity": 0.893017803118917, "compression_loss": 105.11178588867188, "distillation_loss": 4.300522327423096, "epoch": 4.58, "learning_rate": 3.858484658735128e-05, "loss": 109.311, "step": 5421, "task_loss": 3.140900135040283 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9972935704571345, "compression/movement_sparsity/importance_threshold": -1.7565119836989926e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248387041692232, "compression/movement_sparsity/model_sparsity": 0.8930676495443716, "compression_loss": 105.11267852783203, "distillation_loss": 5.671657562255859, "epoch": 4.58, "learning_rate": 3.858171571696932e-05, "loss": 109.0296, "step": 5422, "task_loss": 2.3730099201202393 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9973099399461665, "compression/movement_sparsity/importance_threshold": -1.7458879481583624e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248271377266164, "compression/movement_sparsity/model_sparsity": 0.8930564804446508, "compression_loss": 105.11356353759766, "distillation_loss": 4.516895294189453, "epoch": 4.58, "learning_rate": 3.8578584846587354e-05, "loss": 109.5112, "step": 5423, "task_loss": 2.1537418365478516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.997326243295939, "compression/movement_sparsity/importance_threshold": -1.735306837955322e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248265534424023, "compression/movement_sparsity/model_sparsity": 0.8930559162323969, "compression_loss": 105.1144790649414, "distillation_loss": 5.598310470581055, "epoch": 4.58, "learning_rate": 3.857545397620539e-05, "loss": 109.8144, "step": 5424, "task_loss": 3.2304656505584717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9973424806403371, "compression/movement_sparsity/importance_threshold": -1.724768566196532e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248168471699467, "compression/movement_sparsity/model_sparsity": 0.8930465434002601, "compression_loss": 105.11534118652344, "distillation_loss": 3.8078513145446777, "epoch": 4.59, "learning_rate": 3.8572323105823424e-05, "loss": 109.1073, "step": 5425, "task_loss": 1.538874864578247 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9973586521132461, "compression/movement_sparsity/importance_threshold": -1.7142730459884792e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248475280532737, "compression/movement_sparsity/model_sparsity": 0.8930761703008595, "compression_loss": 105.11625671386719, "distillation_loss": 5.0467376708984375, "epoch": 4.59, "learning_rate": 3.8569192235441456e-05, "loss": 109.4598, "step": 5426, "task_loss": 2.5639138221740723 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9973747578485511, "compression/movement_sparsity/importance_threshold": -1.7038201904379102e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248505687160209, "compression/movement_sparsity/model_sparsity": 0.8930791065074871, "compression_loss": 105.11711883544922, "distillation_loss": 5.840982437133789, "epoch": 4.59, "learning_rate": 3.856606136505949e-05, "loss": 110.2651, "step": 5427, "task_loss": 3.1429085731506348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9973907979801372, "compression/movement_sparsity/importance_threshold": -1.693409912651312e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248707563318284, "compression/movement_sparsity/model_sparsity": 0.8930986006165873, "compression_loss": 105.1180191040039, "distillation_loss": 4.505388259887695, "epoch": 4.59, "learning_rate": 3.8562930494677526e-05, "loss": 108.846, "step": 5428, "task_loss": 2.1659722328186035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9974067726418895, "compression/movement_sparsity/importance_threshold": -1.683042125735345e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248457990489666, "compression/movement_sparsity/model_sparsity": 0.8930745006931693, "compression_loss": 105.11890411376953, "distillation_loss": 3.876424551010132, "epoch": 4.59, "learning_rate": 3.855979962429556e-05, "loss": 109.0756, "step": 5429, "task_loss": 2.4443154335021973 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9974226819676932, "compression/movement_sparsity/importance_threshold": -1.6727167427965824e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248450359022379, "compression/movement_sparsity/model_sparsity": 0.8930737637628784, "compression_loss": 105.11972045898438, "distillation_loss": 3.2960877418518066, "epoch": 4.59, "learning_rate": 3.855666875391359e-05, "loss": 109.3503, "step": 5430, "task_loss": 2.5977606773376465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9974385260914334, "compression/movement_sparsity/importance_threshold": -1.6624336769417714e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248580213207933, "compression/movement_sparsity/model_sparsity": 0.8930863030923587, "compression_loss": 105.12057495117188, "distillation_loss": 3.664245843887329, "epoch": 4.59, "learning_rate": 3.855353788353163e-05, "loss": 109.2039, "step": 5431, "task_loss": 1.3039288520812988 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9974543051469952, "compression/movement_sparsity/importance_threshold": -1.652192841277312e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9247967787958156, "compression/movement_sparsity/model_sparsity": 0.893027164436518, "compression_loss": 105.12139892578125, "distillation_loss": 5.333217620849609, "epoch": 4.59, "learning_rate": 3.855040701314966e-05, "loss": 109.5119, "step": 5432, "task_loss": 3.232448101043701 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9974700192682638, "compression/movement_sparsity/importance_threshold": -1.641994148909951e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248172764399816, "compression/movement_sparsity/model_sparsity": 0.8930469579235487, "compression_loss": 105.1222915649414, "distillation_loss": 4.606888771057129, "epoch": 4.59, "learning_rate": 3.854727614276769e-05, "loss": 109.3881, "step": 5433, "task_loss": 2.971060276031494 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9974856685891245, "compression/movement_sparsity/importance_threshold": -1.6318375129461757e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9249022242102195, "compression/movement_sparsity/model_sparsity": 0.8931289874765491, "compression_loss": 105.12315368652344, "distillation_loss": 5.2242608070373535, "epoch": 4.59, "learning_rate": 3.8544145272385723e-05, "loss": 109.9443, "step": 5434, "task_loss": 2.38887357711792 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975012532434622, "compression/movement_sparsity/importance_threshold": -1.6217228464927326e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9249206589733845, "compression/movement_sparsity/model_sparsity": 0.8931467889488874, "compression_loss": 105.12401580810547, "distillation_loss": 5.861793518066406, "epoch": 4.59, "learning_rate": 3.854101440200376e-05, "loss": 109.8229, "step": 5435, "task_loss": 3.135319709777832 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975167733651621, "compression/movement_sparsity/importance_threshold": -1.611650062656022e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9249058372330131, "compression/movement_sparsity/model_sparsity": 0.8931324763808948, "compression_loss": 105.1248550415039, "distillation_loss": 3.81300950050354, "epoch": 4.59, "learning_rate": 3.8537883531621794e-05, "loss": 109.1883, "step": 5436, "task_loss": 1.936614990234375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975322290881093, "compression/movement_sparsity/importance_threshold": -1.6016190745428774e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9248692538867063, "compression/movement_sparsity/model_sparsity": 0.8930971497850771, "compression_loss": 105.12567138671875, "distillation_loss": 4.949329376220703, "epoch": 4.6, "learning_rate": 3.8534752661239826e-05, "loss": 109.1462, "step": 5437, "task_loss": 3.5151026248931885 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975476205461891, "compression/movement_sparsity/importance_threshold": -1.5916297952598725e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.924925392867936, "compression/movement_sparsity/model_sparsity": 0.8931513602195978, "compression_loss": 105.1264419555664, "distillation_loss": 4.4441633224487305, "epoch": 4.6, "learning_rate": 3.853162179085786e-05, "loss": 109.1856, "step": 5438, "task_loss": 3.7187256813049316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975629478732865, "compression/movement_sparsity/importance_threshold": -1.5816821379134073e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9249619285175722, "compression/movement_sparsity/model_sparsity": 0.8931866407572723, "compression_loss": 105.12722778320312, "distillation_loss": 2.8159220218658447, "epoch": 4.6, "learning_rate": 3.8528490920475896e-05, "loss": 108.6606, "step": 5439, "task_loss": 1.3298004865646362 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975782112032866, "compression/movement_sparsity/importance_threshold": -1.5717760156103155e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250384101287886, "compression/movement_sparsity/model_sparsity": 0.8932604949898585, "compression_loss": 105.12808990478516, "distillation_loss": 5.054182052612305, "epoch": 4.6, "learning_rate": 3.852536005009393e-05, "loss": 109.5044, "step": 5440, "task_loss": 2.5140435695648193 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9975934106700748, "compression/movement_sparsity/importance_threshold": -1.5619113414569973e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250580015362143, "compression/movement_sparsity/model_sparsity": 0.893279413372169, "compression_loss": 105.12886047363281, "distillation_loss": 3.260744571685791, "epoch": 4.6, "learning_rate": 3.852222917971196e-05, "loss": 108.8533, "step": 5441, "task_loss": 2.808159351348877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976085464075359, "compression/movement_sparsity/importance_threshold": -1.5520880285601994e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250291331263679, "compression/movement_sparsity/model_sparsity": 0.8932515366810103, "compression_loss": 105.1296157836914, "distillation_loss": 4.102766990661621, "epoch": 4.6, "learning_rate": 3.851909830932999e-05, "loss": 109.3082, "step": 5442, "task_loss": 2.122342586517334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976236185495553, "compression/movement_sparsity/importance_threshold": -1.5423059900264956e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250651918092988, "compression/movement_sparsity/model_sparsity": 0.8932863566372531, "compression_loss": 105.1303482055664, "distillation_loss": 4.034440994262695, "epoch": 4.6, "learning_rate": 3.851596743894803e-05, "loss": 109.1749, "step": 5443, "task_loss": 1.7374626398086548 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976386272300181, "compression/movement_sparsity/importance_threshold": -1.5325651389624592e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250838769799842, "compression/movement_sparsity/model_sparsity": 0.8933043999148431, "compression_loss": 105.13109588623047, "distillation_loss": 3.6509857177734375, "epoch": 4.6, "learning_rate": 3.851283656856606e-05, "loss": 109.3434, "step": 5444, "task_loss": 1.9409536123275757 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976535725828093, "compression/movement_sparsity/importance_threshold": -1.5228653884747506e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.925023695705926, "compression/movement_sparsity/model_sparsity": 0.8932462860526881, "compression_loss": 105.13178253173828, "distillation_loss": 4.862400054931641, "epoch": 4.6, "learning_rate": 3.850970569818409e-05, "loss": 109.5053, "step": 5445, "task_loss": 2.5034730434417725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976684547418141, "compression/movement_sparsity/importance_threshold": -1.5132066516699433e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250222767299773, "compression/movement_sparsity/model_sparsity": 0.8932449158229285, "compression_loss": 105.13248443603516, "distillation_loss": 3.2507286071777344, "epoch": 4.6, "learning_rate": 3.850657482780213e-05, "loss": 108.858, "step": 5446, "task_loss": 2.5123324394226074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976832738409177, "compression/movement_sparsity/importance_threshold": -1.5035888416546107e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.925028083799616, "compression/movement_sparsity/model_sparsity": 0.8932505234018604, "compression_loss": 105.13312530517578, "distillation_loss": 3.463916778564453, "epoch": 4.6, "learning_rate": 3.8503443957420164e-05, "loss": 108.9719, "step": 5447, "task_loss": 2.070040464401245 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9976980300140053, "compression/movement_sparsity/importance_threshold": -1.4940118715354132e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250275710604077, "compression/movement_sparsity/model_sparsity": 0.8932500282768213, "compression_loss": 105.13383483886719, "distillation_loss": 3.502894401550293, "epoch": 4.6, "learning_rate": 3.8500313087038195e-05, "loss": 108.4811, "step": 5448, "task_loss": 1.5050935745239258 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977127233949618, "compression/movement_sparsity/importance_threshold": -1.484475654419011e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250886227987033, "compression/movement_sparsity/model_sparsity": 0.8933089827000893, "compression_loss": 105.13449096679688, "distillation_loss": 3.0233869552612305, "epoch": 4.61, "learning_rate": 3.849718221665623e-05, "loss": 108.8509, "step": 5449, "task_loss": 1.765134572982788 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977273541176725, "compression/movement_sparsity/importance_threshold": -1.4749801034118042e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9251071887277122, "compression/movement_sparsity/model_sparsity": 0.8933269108323214, "compression_loss": 105.13510131835938, "distillation_loss": 3.886171817779541, "epoch": 4.61, "learning_rate": 3.8494051346274266e-05, "loss": 109.1054, "step": 5450, "task_loss": 2.220212936401367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977419223160227, "compression/movement_sparsity/importance_threshold": -1.4655251316204529e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9251590588569283, "compression/movement_sparsity/model_sparsity": 0.8933769990630277, "compression_loss": 105.13567352294922, "distillation_loss": 5.6045331954956055, "epoch": 4.61, "learning_rate": 3.84909204758923e-05, "loss": 109.3995, "step": 5451, "task_loss": 2.8757314682006836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977564281238972, "compression/movement_sparsity/importance_threshold": -1.456110652151791e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9251112548688761, "compression/movement_sparsity/model_sparsity": 0.8933308372890274, "compression_loss": 105.13631439208984, "distillation_loss": 3.9218149185180664, "epoch": 4.61, "learning_rate": 3.848778960551033e-05, "loss": 109.5039, "step": 5452, "task_loss": 2.547415256500244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977708716751814, "compression/movement_sparsity/importance_threshold": -1.4467365781121316e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9251512723754621, "compression/movement_sparsity/model_sparsity": 0.8933694800711538, "compression_loss": 105.13690948486328, "distillation_loss": 3.49699330329895, "epoch": 4.61, "learning_rate": 3.848465873512837e-05, "loss": 109.3515, "step": 5453, "task_loss": 2.994401693344116 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977852531037603, "compression/movement_sparsity/importance_threshold": -1.437402822608222e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250805978338843, "compression/movement_sparsity/model_sparsity": 0.8933012334174996, "compression_loss": 105.13752746582031, "distillation_loss": 3.404147148132324, "epoch": 4.61, "learning_rate": 3.84815278647464e-05, "loss": 109.129, "step": 5454, "task_loss": 2.8487491607666016 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9977995725435191, "compression/movement_sparsity/importance_threshold": -1.4281092987467221e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9250445391509535, "compression/movement_sparsity/model_sparsity": 0.8932664134612569, "compression_loss": 105.13811492919922, "distillation_loss": 4.893152713775635, "epoch": 4.61, "learning_rate": 3.847839699436444e-05, "loss": 109.6555, "step": 5455, "task_loss": 2.8395915031433105 }, { "compression/movement_sparsity/importance_regularization_factor": 0.997813830128343, "compression/movement_sparsity/importance_threshold": -1.4188559196340322e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.925090137167993, "compression/movement_sparsity/model_sparsity": 0.8933104450461352, "compression_loss": 105.138671875, "distillation_loss": 3.330874443054199, "epoch": 4.61, "learning_rate": 3.847526612398247e-05, "loss": 109.3042, "step": 5456, "task_loss": 1.154024600982666 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9978280259921171, "compression/movement_sparsity/importance_threshold": -1.409642598376986e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9251523217022141, "compression/movement_sparsity/model_sparsity": 0.8933704933503037, "compression_loss": 105.13929748535156, "distillation_loss": 2.9671449661254883, "epoch": 4.61, "learning_rate": 3.847213525360051e-05, "loss": 108.6809, "step": 5457, "task_loss": 2.0611112117767334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9978421602687264, "compression/movement_sparsity/importance_threshold": -1.4004692480820702e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9252088064843051, "compression/movement_sparsity/model_sparsity": 0.8934250377063625, "compression_loss": 105.13992309570312, "distillation_loss": 3.914046049118042, "epoch": 4.61, "learning_rate": 3.846900438321854e-05, "loss": 108.7817, "step": 5458, "task_loss": 1.3501880168914795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9978562330920562, "compression/movement_sparsity/importance_threshold": -1.391335781855945e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9252024985996258, "compression/movement_sparsity/model_sparsity": 0.8934189465169272, "compression_loss": 105.14044952392578, "distillation_loss": 4.135209560394287, "epoch": 4.61, "learning_rate": 3.846587351283657e-05, "loss": 109.7103, "step": 5459, "task_loss": 3.3069474697113037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9978702445959916, "compression/movement_sparsity/importance_threshold": -1.3822421128050974e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9252675807065823, "compression/movement_sparsity/model_sparsity": 0.8934817928532938, "compression_loss": 105.14098358154297, "distillation_loss": 4.304827690124512, "epoch": 4.61, "learning_rate": 3.8462742642454604e-05, "loss": 108.8778, "step": 5460, "task_loss": 1.419272780418396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9978841949144177, "compression/movement_sparsity/importance_threshold": -1.3731881540362742e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9252979754098861, "compression/movement_sparsity/model_sparsity": 0.8935111434050341, "compression_loss": 105.14156341552734, "distillation_loss": 2.5716919898986816, "epoch": 4.62, "learning_rate": 3.845961177207264e-05, "loss": 108.4016, "step": 5461, "task_loss": 1.220475673675537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9978980841812197, "compression/movement_sparsity/importance_threshold": -1.3641738186559622e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.925269619739248, "compression/movement_sparsity/model_sparsity": 0.8934837618389146, "compression_loss": 105.14207458496094, "distillation_loss": 4.18829345703125, "epoch": 4.62, "learning_rate": 3.8456480901690674e-05, "loss": 109.493, "step": 5462, "task_loss": 2.4378135204315186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9979119125302828, "compression/movement_sparsity/importance_threshold": -1.3551990197708218e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9253753394095076, "compression/movement_sparsity/model_sparsity": 0.893585849713269, "compression_loss": 105.14264678955078, "distillation_loss": 3.117889642715454, "epoch": 4.62, "learning_rate": 3.8453350031308706e-05, "loss": 108.3922, "step": 5463, "task_loss": 1.759373664855957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.997925680095492, "compression/movement_sparsity/importance_threshold": -1.3462636704874263e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9254299759456152, "compression/movement_sparsity/model_sparsity": 0.8936386093162797, "compression_loss": 105.1432113647461, "distillation_loss": 3.751377820968628, "epoch": 4.62, "learning_rate": 3.845021916092674e-05, "loss": 109.1616, "step": 5464, "task_loss": 2.6221935749053955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9979393870107324, "compression/movement_sparsity/importance_threshold": -1.3373676839125229e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9255267524901478, "compression/movement_sparsity/model_sparsity": 0.8937320612887881, "compression_loss": 105.14366149902344, "distillation_loss": 4.230052947998047, "epoch": 4.62, "learning_rate": 3.8447088290544776e-05, "loss": 109.4968, "step": 5465, "task_loss": 1.3410485982894897 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9979530334098894, "compression/movement_sparsity/importance_threshold": -1.3285109731524247e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9256067159583139, "compression/movement_sparsity/model_sparsity": 0.8938092777658263, "compression_loss": 105.1441650390625, "distillation_loss": 4.719376564025879, "epoch": 4.62, "learning_rate": 3.844395742016281e-05, "loss": 109.5722, "step": 5466, "task_loss": 2.313566207885742 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9979666194268478, "compression/movement_sparsity/importance_threshold": -1.3196934513140524e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9256769254573538, "compression/movement_sparsity/model_sparsity": 0.8938770753525845, "compression_loss": 105.14468383789062, "distillation_loss": 4.137048721313477, "epoch": 4.62, "learning_rate": 3.844082654978084e-05, "loss": 108.9777, "step": 5467, "task_loss": 2.5386693477630615 }, { "compression/movement_sparsity/importance_regularization_factor": 0.997980145195493, "compression/movement_sparsity/importance_threshold": -1.3109150315038925e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9256694609284137, "compression/movement_sparsity/model_sparsity": 0.8938698672531772, "compression_loss": 105.14519500732422, "distillation_loss": 4.3023271560668945, "epoch": 4.62, "learning_rate": 3.843769567939888e-05, "loss": 109.4844, "step": 5468, "task_loss": 1.8443610668182373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9979936108497102, "compression/movement_sparsity/importance_threshold": -1.302175626828432e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257169787364427, "compression/movement_sparsity/model_sparsity": 0.8939157526783185, "compression_loss": 105.14576721191406, "distillation_loss": 3.6623787879943848, "epoch": 4.62, "learning_rate": 3.843456480901691e-05, "loss": 109.1554, "step": 5469, "task_loss": 1.3797202110290527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9980070165233843, "compression/movement_sparsity/importance_threshold": -1.2934751503943312e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257294752641251, "compression/movement_sparsity/model_sparsity": 0.8939278199118311, "compression_loss": 105.14632415771484, "distillation_loss": 3.234874725341797, "epoch": 4.62, "learning_rate": 3.843143393863494e-05, "loss": 108.8508, "step": 5470, "task_loss": 2.4963691234588623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9980203623504005, "compression/movement_sparsity/importance_threshold": -1.2848135153083369e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257873551738296, "compression/movement_sparsity/model_sparsity": 0.8939837114685778, "compression_loss": 105.1468505859375, "distillation_loss": 4.341855049133301, "epoch": 4.62, "learning_rate": 3.8428303068252974e-05, "loss": 108.6728, "step": 5471, "task_loss": 1.9215161800384521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998033648464644, "compression/movement_sparsity/importance_threshold": -1.2761906346768492e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258237358092866, "compression/movement_sparsity/model_sparsity": 0.894018842317287, "compression_loss": 105.14742279052734, "distillation_loss": 3.787132501602173, "epoch": 4.63, "learning_rate": 3.842517219787101e-05, "loss": 108.6383, "step": 5472, "task_loss": 2.4077484607696533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998046875, "compression/movement_sparsity/importance_threshold": -1.2676064216066152e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258064219178793, "compression/movement_sparsity/model_sparsity": 0.8940021232113133, "compression_loss": 105.14793395996094, "distillation_loss": 3.27209734916687, "epoch": 4.63, "learning_rate": 3.8422041327489044e-05, "loss": 108.9684, "step": 5473, "task_loss": 2.403099536895752 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9980600420903535, "compression/movement_sparsity/importance_threshold": -1.2590607892042083e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258289347463758, "compression/movement_sparsity/model_sparsity": 0.8940238626548934, "compression_loss": 105.14842224121094, "distillation_loss": 4.8819355964660645, "epoch": 4.63, "learning_rate": 3.8418910457107076e-05, "loss": 110.1898, "step": 5474, "task_loss": 3.3823001384735107 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9980731498695897, "compression/movement_sparsity/importance_threshold": -1.250553650576202e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257766949679631, "compression/movement_sparsity/model_sparsity": 0.8939734174735775, "compression_loss": 105.14900970458984, "distillation_loss": 3.5622847080230713, "epoch": 4.63, "learning_rate": 3.841577958672511e-05, "loss": 108.8104, "step": 5475, "task_loss": 1.701818585395813 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9980861984715937, "compression/movement_sparsity/importance_threshold": -1.2420849188292567e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258525922949654, "compression/movement_sparsity/model_sparsity": 0.8940467074939098, "compression_loss": 105.14949798583984, "distillation_loss": 4.4940948486328125, "epoch": 4.63, "learning_rate": 3.8412648716343146e-05, "loss": 109.6733, "step": 5476, "task_loss": 2.364332914352417 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9980991880302509, "compression/movement_sparsity/importance_threshold": -1.2336545070697723e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257799979623982, "compression/movement_sparsity/model_sparsity": 0.8939766069999926, "compression_loss": 105.15001678466797, "distillation_loss": 3.6913392543792725, "epoch": 4.63, "learning_rate": 3.840951784596118e-05, "loss": 108.9585, "step": 5477, "task_loss": 2.9271106719970703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998112118679446, "compression/movement_sparsity/importance_threshold": -1.2252623284046693e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257301072450098, "compression/movement_sparsity/model_sparsity": 0.8939284301822282, "compression_loss": 105.15055084228516, "distillation_loss": 4.065769672393799, "epoch": 4.63, "learning_rate": 3.840638697557921e-05, "loss": 109.4747, "step": 5478, "task_loss": 2.445261240005493 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9981249905530645, "compression/movement_sparsity/importance_threshold": -1.2169082959403478e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257020735268978, "compression/movement_sparsity/model_sparsity": 0.8939013595085753, "compression_loss": 105.15103149414062, "distillation_loss": 4.104911804199219, "epoch": 4.63, "learning_rate": 3.840325610519724e-05, "loss": 108.7139, "step": 5479, "task_loss": 2.1350109577178955 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9981378037849913, "compression/movement_sparsity/importance_threshold": -1.2085923227835547e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257296183541367, "compression/movement_sparsity/model_sparsity": 0.8939279580862607, "compression_loss": 105.15153503417969, "distillation_loss": 5.8017120361328125, "epoch": 4.63, "learning_rate": 3.840012523481528e-05, "loss": 109.2792, "step": 5480, "task_loss": 3.2855920791625977 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9981505585091117, "compression/movement_sparsity/importance_threshold": -1.2003143220406902e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257861031362278, "compression/movement_sparsity/model_sparsity": 0.8939825024423194, "compression_loss": 105.15204620361328, "distillation_loss": 6.229037761688232, "epoch": 4.63, "learning_rate": 3.839699436443331e-05, "loss": 109.7864, "step": 5481, "task_loss": 3.0736749172210693 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9981632548593108, "compression/movement_sparsity/importance_threshold": -1.1920742068185011e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257501994674762, "compression/movement_sparsity/model_sparsity": 0.893947832175042, "compression_loss": 105.15254974365234, "distillation_loss": 3.806746482849121, "epoch": 4.63, "learning_rate": 3.8393863494051344e-05, "loss": 109.4279, "step": 5482, "task_loss": 2.2679524421691895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9981758929694737, "compression/movement_sparsity/importance_threshold": -1.183871890223561e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257838733168798, "compression/movement_sparsity/model_sparsity": 0.8939803492241258, "compression_loss": 105.15311431884766, "distillation_loss": 4.799604892730713, "epoch": 4.63, "learning_rate": 3.839073262366938e-05, "loss": 109.6951, "step": 5483, "task_loss": 3.3111438751220703 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9981884729734857, "compression/movement_sparsity/importance_threshold": -1.1757072853624435e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257720445425851, "compression/movement_sparsity/model_sparsity": 0.8939689268046176, "compression_loss": 105.15357971191406, "distillation_loss": 4.675943851470947, "epoch": 4.64, "learning_rate": 3.8387601753287414e-05, "loss": 109.6592, "step": 5484, "task_loss": 2.593230962753296 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982009950052316, "compression/movement_sparsity/importance_threshold": -1.1675803053418088e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9256771520165389, "compression/movement_sparsity/model_sparsity": 0.8938772941287647, "compression_loss": 105.15412139892578, "distillation_loss": 5.4743452072143555, "epoch": 4.64, "learning_rate": 3.8384470882905446e-05, "loss": 110.0713, "step": 5485, "task_loss": 2.888535261154175 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982134591985969, "compression/movement_sparsity/importance_threshold": -1.1594908632682303e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257531089643793, "compression/movement_sparsity/model_sparsity": 0.8939506417217759, "compression_loss": 105.154541015625, "distillation_loss": 3.588864803314209, "epoch": 4.64, "learning_rate": 3.8381340012523484e-05, "loss": 108.6363, "step": 5486, "task_loss": 2.99826717376709 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982258656874666, "compression/movement_sparsity/importance_threshold": -1.1514388722482817e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257225473227286, "compression/movement_sparsity/model_sparsity": 0.8939211299665345, "compression_loss": 105.15509033203125, "distillation_loss": 4.768405437469482, "epoch": 4.64, "learning_rate": 3.8378209142141516e-05, "loss": 109.4923, "step": 5487, "task_loss": 2.3449032306671143 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982382146057258, "compression/movement_sparsity/importance_threshold": -1.1434242453886231e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.925728890979911, "compression/movement_sparsity/model_sparsity": 0.8939272556995772, "compression_loss": 105.15558624267578, "distillation_loss": 2.6913158893585205, "epoch": 4.64, "learning_rate": 3.837507827175955e-05, "loss": 108.5963, "step": 5488, "task_loss": 1.8931641578674316 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982505060872596, "compression/movement_sparsity/importance_threshold": -1.1354468957959149e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257354611962783, "compression/movement_sparsity/model_sparsity": 0.8939336002088, "compression_loss": 105.15608215332031, "distillation_loss": 4.095090866088867, "epoch": 4.64, "learning_rate": 3.8371947401377587e-05, "loss": 108.9845, "step": 5489, "task_loss": 2.0458030700683594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982627402659533, "compression/movement_sparsity/importance_threshold": -1.1275067365766438e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257169310397722, "compression/movement_sparsity/model_sparsity": 0.8939157066201753, "compression_loss": 105.15648651123047, "distillation_loss": 3.957024574279785, "epoch": 4.64, "learning_rate": 3.836881653099562e-05, "loss": 108.3831, "step": 5490, "task_loss": 2.5465614795684814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998274917275692, "compression/movement_sparsity/importance_threshold": -1.1196036808373833e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.925759571863238, "compression/movement_sparsity/model_sparsity": 0.8939568826001766, "compression_loss": 105.15692901611328, "distillation_loss": 3.6451454162597656, "epoch": 4.64, "learning_rate": 3.836568566061366e-05, "loss": 109.5367, "step": 5491, "task_loss": 2.189373016357422 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982870372503607, "compression/movement_sparsity/importance_threshold": -1.1117376416847936e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258299363764573, "compression/movement_sparsity/model_sparsity": 0.8940248298759002, "compression_loss": 105.15730285644531, "distillation_loss": 2.8968100547790527, "epoch": 4.64, "learning_rate": 3.836255479023169e-05, "loss": 109.1668, "step": 5492, "task_loss": 1.3097339868545532 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9982991003238446, "compression/movement_sparsity/importance_threshold": -1.1039085322256219e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258222929850026, "compression/movement_sparsity/model_sparsity": 0.8940174490584558, "compression_loss": 105.15770721435547, "distillation_loss": 5.471523284912109, "epoch": 4.64, "learning_rate": 3.835942391984972e-05, "loss": 109.5617, "step": 5493, "task_loss": 3.1415367126464844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998311106630029, "compression/movement_sparsity/importance_threshold": -1.096116265566268e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258423732833014, "compression/movement_sparsity/model_sparsity": 0.8940368395367339, "compression_loss": 105.15811920166016, "distillation_loss": 4.231326103210449, "epoch": 4.64, "learning_rate": 3.835629304946776e-05, "loss": 109.3664, "step": 5494, "task_loss": 3.8758065700531006 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9983230563027988, "compression/movement_sparsity/importance_threshold": -1.0883607548133924e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257641507436102, "compression/movement_sparsity/model_sparsity": 0.8939613041819217, "compression_loss": 105.1585922241211, "distillation_loss": 5.411951065063477, "epoch": 4.64, "learning_rate": 3.835316217908579e-05, "loss": 109.6312, "step": 5495, "task_loss": 2.411834716796875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9983349494760394, "compression/movement_sparsity/importance_threshold": -1.0806419130736551e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258036674351554, "compression/movement_sparsity/model_sparsity": 0.8939994633535447, "compression_loss": 105.1590347290039, "distillation_loss": 4.161849021911621, "epoch": 4.65, "learning_rate": 3.835003130870382e-05, "loss": 109.6138, "step": 5496, "task_loss": 2.9240989685058594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9983467862836357, "compression/movement_sparsity/importance_threshold": -1.072959653453543e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258296740447692, "compression/movement_sparsity/model_sparsity": 0.8940245765561127, "compression_loss": 105.15949249267578, "distillation_loss": 4.885004997253418, "epoch": 4.65, "learning_rate": 3.8346900438321854e-05, "loss": 109.1386, "step": 5497, "task_loss": 3.7227299213409424 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998358566859473, "compression/movement_sparsity/importance_threshold": -1.0653138890598032e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258100468648406, "compression/movement_sparsity/model_sparsity": 0.8940056236301949, "compression_loss": 105.159912109375, "distillation_loss": 3.305805206298828, "epoch": 4.65, "learning_rate": 3.834376956793989e-05, "loss": 108.7117, "step": 5498, "task_loss": 2.0963401794433594 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9983702913374364, "compression/movement_sparsity/importance_threshold": -1.0577045329989224e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257734158218632, "compression/movement_sparsity/model_sparsity": 0.8939702509762341, "compression_loss": 105.16030883789062, "distillation_loss": 5.707159996032715, "epoch": 4.65, "learning_rate": 3.8340638697557925e-05, "loss": 109.5723, "step": 5499, "task_loss": 2.8513622283935547 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9983819598514109, "compression/movement_sparsity/importance_threshold": -1.0501314983776475e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257248844595852, "compression/movement_sparsity/model_sparsity": 0.8939233868155503, "compression_loss": 105.16072082519531, "distillation_loss": 4.359749794006348, "epoch": 4.65, "learning_rate": 3.8337507827175957e-05, "loss": 109.5736, "step": 5500, "task_loss": 2.8013970851898193 }, { "epoch": 4.65, "eval_accuracy": 0.5855049504950495, "eval_loss": 108.76239776611328, "eval_runtime": 208.5599, "eval_samples_per_second": 121.068, "eval_steps_per_second": 0.949, "step": 5500 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9983935725352818, "compression/movement_sparsity/importance_threshold": -1.0425946983024653e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257318124009817, "compression/movement_sparsity/model_sparsity": 0.8939300767608469, "compression_loss": 105.16111755371094, "distillation_loss": 5.257047176361084, "epoch": 4.65, "learning_rate": 3.833437695679399e-05, "loss": 109.2185, "step": 5501, "task_loss": 3.3152670860290527 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984051295229343, "compression/movement_sparsity/importance_threshold": -1.0350940458799494e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257434861610973, "compression/movement_sparsity/model_sparsity": 0.8939413494913897, "compression_loss": 105.16153717041016, "distillation_loss": 4.588510513305664, "epoch": 4.65, "learning_rate": 3.833124608641203e-05, "loss": 109.012, "step": 5502, "task_loss": 2.472874164581299 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984166309482533, "compression/movement_sparsity/importance_threshold": -1.0276294542168468e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9257739166369039, "compression/movement_sparsity/model_sparsity": 0.8939707345867374, "compression_loss": 105.16191101074219, "distillation_loss": 3.590587615966797, "epoch": 4.65, "learning_rate": 3.832811521603006e-05, "loss": 109.202, "step": 5503, "task_loss": 3.0143370628356934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984280769451241, "compression/movement_sparsity/importance_threshold": -1.0202008364196442e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258080555288454, "compression/movement_sparsity/model_sparsity": 0.8940037007027172, "compression_loss": 105.16233825683594, "distillation_loss": 5.387524604797363, "epoch": 4.65, "learning_rate": 3.832498434564809e-05, "loss": 109.4645, "step": 5504, "task_loss": 3.0847272872924805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998439467647432, "compression/movement_sparsity/importance_threshold": -1.0128081055949152e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258317726982731, "compression/movement_sparsity/model_sparsity": 0.8940266031144125, "compression_loss": 105.16273498535156, "distillation_loss": 3.8409762382507324, "epoch": 4.65, "learning_rate": 3.832185347526613e-05, "loss": 109.4404, "step": 5505, "task_loss": 2.4944803714752197 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984508031890618, "compression/movement_sparsity/importance_threshold": -1.0054511748494067e-05, "compression/movement_sparsity/linear_layer_sparsity": 0.9258736980716809, "compression/movement_sparsity/model_sparsity": 0.8940670882222661, "compression_loss": 105.16311645507812, "distillation_loss": 3.9612977504730225, "epoch": 4.65, "learning_rate": 3.831872260488416e-05, "loss": 108.6688, "step": 5506, "task_loss": 2.0610318183898926 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984620837038989, "compression/movement_sparsity/importance_threshold": -9.981299572896056e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9258907615555678, "compression/movement_sparsity/model_sparsity": 0.8940835655229881, "compression_loss": 105.1634521484375, "distillation_loss": 3.0728611946105957, "epoch": 4.65, "learning_rate": 3.831559173450219e-05, "loss": 109.1076, "step": 5507, "task_loss": 2.5437867641448975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984733093258282, "compression/movement_sparsity/importance_threshold": -9.90844366022172e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259508593604525, "compression/movement_sparsity/model_sparsity": 0.8941415987833926, "compression_loss": 105.16384887695312, "distillation_loss": 4.05833625793457, "epoch": 4.66, "learning_rate": 3.8312460864120224e-05, "loss": 108.9534, "step": 5508, "task_loss": 1.9262977838516235 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984844801887351, "compression/movement_sparsity/importance_threshold": -9.835943141536796e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259623781063887, "compression/movement_sparsity/model_sparsity": 0.8941527218249701, "compression_loss": 105.16414642333984, "distillation_loss": 3.1709847450256348, "epoch": 4.66, "learning_rate": 3.830932999373826e-05, "loss": 108.697, "step": 5509, "task_loss": 2.172089099884033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9984955964265047, "compression/movement_sparsity/importance_threshold": -9.763797147907018e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259498458062034, "compression/movement_sparsity/model_sparsity": 0.89414062004785, "compression_loss": 105.16445922851562, "distillation_loss": 3.7472143173217773, "epoch": 4.66, "learning_rate": 3.8306199123356295e-05, "loss": 109.1167, "step": 5510, "task_loss": 2.3472213745117188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985066581730219, "compression/movement_sparsity/importance_threshold": -9.692004810399855e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259790242444085, "compression/movement_sparsity/model_sparsity": 0.8941687961169392, "compression_loss": 105.16485595703125, "distillation_loss": 3.2483909130096436, "epoch": 4.66, "learning_rate": 3.8303068252974326e-05, "loss": 108.7451, "step": 5511, "task_loss": 2.174842596054077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985176655621721, "compression/movement_sparsity/importance_threshold": -9.62056526007931e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259868822708804, "compression/movement_sparsity/model_sparsity": 0.8941763841960279, "compression_loss": 105.16523742675781, "distillation_loss": 3.3938636779785156, "epoch": 4.66, "learning_rate": 3.829993738259236e-05, "loss": 109.1717, "step": 5512, "task_loss": 0.9340510368347168 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985286187278404, "compression/movement_sparsity/importance_threshold": -9.54947762801285e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259660388258529, "compression/movement_sparsity/model_sparsity": 0.894156256787459, "compression_loss": 105.16557312011719, "distillation_loss": 6.411905288696289, "epoch": 4.66, "learning_rate": 3.82968065122104e-05, "loss": 109.8375, "step": 5513, "task_loss": 3.8978633880615234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985395178039117, "compression/movement_sparsity/importance_threshold": -9.47874104526708e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9259707607962367, "compression/movement_sparsity/model_sparsity": 0.8941608165436337, "compression_loss": 105.16596984863281, "distillation_loss": 5.095303535461426, "epoch": 4.66, "learning_rate": 3.829367564182843e-05, "loss": 109.3488, "step": 5514, "task_loss": 1.8470959663391113 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985503629242715, "compression/movement_sparsity/importance_threshold": -9.408354642906e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9260321583353939, "compression/movement_sparsity/model_sparsity": 0.8942201048884397, "compression_loss": 105.16632843017578, "distillation_loss": 4.106523513793945, "epoch": 4.66, "learning_rate": 3.829054477144646e-05, "loss": 109.3228, "step": 5515, "task_loss": 2.208195686340332 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985611542228048, "compression/movement_sparsity/importance_threshold": -9.338317551996211e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9261100112258884, "compression/movement_sparsity/model_sparsity": 0.8942952832926423, "compression_loss": 105.16665649414062, "distillation_loss": 3.1844897270202637, "epoch": 4.66, "learning_rate": 3.828741390106449e-05, "loss": 108.9703, "step": 5516, "task_loss": 1.9110333919525146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985718918333966, "compression/movement_sparsity/importance_threshold": -9.268628903605185e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9261129207227915, "compression/movement_sparsity/model_sparsity": 0.8942980928393762, "compression_loss": 105.16697692871094, "distillation_loss": 4.284382343292236, "epoch": 4.66, "learning_rate": 3.828428303068253e-05, "loss": 108.3623, "step": 5517, "task_loss": 2.7876713275909424 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985825758899322, "compression/movement_sparsity/importance_threshold": -9.199287828797788e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9261836906577103, "compression/movement_sparsity/model_sparsity": 0.8943664316093167, "compression_loss": 105.1673583984375, "distillation_loss": 4.435400009155273, "epoch": 4.66, "learning_rate": 3.828115216030056e-05, "loss": 108.7831, "step": 5518, "task_loss": 2.9452011585235596 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9985932065262967, "compression/movement_sparsity/importance_threshold": -9.130293458639756e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9261835356435311, "compression/movement_sparsity/model_sparsity": 0.8943662819203514, "compression_loss": 105.16768646240234, "distillation_loss": 3.2061800956726074, "epoch": 4.66, "learning_rate": 3.8278021289918594e-05, "loss": 108.6611, "step": 5519, "task_loss": 2.259645462036133 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986037838763752, "compression/movement_sparsity/importance_threshold": -9.061644924198559e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9262230404109087, "compression/movement_sparsity/model_sparsity": 0.8944044295774387, "compression_loss": 105.16802215576172, "distillation_loss": 3.0869526863098145, "epoch": 4.67, "learning_rate": 3.827489041953663e-05, "loss": 109.1175, "step": 5520, "task_loss": 0.9609329700469971 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986143080740528, "compression/movement_sparsity/importance_threshold": -8.993341356539064e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9261819974259061, "compression/movement_sparsity/model_sparsity": 0.8943647965452339, "compression_loss": 105.16841888427734, "distillation_loss": 4.504428386688232, "epoch": 4.67, "learning_rate": 3.8271759549154665e-05, "loss": 109.2082, "step": 5521, "task_loss": 1.4417804479599 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986247792532149, "compression/movement_sparsity/importance_threshold": -8.925381886727007e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.926188877670632, "compression/movement_sparsity/model_sparsity": 0.8943714404323874, "compression_loss": 105.1687240600586, "distillation_loss": 4.324798583984375, "epoch": 4.67, "learning_rate": 3.82686286787727e-05, "loss": 109.492, "step": 5522, "task_loss": 2.553222894668579 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986351975477463, "compression/movement_sparsity/importance_threshold": -8.857765645829858e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9261892473198287, "compression/movement_sparsity/model_sparsity": 0.894371797382997, "compression_loss": 105.16905212402344, "distillation_loss": 3.866504192352295, "epoch": 4.67, "learning_rate": 3.8265497808390735e-05, "loss": 108.2679, "step": 5523, "task_loss": 2.1640403270721436 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986455630915324, "compression/movement_sparsity/importance_threshold": -8.790491764912485e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.926233509830093, "compression/movement_sparsity/model_sparsity": 0.8944145393398664, "compression_loss": 105.1693115234375, "distillation_loss": 4.284952640533447, "epoch": 4.67, "learning_rate": 3.8262366938008773e-05, "loss": 109.4593, "step": 5524, "task_loss": 2.164170980453491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986558760184582, "compression/movement_sparsity/importance_threshold": -8.72355937504149e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9262215021932837, "compression/movement_sparsity/model_sparsity": 0.8944029442023212, "compression_loss": 105.16963195800781, "distillation_loss": 3.5136938095092773, "epoch": 4.67, "learning_rate": 3.8259236067626805e-05, "loss": 108.8273, "step": 5525, "task_loss": 2.6103854179382324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986661364624089, "compression/movement_sparsity/importance_threshold": -8.656967607282608e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9262820769648739, "compression/movement_sparsity/model_sparsity": 0.8944614380441576, "compression_loss": 105.16987609863281, "distillation_loss": 3.114741086959839, "epoch": 4.67, "learning_rate": 3.825610519724484e-05, "loss": 108.698, "step": 5526, "task_loss": 2.019852638244629 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986763445572695, "compression/movement_sparsity/importance_threshold": -8.590715592702441e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263229530115297, "compression/movement_sparsity/model_sparsity": 0.8945009098728612, "compression_loss": 105.17005157470703, "distillation_loss": 4.283642768859863, "epoch": 4.67, "learning_rate": 3.825297432686287e-05, "loss": 109.1006, "step": 5527, "task_loss": 2.925391674041748 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986865004369254, "compression/movement_sparsity/importance_threshold": -8.524802462365859e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9262707132331169, "compression/movement_sparsity/model_sparsity": 0.8944504646915453, "compression_loss": 105.17029571533203, "distillation_loss": 3.7527427673339844, "epoch": 4.67, "learning_rate": 3.824984345648091e-05, "loss": 108.6132, "step": 5528, "task_loss": 2.860309600830078 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9986966042352615, "compression/movement_sparsity/importance_threshold": -8.45922734734033e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263155481034278, "compression/movement_sparsity/model_sparsity": 0.8944937593461328, "compression_loss": 105.17048645019531, "distillation_loss": 3.696965217590332, "epoch": 4.67, "learning_rate": 3.824671258609894e-05, "loss": 108.4653, "step": 5529, "task_loss": 1.7292072772979736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987066560861632, "compression/movement_sparsity/importance_threshold": -8.393989378689855e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263800340020025, "compression/movement_sparsity/model_sparsity": 0.8945560299557097, "compression_loss": 105.17066955566406, "distillation_loss": 3.969604969024658, "epoch": 4.67, "learning_rate": 3.824358171571697e-05, "loss": 109.3445, "step": 5530, "task_loss": 1.5265544652938843 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987166561235153, "compression/movement_sparsity/importance_threshold": -8.329087687482771e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264136243822327, "compression/movement_sparsity/model_sparsity": 0.8945884664030429, "compression_loss": 105.17082977294922, "distillation_loss": 4.752014636993408, "epoch": 4.67, "learning_rate": 3.824045084533501e-05, "loss": 109.5428, "step": 5531, "task_loss": 2.9104249477386475 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987266044812032, "compression/movement_sparsity/importance_threshold": -8.264521404783946e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263694334169742, "compression/movement_sparsity/model_sparsity": 0.8945457935333884, "compression_loss": 105.17105102539062, "distillation_loss": 3.736938238143921, "epoch": 4.68, "learning_rate": 3.823731997495304e-05, "loss": 109.08, "step": 5532, "task_loss": 2.687177896499634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.998736501293112, "compression/movement_sparsity/importance_threshold": -8.200289661659983e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263633282431446, "compression/movement_sparsity/model_sparsity": 0.8945398980910615, "compression_loss": 105.17122650146484, "distillation_loss": 5.019606113433838, "epoch": 4.68, "learning_rate": 3.823418910457107e-05, "loss": 108.9038, "step": 5533, "task_loss": 2.7686045169830322 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987463466931267, "compression/movement_sparsity/importance_threshold": -8.136391589176617e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263940687473099, "compression/movement_sparsity/model_sparsity": 0.8945695825643398, "compression_loss": 105.17137145996094, "distillation_loss": 4.460613250732422, "epoch": 4.68, "learning_rate": 3.8231058234189105e-05, "loss": 108.5624, "step": 5534, "task_loss": 2.7495803833007812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987561408151326, "compression/movement_sparsity/importance_threshold": -8.072826318398715e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264049674365291, "compression/movement_sparsity/model_sparsity": 0.8945801068500561, "compression_loss": 105.17155456542969, "distillation_loss": 3.437826633453369, "epoch": 4.68, "learning_rate": 3.8227927363807143e-05, "loss": 109.0161, "step": 5535, "task_loss": 2.3904030323028564 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987658837930148, "compression/movement_sparsity/importance_threshold": -8.009592980394616e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264143517564585, "compression/movement_sparsity/model_sparsity": 0.8945891687897264, "compression_loss": 105.17163848876953, "distillation_loss": 4.378613471984863, "epoch": 4.68, "learning_rate": 3.8224796493425175e-05, "loss": 109.7495, "step": 5536, "task_loss": 3.443814754486084 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987755757606583, "compression/movement_sparsity/importance_threshold": -7.946690706228318e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263804513478697, "compression/movement_sparsity/model_sparsity": 0.8945564329644625, "compression_loss": 105.17179870605469, "distillation_loss": 4.681374549865723, "epoch": 4.68, "learning_rate": 3.822166562304321e-05, "loss": 108.2461, "step": 5537, "task_loss": 2.250774383544922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987852168519485, "compression/movement_sparsity/importance_threshold": -7.884118626967293e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9263854833466121, "compression/movement_sparsity/model_sparsity": 0.8945612920985678, "compression_loss": 105.1719970703125, "distillation_loss": 3.9234302043914795, "epoch": 4.68, "learning_rate": 3.821853475266124e-05, "loss": 108.6919, "step": 5538, "task_loss": 2.0033278465270996 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9987948072007704, "compression/movement_sparsity/importance_threshold": -7.821875873676408e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264266098007882, "compression/movement_sparsity/model_sparsity": 0.8946010057325232, "compression_loss": 105.17220306396484, "distillation_loss": 3.8876285552978516, "epoch": 4.68, "learning_rate": 3.821540388227928e-05, "loss": 108.9226, "step": 5539, "task_loss": 2.638014316558838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988043469410092, "compression/movement_sparsity/importance_threshold": -7.759961577422265e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264850620705392, "compression/movement_sparsity/model_sparsity": 0.894657449986988, "compression_loss": 105.17229461669922, "distillation_loss": 2.566643714904785, "epoch": 4.68, "learning_rate": 3.821227301189731e-05, "loss": 108.692, "step": 5540, "task_loss": 2.013514280319214 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988138362065498, "compression/movement_sparsity/importance_threshold": -7.698374869272336e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.926457231063277, "compression/movement_sparsity/model_sparsity": 0.8946305750604435, "compression_loss": 105.17242431640625, "distillation_loss": 3.8205931186676025, "epoch": 4.68, "learning_rate": 3.820914214151534e-05, "loss": 109.1307, "step": 5541, "task_loss": 2.13059401512146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988232751312777, "compression/movement_sparsity/importance_threshold": -7.637114880289751e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264359344998794, "compression/movement_sparsity/model_sparsity": 0.8946100100995145, "compression_loss": 105.17259979248047, "distillation_loss": 4.249103546142578, "epoch": 4.68, "learning_rate": 3.820601127113338e-05, "loss": 109.3943, "step": 5542, "task_loss": 2.140233039855957 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988326638490778, "compression/movement_sparsity/importance_threshold": -7.57618074154285e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264558240114961, "compression/movement_sparsity/model_sparsity": 0.8946292163452197, "compression_loss": 105.17280578613281, "distillation_loss": 4.0050458908081055, "epoch": 4.69, "learning_rate": 3.820288040075141e-05, "loss": 109.1276, "step": 5543, "task_loss": 1.6777533292770386 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988420024938354, "compression/movement_sparsity/importance_threshold": -7.515571584095632e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9264735314004353, "compression/movement_sparsity/model_sparsity": 0.8946463154308747, "compression_loss": 105.17301940917969, "distillation_loss": 3.1025309562683105, "epoch": 4.69, "learning_rate": 3.819974953036944e-05, "loss": 108.7996, "step": 5544, "task_loss": 1.3356029987335205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988512911994354, "compression/movement_sparsity/importance_threshold": -7.4552865390173015e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9265193798249952, "compression/movement_sparsity/model_sparsity": 0.8946905888210047, "compression_loss": 105.17317962646484, "distillation_loss": 2.925321102142334, "epoch": 4.69, "learning_rate": 3.8196618659987475e-05, "loss": 109.0552, "step": 5545, "task_loss": 1.525267243385315 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988605300997632, "compression/movement_sparsity/importance_threshold": -7.395324737370125e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9265679469597761, "compression/movement_sparsity/model_sparsity": 0.8947374875252958, "compression_loss": 105.17334747314453, "distillation_loss": 4.556665420532227, "epoch": 4.69, "learning_rate": 3.819348778960551e-05, "loss": 109.2586, "step": 5546, "task_loss": 2.8845741748809814 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988697193287037, "compression/movement_sparsity/importance_threshold": -7.335685310223307e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266090972622875, "compression/movement_sparsity/model_sparsity": 0.8947772241883228, "compression_loss": 105.17359161376953, "distillation_loss": 3.7846620082855225, "epoch": 4.69, "learning_rate": 3.8190356919223545e-05, "loss": 109.3713, "step": 5547, "task_loss": 2.7701807022094727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988788590201423, "compression/movement_sparsity/importance_threshold": -7.2763673886408475e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266404458990021, "compression/movement_sparsity/model_sparsity": 0.8948074959029266, "compression_loss": 105.17377471923828, "distillation_loss": 3.3148512840270996, "epoch": 4.69, "learning_rate": 3.818722604884158e-05, "loss": 108.5992, "step": 5548, "task_loss": 2.0378522872924805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988879493079639, "compression/movement_sparsity/importance_threshold": -7.217370103690217e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266696362613747, "compression/movement_sparsity/model_sparsity": 0.8948356834865517, "compression_loss": 105.17394256591797, "distillation_loss": 3.1718435287475586, "epoch": 4.69, "learning_rate": 3.818409517845961e-05, "loss": 109.0555, "step": 5549, "task_loss": 1.47273588180542 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9988969903260538, "compression/movement_sparsity/importance_threshold": -7.158692586436283e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266904558580669, "compression/movement_sparsity/model_sparsity": 0.8948557878660489, "compression_loss": 105.17410278320312, "distillation_loss": 5.971027374267578, "epoch": 4.69, "learning_rate": 3.818096430807765e-05, "loss": 109.3803, "step": 5550, "task_loss": 2.9477932453155518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989059822082972, "compression/movement_sparsity/importance_threshold": -7.100333967945648e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9267275280952468, "compression/movement_sparsity/model_sparsity": 0.8948915865578342, "compression_loss": 105.17430114746094, "distillation_loss": 3.131446599960327, "epoch": 4.69, "learning_rate": 3.817783343769568e-05, "loss": 108.602, "step": 5551, "task_loss": 2.2974634170532227 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989149250885789, "compression/movement_sparsity/importance_threshold": -7.042293379284914e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9267453070291919, "compression/movement_sparsity/model_sparsity": 0.8949087547307039, "compression_loss": 105.17443084716797, "distillation_loss": 3.5285887718200684, "epoch": 4.69, "learning_rate": 3.817470256731371e-05, "loss": 108.6185, "step": 5552, "task_loss": 2.4092583656311035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989238191007845, "compression/movement_sparsity/importance_threshold": -6.9845699515180826e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266917198198363, "compression/movement_sparsity/model_sparsity": 0.8948570084068432, "compression_loss": 105.174560546875, "distillation_loss": 3.582732677459717, "epoch": 4.69, "learning_rate": 3.817157169693175e-05, "loss": 109.0138, "step": 5553, "task_loss": 2.291818857192993 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989326643787988, "compression/movement_sparsity/importance_threshold": -6.927162815714358e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.926660979315671, "compression/movement_sparsity/model_sparsity": 0.8948273239335649, "compression_loss": 105.17466735839844, "distillation_loss": 4.950085639953613, "epoch": 4.69, "learning_rate": 3.816844082654978e-05, "loss": 109.1714, "step": 5554, "task_loss": 2.706205368041992 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989414610565072, "compression/movement_sparsity/importance_threshold": -6.870071102936005e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266280328404932, "compression/movement_sparsity/model_sparsity": 0.8947955092711646, "compression_loss": 105.17479705810547, "distillation_loss": 4.543703556060791, "epoch": 4.7, "learning_rate": 3.816530995616781e-05, "loss": 109.7805, "step": 5555, "task_loss": 2.605151653289795 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989502092677945, "compression/movement_sparsity/importance_threshold": -6.81329394425223e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266784243729224, "compression/movement_sparsity/model_sparsity": 0.8948441696994323, "compression_loss": 105.17493438720703, "distillation_loss": 3.374993085861206, "epoch": 4.7, "learning_rate": 3.816217908578585e-05, "loss": 108.7021, "step": 5556, "task_loss": 1.7889169454574585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989589091465461, "compression/movement_sparsity/importance_threshold": -6.7568304707278995e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266233585667799, "compression/movement_sparsity/model_sparsity": 0.8947909955731331, "compression_loss": 105.17504119873047, "distillation_loss": 5.11016845703125, "epoch": 4.7, "learning_rate": 3.815904821540388e-05, "loss": 109.2552, "step": 5557, "task_loss": 4.1361308097839355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989675608266471, "compression/movement_sparsity/importance_threshold": -6.700679813429616e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9265928446217997, "compression/movement_sparsity/model_sparsity": 0.8947615298760349, "compression_loss": 105.17512512207031, "distillation_loss": 3.848191261291504, "epoch": 4.7, "learning_rate": 3.815591734502192e-05, "loss": 108.7094, "step": 5558, "task_loss": 1.6348544359207153 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989761644419826, "compression/movement_sparsity/importance_threshold": -6.6448411034222485e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9265534113994279, "compression/movement_sparsity/model_sparsity": 0.8947234513061623, "compression_loss": 105.17520141601562, "distillation_loss": 4.06427001953125, "epoch": 4.7, "learning_rate": 3.8152786474639954e-05, "loss": 108.9647, "step": 5559, "task_loss": 2.1815524101257324 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989847201264378, "compression/movement_sparsity/importance_threshold": -6.589313471771531e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266048641827767, "compression/movement_sparsity/model_sparsity": 0.8947731365281157, "compression_loss": 105.17527770996094, "distillation_loss": 4.414445877075195, "epoch": 4.7, "learning_rate": 3.8149655604257985e-05, "loss": 109.175, "step": 5560, "task_loss": 3.4308266639709473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9989932280138978, "compression/movement_sparsity/importance_threshold": -6.534096049544934e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.926567851566435, "compression/movement_sparsity/model_sparsity": 0.8947373954090095, "compression_loss": 105.17532348632812, "distillation_loss": 4.377659797668457, "epoch": 4.7, "learning_rate": 3.8146524733876024e-05, "loss": 109.048, "step": 5561, "task_loss": 2.2091407775878906 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990016882382476, "compression/movement_sparsity/importance_threshold": -6.479187967809059e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266255049169544, "compression/movement_sparsity/model_sparsity": 0.894793068189576, "compression_loss": 105.17537689208984, "distillation_loss": 3.518171548843384, "epoch": 4.7, "learning_rate": 3.8143393863494056e-05, "loss": 108.9943, "step": 5562, "task_loss": 2.501857280731201 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990101009333726, "compression/movement_sparsity/importance_threshold": -6.424588357627908e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266072132438009, "compression/movement_sparsity/model_sparsity": 0.8947754048916673, "compression_loss": 105.17540740966797, "distillation_loss": 4.434576988220215, "epoch": 4.7, "learning_rate": 3.814026299311209e-05, "loss": 109.7419, "step": 5563, "task_loss": 2.501779556274414 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990184662331578, "compression/movement_sparsity/importance_threshold": -6.3702963500689505e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266471234328781, "compression/movement_sparsity/model_sparsity": 0.8948139440429717, "compression_loss": 105.1754379272461, "distillation_loss": 4.800159454345703, "epoch": 4.7, "learning_rate": 3.813713212273012e-05, "loss": 109.5195, "step": 5564, "task_loss": 2.8439595699310303 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990267842714884, "compression/movement_sparsity/importance_threshold": -6.316311076197921e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266723549715957, "compression/movement_sparsity/model_sparsity": 0.8948383088007128, "compression_loss": 105.17546081542969, "distillation_loss": 4.310975074768066, "epoch": 4.7, "learning_rate": 3.813400125234816e-05, "loss": 109.5623, "step": 5565, "task_loss": 2.5220627784729004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990350551822496, "compression/movement_sparsity/importance_threshold": -6.262631667079688e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266421987516446, "compression/movement_sparsity/model_sparsity": 0.8948091885396885, "compression_loss": 105.17549133300781, "distillation_loss": 5.584038734436035, "epoch": 4.7, "learning_rate": 3.813087038196619e-05, "loss": 109.7877, "step": 5566, "task_loss": 3.0400047302246094 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990432790993262, "compression/movement_sparsity/importance_threshold": -6.209257253782588e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9267027139023966, "compression/movement_sparsity/model_sparsity": 0.8948676248088457, "compression_loss": 105.17554473876953, "distillation_loss": 4.327881813049316, "epoch": 4.71, "learning_rate": 3.812773951158422e-05, "loss": 109.2614, "step": 5567, "task_loss": 1.453777551651001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990514561566037, "compression/movement_sparsity/importance_threshold": -6.156186967370622e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9266686584796285, "compression/movement_sparsity/model_sparsity": 0.8948347392946165, "compression_loss": 105.17559814453125, "distillation_loss": 3.622878074645996, "epoch": 4.71, "learning_rate": 3.812460864120226e-05, "loss": 109.3756, "step": 5568, "task_loss": 2.611117362976074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990595864879671, "compression/movement_sparsity/importance_threshold": -6.103419938910393e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9267234023332448, "compression/movement_sparsity/model_sparsity": 0.8948876025284492, "compression_loss": 105.17566680908203, "distillation_loss": 4.083302974700928, "epoch": 4.71, "learning_rate": 3.812147777082029e-05, "loss": 109.6934, "step": 5569, "task_loss": 2.715667724609375 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990676702273016, "compression/movement_sparsity/importance_threshold": -6.05095529946937e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9267141968758299, "compression/movement_sparsity/model_sparsity": 0.8948787133068159, "compression_loss": 105.17572021484375, "distillation_loss": 5.780909538269043, "epoch": 4.71, "learning_rate": 3.8118346900438324e-05, "loss": 109.5712, "step": 5570, "task_loss": 3.6009857654571533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990757075084923, "compression/movement_sparsity/importance_threshold": -5.9987921801115535e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9268093040368935, "compression/movement_sparsity/model_sparsity": 0.8949705532443132, "compression_loss": 105.17578125, "distillation_loss": 5.239343643188477, "epoch": 4.71, "learning_rate": 3.8115216030056355e-05, "loss": 109.4272, "step": 5571, "task_loss": 2.882859468460083 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990836984654243, "compression/movement_sparsity/importance_threshold": -5.946929711903547e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9268444564830841, "compression/movement_sparsity/model_sparsity": 0.8950044980958355, "compression_loss": 105.17587280273438, "distillation_loss": 4.190180778503418, "epoch": 4.71, "learning_rate": 3.8112085159674394e-05, "loss": 109.1389, "step": 5572, "task_loss": 2.388023614883423 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990916432319829, "compression/movement_sparsity/importance_threshold": -5.895367025911952e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9268783807400082, "compression/movement_sparsity/model_sparsity": 0.8950372569501709, "compression_loss": 105.17589569091797, "distillation_loss": 5.0588788986206055, "epoch": 4.71, "learning_rate": 3.8108954289292426e-05, "loss": 109.2845, "step": 5573, "task_loss": 2.9973626136779785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9990995419420531, "compression/movement_sparsity/importance_threshold": -5.8441032532016365e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9268891959600538, "compression/movement_sparsity/model_sparsity": 0.8950477006341365, "compression_loss": 105.17591857910156, "distillation_loss": 3.5846149921417236, "epoch": 4.71, "learning_rate": 3.810582341891046e-05, "loss": 109.5452, "step": 5574, "task_loss": 1.670130968093872 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991073947295199, "compression/movement_sparsity/importance_threshold": -5.793137524840938e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9269237283495273, "compression/movement_sparsity/model_sparsity": 0.8950810467297976, "compression_loss": 105.17593383789062, "distillation_loss": 3.4605393409729004, "epoch": 4.71, "learning_rate": 3.810269254852849e-05, "loss": 108.4199, "step": 5575, "task_loss": 1.792089819908142 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991152017282688, "compression/movement_sparsity/importance_threshold": -5.74246897189299e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9269855432345516, "compression/movement_sparsity/model_sparsity": 0.8951407380833565, "compression_loss": 105.17586517333984, "distillation_loss": 3.691481590270996, "epoch": 4.71, "learning_rate": 3.809956167814653e-05, "loss": 108.8179, "step": 5576, "task_loss": 1.4901686906814575 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991229630721846, "compression/movement_sparsity/importance_threshold": -5.692096725426997e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270294837922898, "compression/movement_sparsity/model_sparsity": 0.8951831691477593, "compression_loss": 105.17579650878906, "distillation_loss": 4.215408802032471, "epoch": 4.71, "learning_rate": 3.809643080776456e-05, "loss": 109.3368, "step": 5577, "task_loss": 2.1374082565307617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991306788951527, "compression/movement_sparsity/importance_threshold": -5.642019916506091e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270639327125898, "compression/movement_sparsity/model_sparsity": 0.8952164346416698, "compression_loss": 105.17577362060547, "distillation_loss": 3.085242509841919, "epoch": 4.71, "learning_rate": 3.809329993738259e-05, "loss": 108.5494, "step": 5578, "task_loss": 1.6982921361923218 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991383493310582, "compression/movement_sparsity/importance_threshold": -5.5922376761977435e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9271144434866952, "compression/movement_sparsity/model_sparsity": 0.8952652102152955, "compression_loss": 105.17575073242188, "distillation_loss": 4.2885942459106445, "epoch": 4.72, "learning_rate": 3.809016906700063e-05, "loss": 108.8948, "step": 5579, "task_loss": 3.002333164215088 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991459745137861, "compression/movement_sparsity/importance_threshold": -5.542749135567689e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270518773791099, "compression/movement_sparsity/model_sparsity": 0.8952047934459815, "compression_loss": 105.17569732666016, "distillation_loss": 4.584970951080322, "epoch": 4.72, "learning_rate": 3.808703819661866e-05, "loss": 109.3645, "step": 5580, "task_loss": 2.9372076988220215 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991535545772214, "compression/movement_sparsity/importance_threshold": -5.493553425682529e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270122652942235, "compression/movement_sparsity/model_sparsity": 0.895166542158072, "compression_loss": 105.17567443847656, "distillation_loss": 4.3118577003479, "epoch": 4.72, "learning_rate": 3.8083907326236693e-05, "loss": 109.1682, "step": 5581, "task_loss": 2.3679845333099365 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991610896552497, "compression/movement_sparsity/importance_threshold": -5.444649677607133e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9269960603504065, "compression/movement_sparsity/model_sparsity": 0.8951508939039272, "compression_loss": 105.17564392089844, "distillation_loss": 3.8330936431884766, "epoch": 4.72, "learning_rate": 3.8080776455854725e-05, "loss": 108.8739, "step": 5582, "task_loss": 2.781050443649292 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991685798817559, "compression/movement_sparsity/importance_threshold": -5.3960370224081025e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270061839687292, "compression/movement_sparsity/model_sparsity": 0.8951606697448168, "compression_loss": 105.17558288574219, "distillation_loss": 5.359485626220703, "epoch": 4.72, "learning_rate": 3.8077645585472764e-05, "loss": 109.6429, "step": 5583, "task_loss": 3.7263500690460205 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999176025390625, "compression/movement_sparsity/importance_threshold": -5.347714591152908e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9269975747196962, "compression/movement_sparsity/model_sparsity": 0.8951523562499731, "compression_loss": 105.17554473876953, "distillation_loss": 3.7659897804260254, "epoch": 4.72, "learning_rate": 3.8074514715090796e-05, "loss": 108.9507, "step": 5584, "task_loss": 4.226146697998047 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991834263157423, "compression/movement_sparsity/importance_threshold": -5.299681514905549e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9269999476290558, "compression/movement_sparsity/model_sparsity": 0.8951546476425962, "compression_loss": 105.17549133300781, "distillation_loss": 5.4948625564575195, "epoch": 4.72, "learning_rate": 3.807138384470883e-05, "loss": 109.5163, "step": 5585, "task_loss": 4.467922210693359 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991907827909929, "compression/movement_sparsity/importance_threshold": -5.251936924733497e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9269528948635646, "compression/movement_sparsity/model_sparsity": 0.8951092112843511, "compression_loss": 105.17542266845703, "distillation_loss": 4.0091753005981445, "epoch": 4.72, "learning_rate": 3.806825297432686e-05, "loss": 109.4922, "step": 5586, "task_loss": 2.174107313156128 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9991980949502621, "compression/movement_sparsity/importance_threshold": -5.204479951700751e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.926985006647008, "compression/movement_sparsity/model_sparsity": 0.8951402199292458, "compression_loss": 105.17533111572266, "distillation_loss": 5.01429557800293, "epoch": 4.72, "learning_rate": 3.80651221039449e-05, "loss": 108.946, "step": 5587, "task_loss": 2.608403205871582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992053629274348, "compression/movement_sparsity/importance_threshold": -5.157309726875649e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270455337219277, "compression/movement_sparsity/model_sparsity": 0.8951986677129388, "compression_loss": 105.17529296875, "distillation_loss": 3.3934690952301025, "epoch": 4.72, "learning_rate": 3.806199123356293e-05, "loss": 109.0639, "step": 5588, "task_loss": 2.0721724033355713 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992125868563962, "compression/movement_sparsity/importance_threshold": -5.110425381322191e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270180962121976, "compression/movement_sparsity/model_sparsity": 0.8951721727660755, "compression_loss": 105.1751480102539, "distillation_loss": 5.66282844543457, "epoch": 4.72, "learning_rate": 3.805886036318097e-05, "loss": 108.9026, "step": 5589, "task_loss": 2.9394891262054443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992197668710315, "compression/movement_sparsity/importance_threshold": -5.0638260461087145e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270768585103071, "compression/movement_sparsity/model_sparsity": 0.895228916398471, "compression_loss": 105.17507934570312, "distillation_loss": 3.102369546890259, "epoch": 4.72, "learning_rate": 3.8055729492799e-05, "loss": 108.9947, "step": 5590, "task_loss": 1.2442781925201416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992269031052259, "compression/movement_sparsity/importance_threshold": -5.0175108523000875e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270465234278414, "compression/movement_sparsity/model_sparsity": 0.8951996234194097, "compression_loss": 105.17491149902344, "distillation_loss": 3.2460999488830566, "epoch": 4.73, "learning_rate": 3.805259862241704e-05, "loss": 108.383, "step": 5591, "task_loss": 2.0337064266204834 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992339956928643, "compression/movement_sparsity/importance_threshold": -4.971478930962045e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270773116286772, "compression/movement_sparsity/model_sparsity": 0.8952293539508313, "compression_loss": 105.17475128173828, "distillation_loss": 4.370203495025635, "epoch": 4.73, "learning_rate": 3.804946775203507e-05, "loss": 109.1093, "step": 5592, "task_loss": 2.351027727127075 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992410447678322, "compression/movement_sparsity/importance_threshold": -4.925729413160322e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270402632398326, "compression/movement_sparsity/model_sparsity": 0.8951935782881176, "compression_loss": 105.17456817626953, "distillation_loss": 5.534505844116211, "epoch": 4.73, "learning_rate": 3.80463368816531e-05, "loss": 109.1741, "step": 5593, "task_loss": 3.2604806423187256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992480504640143, "compression/movement_sparsity/importance_threshold": -4.880261429962389e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270870059769651, "compression/movement_sparsity/model_sparsity": 0.8952387152684322, "compression_loss": 105.17438507080078, "distillation_loss": 4.598031044006348, "epoch": 4.73, "learning_rate": 3.804320601127114e-05, "loss": 109.3597, "step": 5594, "task_loss": 2.107825517654419 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992550129152962, "compression/movement_sparsity/importance_threshold": -4.835074112433113e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270635630633931, "compression/movement_sparsity/model_sparsity": 0.8952160776910602, "compression_loss": 105.1742935180664, "distillation_loss": 5.4908061027526855, "epoch": 4.73, "learning_rate": 3.804007514088917e-05, "loss": 109.1034, "step": 5595, "task_loss": 3.5702922344207764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992619322555627, "compression/movement_sparsity/importance_threshold": -4.790166591639097e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270466188211826, "compression/movement_sparsity/model_sparsity": 0.8951997155356961, "compression_loss": 105.17412567138672, "distillation_loss": 4.021405220031738, "epoch": 4.73, "learning_rate": 3.8036944270507204e-05, "loss": 109.4028, "step": 5596, "task_loss": 2.2533118724823 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999268808618699, "compression/movement_sparsity/importance_threshold": -4.745537998646077e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270593061355471, "compression/movement_sparsity/model_sparsity": 0.8952119670017815, "compression_loss": 105.17396545410156, "distillation_loss": 3.674778938293457, "epoch": 4.73, "learning_rate": 3.8033813400125236e-05, "loss": 109.7735, "step": 5597, "task_loss": 2.621121644973755 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992756421385904, "compression/movement_sparsity/importance_threshold": -4.701187464519786e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270764411644398, "compression/movement_sparsity/model_sparsity": 0.8952285133897182, "compression_loss": 105.17383575439453, "distillation_loss": 3.977170467376709, "epoch": 4.73, "learning_rate": 3.8030682529743274e-05, "loss": 109.2377, "step": 5598, "task_loss": 3.638206720352173 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999282432949122, "compression/movement_sparsity/importance_threshold": -4.657114120326829e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270844422809235, "compression/movement_sparsity/model_sparsity": 0.8952362396432364, "compression_loss": 105.17371368408203, "distillation_loss": 4.477200508117676, "epoch": 4.73, "learning_rate": 3.8027551659361306e-05, "loss": 109.3179, "step": 5599, "task_loss": 2.23447585105896 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9992891811841788, "compression/movement_sparsity/importance_threshold": -4.613317097132939e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9271168044718872, "compression/movement_sparsity/model_sparsity": 0.8952674900933828, "compression_loss": 105.17363739013672, "distillation_loss": 4.93639612197876, "epoch": 4.73, "learning_rate": 3.802442078897934e-05, "loss": 109.3878, "step": 5600, "task_loss": 2.8831794261932373 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999295886977646, "compression/movement_sparsity/importance_threshold": -4.5697955260047204e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270956629226688, "compression/movement_sparsity/model_sparsity": 0.8952470748214191, "compression_loss": 105.17346954345703, "distillation_loss": 4.509274959564209, "epoch": 4.73, "learning_rate": 3.802128991859737e-05, "loss": 109.739, "step": 5601, "task_loss": 3.2465507984161377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993025504634088, "compression/movement_sparsity/importance_threshold": -4.526548538007907e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9270924314732395, "compression/movement_sparsity/model_sparsity": 0.8952439543822187, "compression_loss": 105.1733627319336, "distillation_loss": 4.101229667663574, "epoch": 4.73, "learning_rate": 3.801815904821541e-05, "loss": 108.7818, "step": 5602, "task_loss": 1.3042793273925781 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993091717753523, "compression/movement_sparsity/importance_threshold": -4.483575264207368e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9271321985723051, "compression/movement_sparsity/model_sparsity": 0.8952823553590935, "compression_loss": 105.17317199707031, "distillation_loss": 4.678112983703613, "epoch": 4.74, "learning_rate": 3.801502817783344e-05, "loss": 110.1301, "step": 5603, "task_loss": 2.174325704574585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993157510473617, "compression/movement_sparsity/importance_threshold": -4.440874835670572e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272083224584925, "compression/movement_sparsity/model_sparsity": 0.8953558641556059, "compression_loss": 105.17305755615234, "distillation_loss": 4.019852638244629, "epoch": 4.74, "learning_rate": 3.801189730745147e-05, "loss": 109.0665, "step": 5604, "task_loss": 1.8070100545883179 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999322288413322, "compression/movement_sparsity/importance_threshold": -4.3984463834632545e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272053414165835, "compression/movement_sparsity/model_sparsity": 0.8953529855216572, "compression_loss": 105.17294311523438, "distillation_loss": 3.370327949523926, "epoch": 4.74, "learning_rate": 3.800876643706951e-05, "loss": 108.7844, "step": 5605, "task_loss": 1.9210201501846313 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993287840071184, "compression/movement_sparsity/importance_threshold": -4.356289038652018e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.92711441963836, "compression/movement_sparsity/model_sparsity": 0.8952651871862238, "compression_loss": 105.17281341552734, "distillation_loss": 4.274309158325195, "epoch": 4.74, "learning_rate": 3.800563556668754e-05, "loss": 109.2517, "step": 5606, "task_loss": 2.573810338973999 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993352379626362, "compression/movement_sparsity/importance_threshold": -4.314401932301731e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9271443731474613, "compression/movement_sparsity/model_sparsity": 0.8952941117001397, "compression_loss": 105.17268371582031, "distillation_loss": 3.3102352619171143, "epoch": 4.74, "learning_rate": 3.8002504696305574e-05, "loss": 109.2903, "step": 5607, "task_loss": 2.3443877696990967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993416504137603, "compression/movement_sparsity/importance_threshold": -4.272784195478996e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9271539005574023, "compression/movement_sparsity/model_sparsity": 0.8953033118142396, "compression_loss": 105.17250061035156, "distillation_loss": 4.373867988586426, "epoch": 4.74, "learning_rate": 3.7999373825923606e-05, "loss": 109.2466, "step": 5608, "task_loss": 2.658832311630249 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993480214943761, "compression/movement_sparsity/importance_threshold": -4.231434959249547e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9271947408315552, "compression/movement_sparsity/model_sparsity": 0.8953427490993359, "compression_loss": 105.17236328125, "distillation_loss": 3.2238526344299316, "epoch": 4.74, "learning_rate": 3.7996242955541644e-05, "loss": 108.8326, "step": 5609, "task_loss": 1.1932510137557983 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993543513383685, "compression/movement_sparsity/importance_threshold": -4.19035335467912e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272495800785125, "compression/movement_sparsity/model_sparsity": 0.8953957044494549, "compression_loss": 105.17225646972656, "distillation_loss": 3.6951887607574463, "epoch": 4.74, "learning_rate": 3.7993112085159676e-05, "loss": 109.1984, "step": 5610, "task_loss": 2.638152599334717 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993606400796227, "compression/movement_sparsity/importance_threshold": -4.149538512835185e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272816918619559, "compression/movement_sparsity/model_sparsity": 0.8954267130943497, "compression_loss": 105.17206573486328, "distillation_loss": 4.539162635803223, "epoch": 4.74, "learning_rate": 3.798998121477771e-05, "loss": 108.9227, "step": 5611, "task_loss": 2.9690115451812744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993668878520238, "compression/movement_sparsity/importance_threshold": -4.108989564783477e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272943195554822, "compression/movement_sparsity/model_sparsity": 0.8954389069877561, "compression_loss": 105.17183685302734, "distillation_loss": 3.8183960914611816, "epoch": 4.74, "learning_rate": 3.798685034439574e-05, "loss": 109.1079, "step": 5612, "task_loss": 2.134345769882202 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993730947894571, "compression/movement_sparsity/importance_threshold": -4.068705641588864e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927309785200906, "compression/movement_sparsity/model_sparsity": 0.8954538413406816, "compression_loss": 105.17172241210938, "distillation_loss": 4.689839839935303, "epoch": 4.74, "learning_rate": 3.798371947401378e-05, "loss": 109.5208, "step": 5613, "task_loss": 2.920403003692627 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993792610258077, "compression/movement_sparsity/importance_threshold": -4.0286858743179485e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9273120269444215, "compression/movement_sparsity/model_sparsity": 0.895456006073411, "compression_loss": 105.17162322998047, "distillation_loss": 5.262742042541504, "epoch": 4.75, "learning_rate": 3.798058860363181e-05, "loss": 109.8888, "step": 5614, "task_loss": 2.997758626937866 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993853866949606, "compression/movement_sparsity/importance_threshold": -3.988929394037333e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272755866881264, "compression/movement_sparsity/model_sparsity": 0.8954208176520229, "compression_loss": 105.17143249511719, "distillation_loss": 4.130698204040527, "epoch": 4.75, "learning_rate": 3.797745773324984e-05, "loss": 109.336, "step": 5615, "task_loss": 3.2911829948425293 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999391471930801, "compression/movement_sparsity/importance_threshold": -3.949435331811885e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272688018367415, "compression/movement_sparsity/model_sparsity": 0.8954142658811558, "compression_loss": 105.17127990722656, "distillation_loss": 3.3833725452423096, "epoch": 4.75, "learning_rate": 3.797432686286788e-05, "loss": 108.8575, "step": 5616, "task_loss": 2.2924909591674805 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9993975168672142, "compression/movement_sparsity/importance_threshold": -3.910202818708207e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927343614064489, "compression/movement_sparsity/model_sparsity": 0.8954865080787308, "compression_loss": 105.17107391357422, "distillation_loss": 4.490034103393555, "epoch": 4.75, "learning_rate": 3.797119599248591e-05, "loss": 109.8844, "step": 5617, "task_loss": 1.9390512704849243 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994035216380851, "compression/movement_sparsity/importance_threshold": -3.871230985792902e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9273450330404376, "compression/movement_sparsity/model_sparsity": 0.8954878783084903, "compression_loss": 105.17086791992188, "distillation_loss": 5.24521541595459, "epoch": 4.75, "learning_rate": 3.7968065122103944e-05, "loss": 109.438, "step": 5618, "task_loss": 2.526980400085449 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994094863772991, "compression/movement_sparsity/importance_threshold": -3.832518964130838e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9272930317453776, "compression/movement_sparsity/model_sparsity": 0.8954376634178903, "compression_loss": 105.17066955566406, "distillation_loss": 4.407623291015625, "epoch": 4.75, "learning_rate": 3.7964934251721976e-05, "loss": 110.1086, "step": 5619, "task_loss": 3.2372076511383057 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999415411218741, "compression/movement_sparsity/importance_threshold": -3.7940658847903513e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927320528875946, "compression/movement_sparsity/model_sparsity": 0.8954642159374325, "compression_loss": 105.17052459716797, "distillation_loss": 3.0262563228607178, "epoch": 4.75, "learning_rate": 3.7961803381340014e-05, "loss": 109.148, "step": 5620, "task_loss": 1.6418359279632568 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994212962962963, "compression/movement_sparsity/importance_threshold": -3.755870878834576e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927347215163115, "compression/movement_sparsity/model_sparsity": 0.8954899854685406, "compression_loss": 105.17036437988281, "distillation_loss": 4.117809295654297, "epoch": 4.75, "learning_rate": 3.7958672510958046e-05, "loss": 108.806, "step": 5621, "task_loss": 1.6671500205993652 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99942714174385, "compression/movement_sparsity/importance_threshold": -3.7179330773309816e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9273546558437198, "compression/movement_sparsity/model_sparsity": 0.8954971705388765, "compression_loss": 105.17017364501953, "distillation_loss": 4.129779815673828, "epoch": 4.75, "learning_rate": 3.795554164057608e-05, "loss": 109.3139, "step": 5622, "task_loss": 2.314971446990967 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999432947695287, "compression/movement_sparsity/importance_threshold": -3.6802516113461706e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9274035210826915, "compression/movement_sparsity/model_sparsity": 0.8955443571065625, "compression_loss": 105.17001342773438, "distillation_loss": 4.222329139709473, "epoch": 4.75, "learning_rate": 3.7952410770194116e-05, "loss": 108.5758, "step": 5623, "task_loss": 1.9707956314086914 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994387142844928, "compression/movement_sparsity/importance_threshold": -3.642825611945011e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9274098051190356, "compression/movement_sparsity/model_sparsity": 0.8955504252669262, "compression_loss": 105.16983795166016, "distillation_loss": 2.909404754638672, "epoch": 4.75, "learning_rate": 3.794927989981215e-05, "loss": 109.1444, "step": 5624, "task_loss": 1.6796776056289673 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994444416453524, "compression/movement_sparsity/importance_threshold": -3.605654210194105e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9274863582752578, "compression/movement_sparsity/model_sparsity": 0.8956243485867271, "compression_loss": 105.16967010498047, "distillation_loss": 4.857074737548828, "epoch": 4.75, "learning_rate": 3.794614902943019e-05, "loss": 109.4302, "step": 5625, "task_loss": 2.6717443466186523 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999450129911751, "compression/movement_sparsity/importance_threshold": -3.568736537159188e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275273297152546, "compression/movement_sparsity/model_sparsity": 0.8956639125317172, "compression_loss": 105.16941833496094, "distillation_loss": 3.745906114578247, "epoch": 4.76, "learning_rate": 3.794301815904822e-05, "loss": 108.9274, "step": 5626, "task_loss": 1.7672632932662964 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994557792175736, "compression/movement_sparsity/importance_threshold": -3.5320717239068627e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275061285451981, "compression/movement_sparsity/model_sparsity": 0.8956434396870745, "compression_loss": 105.16918182373047, "distillation_loss": 3.289419651031494, "epoch": 4.76, "learning_rate": 3.793988728866626e-05, "loss": 108.6461, "step": 5627, "task_loss": 1.8858476877212524 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994613896967054, "compression/movement_sparsity/importance_threshold": -3.495658901502864e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275641634690819, "compression/movement_sparsity/model_sparsity": 0.8956994809327865, "compression_loss": 105.16899871826172, "distillation_loss": 4.8499436378479, "epoch": 4.76, "learning_rate": 3.793675641828429e-05, "loss": 109.0989, "step": 5628, "task_loss": 2.2408409118652344 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994669614830316, "compression/movement_sparsity/importance_threshold": -3.4594972010129274e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275803922612342, "compression/movement_sparsity/model_sparsity": 0.895715152216003, "compression_loss": 105.16878509521484, "distillation_loss": 3.1620476245880127, "epoch": 4.76, "learning_rate": 3.793362554790232e-05, "loss": 109.3879, "step": 5629, "task_loss": 2.9544825553894043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994724947104373, "compression/movement_sparsity/importance_threshold": -3.423585753503655e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276143999873318, "compression/movement_sparsity/model_sparsity": 0.895747991672089, "compression_loss": 105.16853332519531, "distillation_loss": 4.325157165527344, "epoch": 4.76, "learning_rate": 3.793049467752035e-05, "loss": 109.3851, "step": 5630, "task_loss": 3.3323140144348145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994779895128076, "compression/movement_sparsity/importance_threshold": -3.3879236900407825e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275926980022345, "compression/movement_sparsity/model_sparsity": 0.8957270352169429, "compression_loss": 105.16829681396484, "distillation_loss": 3.586042642593384, "epoch": 4.76, "learning_rate": 3.792736380713839e-05, "loss": 109.1204, "step": 5631, "task_loss": 1.4665666818618774 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994834460240277, "compression/movement_sparsity/importance_threshold": -3.3525101416900446e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275908855287538, "compression/movement_sparsity/model_sparsity": 0.8957252850075021, "compression_loss": 105.16802215576172, "distillation_loss": 4.750227451324463, "epoch": 4.76, "learning_rate": 3.792423293675642e-05, "loss": 109.3753, "step": 5632, "task_loss": 2.7582952976226807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994888643779828, "compression/movement_sparsity/importance_threshold": -3.317344239518044e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276396792227198, "compression/movement_sparsity/model_sparsity": 0.8957724024879734, "compression_loss": 105.16783142089844, "distillation_loss": 3.9537556171417236, "epoch": 4.76, "learning_rate": 3.7921102066374454e-05, "loss": 108.6935, "step": 5633, "task_loss": 1.7923908233642578 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999494244708558, "compression/movement_sparsity/importance_threshold": -3.282425114590516e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276359827307527, "compression/movement_sparsity/model_sparsity": 0.895768832981877, "compression_loss": 105.16761016845703, "distillation_loss": 3.088893413543701, "epoch": 4.76, "learning_rate": 3.7917971195992486e-05, "loss": 108.4697, "step": 5634, "task_loss": 2.380194902420044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9994995871496383, "compression/movement_sparsity/importance_threshold": -3.2477518979740633e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276404423694485, "compression/movement_sparsity/model_sparsity": 0.8957731394182642, "compression_loss": 105.16724395751953, "distillation_loss": 2.819814920425415, "epoch": 4.76, "learning_rate": 3.7914840325610525e-05, "loss": 109.1464, "step": 5635, "task_loss": 1.66770339012146 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995048918351089, "compression/movement_sparsity/importance_threshold": -3.2133237207335533e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276296033010675, "compression/movement_sparsity/model_sparsity": 0.895762672705227, "compression_loss": 105.16696166992188, "distillation_loss": 5.193902015686035, "epoch": 4.76, "learning_rate": 3.7911709455228557e-05, "loss": 109.4587, "step": 5636, "task_loss": 2.1770036220550537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995101588988551, "compression/movement_sparsity/importance_threshold": -3.179139713936456e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276041332789973, "compression/movement_sparsity/model_sparsity": 0.8957380776567698, "compression_loss": 105.1666488647461, "distillation_loss": 4.229099273681641, "epoch": 4.76, "learning_rate": 3.790857858484659e-05, "loss": 108.8378, "step": 5637, "task_loss": 2.3797640800476074 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995153884747618, "compression/movement_sparsity/importance_threshold": -3.1451990086476395e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276028216205573, "compression/movement_sparsity/model_sparsity": 0.8957368110578324, "compression_loss": 105.16636657714844, "distillation_loss": 2.2725024223327637, "epoch": 4.77, "learning_rate": 3.790544771446462e-05, "loss": 108.5281, "step": 5638, "task_loss": 1.6470874547958374 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995205806967143, "compression/movement_sparsity/importance_threshold": -3.111500735933706e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9276054807099401, "compression/movement_sparsity/model_sparsity": 0.8957393787993146, "compression_loss": 105.1659927368164, "distillation_loss": 3.8390512466430664, "epoch": 4.77, "learning_rate": 3.790231684408266e-05, "loss": 109.3017, "step": 5639, "task_loss": 2.5434184074401855 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995257356985978, "compression/movement_sparsity/importance_threshold": -3.078044026860391e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9275752052483127, "compression/movement_sparsity/model_sparsity": 0.8957101433929323, "compression_loss": 105.16567993164062, "distillation_loss": 3.701099395751953, "epoch": 4.77, "learning_rate": 3.789918597370069e-05, "loss": 109.1517, "step": 5640, "task_loss": 3.489248752593994 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995308536142973, "compression/movement_sparsity/importance_threshold": -3.0448280124942967e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927683715173799, "compression/movement_sparsity/model_sparsity": 0.8958149256686626, "compression_loss": 105.16533660888672, "distillation_loss": 5.723487854003906, "epoch": 4.77, "learning_rate": 3.789605510331872e-05, "loss": 109.6504, "step": 5641, "task_loss": 3.634521245956421 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999535934577698, "compression/movement_sparsity/importance_threshold": -3.0118518239011585e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277491192332817, "compression/movement_sparsity/model_sparsity": 0.8958780828974957, "compression_loss": 105.16497039794922, "distillation_loss": 3.3670592308044434, "epoch": 4.77, "learning_rate": 3.789292423293676e-05, "loss": 108.7991, "step": 5642, "task_loss": 1.3536770343780518 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999540978722685, "compression/movement_sparsity/importance_threshold": -2.9791145921467116e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277172578573587, "compression/movement_sparsity/model_sparsity": 0.8958473160578526, "compression_loss": 105.1646957397461, "distillation_loss": 3.135817527770996, "epoch": 4.77, "learning_rate": 3.788979336255479e-05, "loss": 109.0314, "step": 5643, "task_loss": 2.1893410682678223 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995459861831435, "compression/movement_sparsity/importance_threshold": -2.946615448296691e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277332004694878, "compression/movement_sparsity/model_sparsity": 0.8958627109922099, "compression_loss": 105.16436767578125, "distillation_loss": 4.09429931640625, "epoch": 4.77, "learning_rate": 3.7886662492172824e-05, "loss": 108.6308, "step": 5644, "task_loss": 3.764420747756958 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995509570929586, "compression/movement_sparsity/importance_threshold": -2.9143535234176995e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927725402063854, "compression/movement_sparsity/model_sparsity": 0.8958551804858004, "compression_loss": 105.16400909423828, "distillation_loss": 3.5779495239257812, "epoch": 4.77, "learning_rate": 3.7883531621790856e-05, "loss": 108.8473, "step": 5645, "task_loss": 2.109835624694824 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995558915860154, "compression/movement_sparsity/importance_threshold": -2.8823279485763395e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277410704201275, "compression/movement_sparsity/model_sparsity": 0.8958703105858343, "compression_loss": 105.1636734008789, "distillation_loss": 2.825469970703125, "epoch": 4.77, "learning_rate": 3.7880400751408895e-05, "loss": 108.6366, "step": 5646, "task_loss": 1.1203947067260742 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995607897961991, "compression/movement_sparsity/importance_threshold": -2.850537854838346e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927812245776746, "compression/movement_sparsity/model_sparsity": 0.895939040849992, "compression_loss": 105.1633071899414, "distillation_loss": 5.232424736022949, "epoch": 4.77, "learning_rate": 3.7877269881026927e-05, "loss": 109.9169, "step": 5647, "task_loss": 2.6240832805633545 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995656518573949, "compression/movement_sparsity/importance_threshold": -2.8189823732694547e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277784646098336, "compression/movement_sparsity/model_sparsity": 0.895906420170086, "compression_loss": 105.1629409790039, "distillation_loss": 3.316263437271118, "epoch": 4.77, "learning_rate": 3.787413901064496e-05, "loss": 108.7016, "step": 5648, "task_loss": 2.361022472381592 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995704779034879, "compression/movement_sparsity/importance_threshold": -2.787660634934533e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9278055563187023, "compression/movement_sparsity/model_sparsity": 0.8959325811954112, "compression_loss": 105.16254425048828, "distillation_loss": 5.069239139556885, "epoch": 4.77, "learning_rate": 3.787100814026299e-05, "loss": 109.431, "step": 5649, "task_loss": 2.7190113067626953 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995752680683632, "compression/movement_sparsity/importance_threshold": -2.756571770901918e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9278021698550937, "compression/movement_sparsity/model_sparsity": 0.8959293110672456, "compression_loss": 105.16211700439453, "distillation_loss": 3.864623546600342, "epoch": 4.78, "learning_rate": 3.786787726988103e-05, "loss": 108.9246, "step": 5650, "task_loss": 1.445967197418213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995800224859058, "compression/movement_sparsity/importance_threshold": -2.725714912236478e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9278754080927132, "compression/movement_sparsity/model_sparsity": 0.8960000333460957, "compression_loss": 105.16162872314453, "distillation_loss": 4.760858535766602, "epoch": 4.78, "learning_rate": 3.786474639949906e-05, "loss": 109.4685, "step": 5651, "task_loss": 1.7276077270507812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995847412900012, "compression/movement_sparsity/importance_threshold": -2.695089190003948e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9278666319053331, "compression/movement_sparsity/model_sparsity": 0.8959915586477509, "compression_loss": 105.1612548828125, "distillation_loss": 3.952158212661743, "epoch": 4.78, "learning_rate": 3.786161552911709e-05, "loss": 108.8091, "step": 5652, "task_loss": 2.209183931350708 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995894246145341, "compression/movement_sparsity/importance_threshold": -2.66469373527093e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277965774204725, "compression/movement_sparsity/model_sparsity": 0.8959239107499579, "compression_loss": 105.16077423095703, "distillation_loss": 2.907589912414551, "epoch": 4.78, "learning_rate": 3.785848465873513e-05, "loss": 108.8527, "step": 5653, "task_loss": 2.582805633544922 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9995940725933901, "compression/movement_sparsity/importance_threshold": -2.6345276791031602e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277693545457598, "compression/movement_sparsity/model_sparsity": 0.895897623064739, "compression_loss": 105.16033935546875, "distillation_loss": 4.12917423248291, "epoch": 4.78, "learning_rate": 3.785535378835316e-05, "loss": 109.2267, "step": 5654, "task_loss": 1.5618934631347656 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999598685360454, "compression/movement_sparsity/importance_threshold": -2.604590152566373e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9278130804684805, "compression/movement_sparsity/model_sparsity": 0.8959398468674976, "compression_loss": 105.15992736816406, "distillation_loss": 4.940338134765625, "epoch": 4.78, "learning_rate": 3.7852222917971194e-05, "loss": 109.6535, "step": 5655, "task_loss": 2.576239824295044 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996032630496111, "compression/movement_sparsity/importance_threshold": -2.5748802867280388e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9277978294580742, "compression/movement_sparsity/model_sparsity": 0.8959251197762164, "compression_loss": 105.15950775146484, "distillation_loss": 3.9481563568115234, "epoch": 4.78, "learning_rate": 3.784909204758923e-05, "loss": 108.5636, "step": 5656, "task_loss": 1.92830491065979 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996078057947465, "compression/movement_sparsity/importance_threshold": -2.545397212653025e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.927883075332503, "compression/movement_sparsity/model_sparsity": 0.8960074371926116, "compression_loss": 105.15904235839844, "distillation_loss": 3.5926618576049805, "epoch": 4.78, "learning_rate": 3.7845961177207265e-05, "loss": 109.4275, "step": 5657, "task_loss": 2.4622955322265625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996123137297453, "compression/movement_sparsity/importance_threshold": -2.516140061407067e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9279415156780864, "compression/movement_sparsity/model_sparsity": 0.8960638699325406, "compression_loss": 105.15859985351562, "distillation_loss": 3.332850217819214, "epoch": 4.78, "learning_rate": 3.78428303068253e-05, "loss": 109.0933, "step": 5658, "task_loss": 1.9004569053649902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996167869884928, "compression/movement_sparsity/importance_threshold": -2.4871079640567673e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280420721837597, "compression/movement_sparsity/model_sparsity": 0.8961609720128959, "compression_loss": 105.15824127197266, "distillation_loss": 4.395236015319824, "epoch": 4.78, "learning_rate": 3.7839699436443335e-05, "loss": 108.5013, "step": 5659, "task_loss": 1.7540792226791382 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996212257048739, "compression/movement_sparsity/importance_threshold": -2.458300051667861e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280043441173598, "compression/movement_sparsity/model_sparsity": 0.896124540021642, "compression_loss": 105.15776062011719, "distillation_loss": 3.2006683349609375, "epoch": 4.78, "learning_rate": 3.783656856606137e-05, "loss": 108.6742, "step": 5660, "task_loss": 1.020683765411377 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996256300127739, "compression/movement_sparsity/importance_threshold": -2.4297154553060835e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280648473439441, "compression/movement_sparsity/model_sparsity": 0.8961829647762636, "compression_loss": 105.15736389160156, "distillation_loss": 3.6531529426574707, "epoch": 4.78, "learning_rate": 3.7833437695679405e-05, "loss": 108.8637, "step": 5661, "task_loss": 1.7094944715499878 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996300000460779, "compression/movement_sparsity/importance_threshold": -2.4013533060389045e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280528754796378, "compression/movement_sparsity/model_sparsity": 0.8961714041823258, "compression_loss": 105.1568374633789, "distillation_loss": 4.255620002746582, "epoch": 4.79, "learning_rate": 3.783030682529744e-05, "loss": 109.7569, "step": 5662, "task_loss": 1.5422857999801636 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999634335938671, "compression/movement_sparsity/importance_threshold": -2.373212734932059e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280008384120748, "compression/movement_sparsity/model_sparsity": 0.8961211547481185, "compression_loss": 105.15644073486328, "distillation_loss": 4.779195785522461, "epoch": 4.79, "learning_rate": 3.782717595491547e-05, "loss": 109.0893, "step": 5663, "task_loss": 2.192225933074951 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996386378244385, "compression/movement_sparsity/importance_threshold": -2.345292873049548e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280157436216196, "compression/movement_sparsity/model_sparsity": 0.8961355479178617, "compression_loss": 105.15599060058594, "distillation_loss": 4.366491317749023, "epoch": 4.79, "learning_rate": 3.782404508453351e-05, "loss": 109.5473, "step": 5664, "task_loss": 2.5201008319854736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996429058372653, "compression/movement_sparsity/importance_threshold": -2.3175928514597086e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9279991571044381, "compression/movement_sparsity/model_sparsity": 0.8961195311985714, "compression_loss": 105.155517578125, "distillation_loss": 3.8546900749206543, "epoch": 4.79, "learning_rate": 3.782091421415154e-05, "loss": 109.5276, "step": 5665, "task_loss": 2.2374320030212402 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996471401110367, "compression/movement_sparsity/importance_threshold": -2.2901118012274083e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.928034464564808, "compression/movement_sparsity/model_sparsity": 0.896153625739059, "compression_loss": 105.15509033203125, "distillation_loss": 5.075146675109863, "epoch": 4.79, "learning_rate": 3.781778334376957e-05, "loss": 109.2981, "step": 5666, "task_loss": 2.9911551475524902 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996513407796378, "compression/movement_sparsity/importance_threshold": -2.26284885341925e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280556538106969, "compression/movement_sparsity/model_sparsity": 0.8961740870691659, "compression_loss": 105.15460968017578, "distillation_loss": 3.323672294616699, "epoch": 4.79, "learning_rate": 3.78146524733876e-05, "loss": 109.5624, "step": 5667, "task_loss": 1.9575350284576416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996555079769538, "compression/movement_sparsity/importance_threshold": -2.2358031391009683e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9280904246835231, "compression/movement_sparsity/model_sparsity": 0.8962076634555428, "compression_loss": 105.15418243408203, "distillation_loss": 6.101563453674316, "epoch": 4.79, "learning_rate": 3.781152160300564e-05, "loss": 109.5092, "step": 5668, "task_loss": 2.9339373111724854 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996596418368697, "compression/movement_sparsity/importance_threshold": -2.208973789338299e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9281336974878736, "compression/movement_sparsity/model_sparsity": 0.8962494497059412, "compression_loss": 105.15373229980469, "distillation_loss": 3.7804789543151855, "epoch": 4.79, "learning_rate": 3.780839073262367e-05, "loss": 109.1008, "step": 5669, "task_loss": 2.4357709884643555 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996637424932708, "compression/movement_sparsity/importance_threshold": -2.1823599351978445e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9281261733380953, "compression/movement_sparsity/model_sparsity": 0.8962421840338548, "compression_loss": 105.1533203125, "distillation_loss": 4.596798896789551, "epoch": 4.79, "learning_rate": 3.7805259862241705e-05, "loss": 109.1912, "step": 5670, "task_loss": 2.3184292316436768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996678100800421, "compression/movement_sparsity/importance_threshold": -2.15596070774534e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9281172182882008, "compression/movement_sparsity/model_sparsity": 0.8962335366174732, "compression_loss": 105.15281677246094, "distillation_loss": 4.150029182434082, "epoch": 4.79, "learning_rate": 3.780212899185974e-05, "loss": 108.9138, "step": 5671, "task_loss": 3.5438995361328125 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996718447310688, "compression/movement_sparsity/importance_threshold": -2.1297752380473878e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9281631621061018, "compression/movement_sparsity/model_sparsity": 0.8962779021238895, "compression_loss": 105.15233612060547, "distillation_loss": 2.3976187705993652, "epoch": 4.79, "learning_rate": 3.7798998121477775e-05, "loss": 108.5502, "step": 5672, "task_loss": 2.4671525955200195 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996758465802362, "compression/movement_sparsity/importance_threshold": -2.103802657168856e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9281820619118046, "compression/movement_sparsity/model_sparsity": 0.8962961526631239, "compression_loss": 105.15192413330078, "distillation_loss": 2.6796724796295166, "epoch": 4.79, "learning_rate": 3.779586725109581e-05, "loss": 109.0561, "step": 5673, "task_loss": 2.377904176712036 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996798157614292, "compression/movement_sparsity/importance_threshold": -2.078042096177214e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9282271233413005, "compression/movement_sparsity/model_sparsity": 0.8963396660938915, "compression_loss": 105.15143585205078, "distillation_loss": 3.775479793548584, "epoch": 4.8, "learning_rate": 3.779273638071384e-05, "loss": 108.8999, "step": 5674, "task_loss": 1.8431401252746582 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999683752408533, "compression/movement_sparsity/importance_threshold": -2.0524926861373305e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9282361260878657, "compression/movement_sparsity/model_sparsity": 0.8963483595684163, "compression_loss": 105.15095520019531, "distillation_loss": 3.4375410079956055, "epoch": 4.8, "learning_rate": 3.778960551033187e-05, "loss": 109.0596, "step": 5675, "task_loss": 2.2854788303375244 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996876566554328, "compression/movement_sparsity/importance_threshold": -2.0271535581158073e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9282125281601142, "compression/movement_sparsity/model_sparsity": 0.8963255723020789, "compression_loss": 105.15046691894531, "distillation_loss": 3.0544443130493164, "epoch": 4.8, "learning_rate": 3.778647463994991e-05, "loss": 108.7629, "step": 5676, "task_loss": 0.9625123739242554 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996915286360136, "compression/movement_sparsity/importance_threshold": -2.0020238431792473e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9282812113656969, "compression/movement_sparsity/model_sparsity": 0.8963918960282554, "compression_loss": 105.15006256103516, "distillation_loss": 4.308782577514648, "epoch": 4.8, "learning_rate": 3.778334376956794e-05, "loss": 109.04, "step": 5677, "task_loss": 2.905580520629883 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996953684841609, "compression/movement_sparsity/importance_threshold": -1.9771026723916507e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283116656898388, "compression/movement_sparsity/model_sparsity": 0.8964213041526747, "compression_loss": 105.14957427978516, "distillation_loss": 5.329658508300781, "epoch": 4.8, "learning_rate": 3.778021289918597e-05, "loss": 109.6446, "step": 5678, "task_loss": 3.6057379245758057 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9996991763337594, "compression/movement_sparsity/importance_threshold": -1.952389176821355e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283273578944476, "compression/movement_sparsity/model_sparsity": 0.8964364572817803, "compression_loss": 105.14912414550781, "distillation_loss": 3.8374180793762207, "epoch": 4.8, "learning_rate": 3.777708202880401e-05, "loss": 108.6666, "step": 5679, "task_loss": 2.265000343322754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997029523186944, "compression/movement_sparsity/importance_threshold": -1.927882487533228e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283355617217811, "compression/movement_sparsity/model_sparsity": 0.896444379282407, "compression_loss": 105.14866638183594, "distillation_loss": 5.119097709655762, "epoch": 4.8, "learning_rate": 3.777395115842204e-05, "loss": 109.1235, "step": 5680, "task_loss": 3.119065761566162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997066965728512, "compression/movement_sparsity/importance_threshold": -1.903581735593872e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283329741774041, "compression/movement_sparsity/model_sparsity": 0.8964418806281396, "compression_loss": 105.14824676513672, "distillation_loss": 4.361950874328613, "epoch": 4.8, "learning_rate": 3.7770820288040075e-05, "loss": 109.3964, "step": 5681, "task_loss": 2.719069004058838 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997104092301147, "compression/movement_sparsity/importance_threshold": -1.879486052068155e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283283952970319, "compression/movement_sparsity/model_sparsity": 0.8964374590463945, "compression_loss": 105.14778900146484, "distillation_loss": 5.382049560546875, "epoch": 4.8, "learning_rate": 3.776768941765811e-05, "loss": 109.2202, "step": 5682, "task_loss": 2.760979652404785 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997140904243702, "compression/movement_sparsity/importance_threshold": -1.855594568023547e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283245199425503, "compression/movement_sparsity/model_sparsity": 0.8964337168222613, "compression_loss": 105.14735412597656, "distillation_loss": 3.0451557636260986, "epoch": 4.8, "learning_rate": 3.7764558547276145e-05, "loss": 108.5832, "step": 5683, "task_loss": 1.4524154663085938 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997177402895029, "compression/movement_sparsity/importance_threshold": -1.8319064145249156e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283404029338412, "compression/movement_sparsity/model_sparsity": 0.8964490541839396, "compression_loss": 105.14691162109375, "distillation_loss": 5.0012006759643555, "epoch": 4.8, "learning_rate": 3.776142767689418e-05, "loss": 108.7265, "step": 5684, "task_loss": 2.8204026222229004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997213589593976, "compression/movement_sparsity/importance_threshold": -1.8084207226405982e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9283277275436443, "compression/movement_sparsity/model_sparsity": 0.89643681423239, "compression_loss": 105.14641571044922, "distillation_loss": 3.9979729652404785, "epoch": 4.81, "learning_rate": 3.775829680651221e-05, "loss": 109.3176, "step": 5685, "task_loss": 2.732560157775879 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997249465679399, "compression/movement_sparsity/importance_threshold": -1.785136623433728e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.92839755546932, "compression/movement_sparsity/model_sparsity": 0.8965042433540028, "compression_loss": 105.14590454101562, "distillation_loss": 6.056184768676758, "epoch": 4.81, "learning_rate": 3.775516593613024e-05, "loss": 109.0438, "step": 5686, "task_loss": 2.952420949935913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997285032490146, "compression/movement_sparsity/importance_threshold": -1.7620532479709072e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9284437735430766, "compression/movement_sparsity/model_sparsity": 0.8965488736947425, "compression_loss": 105.14545440673828, "distillation_loss": 4.826766014099121, "epoch": 4.81, "learning_rate": 3.775203506574828e-05, "loss": 109.1896, "step": 5687, "task_loss": 2.492788553237915 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999732029136507, "compression/movement_sparsity/importance_threshold": -1.739169727319606e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.928488787275902, "compression/movement_sparsity/model_sparsity": 0.8965923410673668, "compression_loss": 105.14495849609375, "distillation_loss": 5.068090438842773, "epoch": 4.81, "learning_rate": 3.774890419536631e-05, "loss": 109.6649, "step": 5688, "task_loss": 2.0418059825897217 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997355243643022, "compression/movement_sparsity/importance_threshold": -1.7164851925455596e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285376286665384, "compression/movement_sparsity/model_sparsity": 0.8966395046059813, "compression_loss": 105.14447784423828, "distillation_loss": 4.4825263023376465, "epoch": 4.81, "learning_rate": 3.774577332498434e-05, "loss": 109.2077, "step": 5689, "task_loss": 4.041322231292725 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997389890662853, "compression/movement_sparsity/importance_threshold": -1.6939987747136356e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9284897173609775, "compression/movement_sparsity/model_sparsity": 0.8965932392011589, "compression_loss": 105.14395904541016, "distillation_loss": 1.916048526763916, "epoch": 4.81, "learning_rate": 3.774264245460238e-05, "loss": 108.9146, "step": 5690, "task_loss": 1.2966583967208862 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997424233763416, "compression/movement_sparsity/importance_threshold": -1.6717096048904367e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9284685281150886, "compression/movement_sparsity/model_sparsity": 0.8965727778710519, "compression_loss": 105.14337921142578, "distillation_loss": 2.6985087394714355, "epoch": 4.81, "learning_rate": 3.773951158422041e-05, "loss": 108.3378, "step": 5691, "task_loss": 1.8842496871948242 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997458274283559, "compression/movement_sparsity/importance_threshold": -1.6496168141425654e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285040979071464, "compression/movement_sparsity/model_sparsity": 0.8966071257313271, "compression_loss": 105.14291381835938, "distillation_loss": 3.9621620178222656, "epoch": 4.81, "learning_rate": 3.773638071383845e-05, "loss": 109.7357, "step": 5692, "task_loss": 3.7692806720733643 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997492013562138, "compression/movement_sparsity/importance_threshold": -1.6277195335348896e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9284920187253313, "compression/movement_sparsity/model_sparsity": 0.8965954615065672, "compression_loss": 105.14236450195312, "distillation_loss": 4.358094215393066, "epoch": 4.81, "learning_rate": 3.773324984345648e-05, "loss": 109.2179, "step": 5693, "task_loss": 2.5227670669555664 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997525452938001, "compression/movement_sparsity/importance_threshold": -1.606016894134879e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285256091055615, "compression/movement_sparsity/model_sparsity": 0.8966278979539004, "compression_loss": 105.14187622070312, "distillation_loss": 3.936018943786621, "epoch": 4.81, "learning_rate": 3.773011897307452e-05, "loss": 109.1957, "step": 5694, "task_loss": 2.3310346603393555 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999755859375, "compression/movement_sparsity/importance_threshold": -1.584508027008269e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285574943298198, "compression/movement_sparsity/model_sparsity": 0.8966586878226149, "compression_loss": 105.14134216308594, "distillation_loss": 3.8969006538391113, "epoch": 4.81, "learning_rate": 3.7726988102692554e-05, "loss": 108.8044, "step": 5695, "task_loss": 2.731745719909668 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997591437336987, "compression/movement_sparsity/importance_threshold": -1.5631920632207946e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285024881445155, "compression/movement_sparsity/model_sparsity": 0.8966055712689948, "compression_loss": 105.14086151123047, "distillation_loss": 3.358635902404785, "epoch": 4.81, "learning_rate": 3.7723857232310585e-05, "loss": 108.7996, "step": 5696, "task_loss": 1.7261542081832886 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997623985037813, "compression/movement_sparsity/importance_threshold": -1.5420681338373238e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9284886322617227, "compression/movement_sparsity/model_sparsity": 0.8965921913784015, "compression_loss": 105.1404037475586, "distillation_loss": 4.911891937255859, "epoch": 4.82, "learning_rate": 3.772072636192862e-05, "loss": 108.8121, "step": 5697, "task_loss": 3.5406105518341064 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999765623819133, "compression/movement_sparsity/importance_threshold": -1.5211353699253263e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9284820858936906, "compression/movement_sparsity/model_sparsity": 0.8965858698982503, "compression_loss": 105.13993835449219, "distillation_loss": 4.28127908706665, "epoch": 4.82, "learning_rate": 3.7717595491546656e-05, "loss": 109.0779, "step": 5698, "task_loss": 2.6752474308013916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999768819813639, "compression/movement_sparsity/importance_threshold": -1.5003929025505375e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285261576172728, "compression/movement_sparsity/model_sparsity": 0.8966284276225469, "compression_loss": 105.13948059082031, "distillation_loss": 4.4638824462890625, "epoch": 4.82, "learning_rate": 3.771446462116469e-05, "loss": 109.32, "step": 5699, "task_loss": 2.550107955932617 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997719866211843, "compression/movement_sparsity/importance_threshold": -1.4798398627786924e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285689057582474, "compression/movement_sparsity/model_sparsity": 0.8966697072333704, "compression_loss": 105.1390380859375, "distillation_loss": 2.9985618591308594, "epoch": 4.82, "learning_rate": 3.771133375078272e-05, "loss": 108.3931, "step": 5700, "task_loss": 1.3278377056121826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999775124375654, "compression/movement_sparsity/importance_threshold": -1.459475381677261e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285766564672107, "compression/movement_sparsity/model_sparsity": 0.8966771916816368, "compression_loss": 105.13862609863281, "distillation_loss": 5.424281120300293, "epoch": 4.82, "learning_rate": 3.770820288040076e-05, "loss": 109.7177, "step": 5701, "task_loss": 2.5882558822631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997782332109333, "compression/movement_sparsity/importance_threshold": -1.4392985903102437e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285442465795763, "compression/movement_sparsity/model_sparsity": 0.8966458951733473, "compression_loss": 105.13821411132812, "distillation_loss": 4.760251998901367, "epoch": 4.82, "learning_rate": 3.770507201001879e-05, "loss": 109.1537, "step": 5702, "task_loss": 2.0732247829437256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997813132609075, "compression/movement_sparsity/importance_threshold": -1.4193086197451105e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285597718458383, "compression/movement_sparsity/model_sparsity": 0.8966608870989518, "compression_loss": 105.13782501220703, "distillation_loss": 3.0764834880828857, "epoch": 4.82, "learning_rate": 3.770194113963682e-05, "loss": 109.0119, "step": 5703, "task_loss": 2.2881557941436768 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997843646594615, "compression/movement_sparsity/importance_threshold": -1.399504601046729e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9285947573536819, "compression/movement_sparsity/model_sparsity": 0.896694670746973, "compression_loss": 105.13744354248047, "distillation_loss": 3.5574026107788086, "epoch": 4.82, "learning_rate": 3.769881026925485e-05, "loss": 109.2389, "step": 5704, "task_loss": 2.0909719467163086 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997873875404806, "compression/movement_sparsity/importance_threshold": -1.379885665281702e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286185579922831, "compression/movement_sparsity/model_sparsity": 0.8967176537604188, "compression_loss": 105.13697814941406, "distillation_loss": 4.950407028198242, "epoch": 4.82, "learning_rate": 3.769567939887289e-05, "loss": 109.2816, "step": 5705, "task_loss": 3.56109356880188 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997903820378499, "compression/movement_sparsity/importance_threshold": -1.360450943516632e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286697126714409, "compression/movement_sparsity/model_sparsity": 0.8967670511189774, "compression_loss": 105.13648986816406, "distillation_loss": 4.1877617835998535, "epoch": 4.82, "learning_rate": 3.7692548528490924e-05, "loss": 109.2286, "step": 5706, "task_loss": 2.338343381881714 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997933482854545, "compression/movement_sparsity/importance_threshold": -1.3411995668163867e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286213482475099, "compression/movement_sparsity/model_sparsity": 0.8967203481617948, "compression_loss": 105.13603973388672, "distillation_loss": 4.229730606079102, "epoch": 4.82, "learning_rate": 3.7689417658108955e-05, "loss": 108.7791, "step": 5707, "task_loss": 2.1339974403381348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997962864171795, "compression/movement_sparsity/importance_threshold": -1.3221306662484361e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286660161794738, "compression/movement_sparsity/model_sparsity": 0.8967634816128811, "compression_loss": 105.13548278808594, "distillation_loss": 5.187078952789307, "epoch": 4.82, "learning_rate": 3.768628678772699e-05, "loss": 109.2501, "step": 5708, "task_loss": 2.57895565032959 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9997991965669102, "compression/movement_sparsity/importance_threshold": -1.3032433728785153e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9287586431136692, "compression/movement_sparsity/model_sparsity": 0.896852926526933, "compression_loss": 105.13494873046875, "distillation_loss": 5.104189872741699, "epoch": 4.83, "learning_rate": 3.7683155917345026e-05, "loss": 109.1482, "step": 5709, "task_loss": 2.331618547439575 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998020788685317, "compression/movement_sparsity/importance_threshold": -1.2845368177706248e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9287861283200698, "compression/movement_sparsity/model_sparsity": 0.8968794675319395, "compression_loss": 105.1343994140625, "distillation_loss": 4.38733434677124, "epoch": 4.83, "learning_rate": 3.768002504696306e-05, "loss": 108.8024, "step": 5710, "task_loss": 2.4929614067077637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999804933455929, "compression/movement_sparsity/importance_threshold": -1.266010131993102e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9287923885080787, "compression/movement_sparsity/model_sparsity": 0.8968855126632317, "compression_loss": 105.13374328613281, "distillation_loss": 3.201449155807495, "epoch": 4.83, "learning_rate": 3.767689417658109e-05, "loss": 108.8942, "step": 5711, "task_loss": 1.8522369861602783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998077604629874, "compression/movement_sparsity/importance_threshold": -1.2476624466116817e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9287495807462659, "compression/movement_sparsity/model_sparsity": 0.8968441754797293, "compression_loss": 105.13317108154297, "distillation_loss": 4.090569496154785, "epoch": 4.83, "learning_rate": 3.767376330619912e-05, "loss": 108.7453, "step": 5712, "task_loss": 2.556514263153076 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998105600235919, "compression/movement_sparsity/importance_threshold": -1.2294928926920995e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286743392484836, "compression/movement_sparsity/model_sparsity": 0.8967715187588657, "compression_loss": 105.13253784179688, "distillation_loss": 3.6007018089294434, "epoch": 4.83, "learning_rate": 3.767063243581716e-05, "loss": 108.8393, "step": 5713, "task_loss": 2.99871563911438 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998133322716277, "compression/movement_sparsity/importance_threshold": -1.2115006013000904e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286638817534669, "compression/movement_sparsity/model_sparsity": 0.8967614205109739, "compression_loss": 105.13185119628906, "distillation_loss": 5.3461833000183105, "epoch": 4.83, "learning_rate": 3.766750156543519e-05, "loss": 109.444, "step": 5714, "task_loss": 3.4124836921691895 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99981607734098, "compression/movement_sparsity/importance_threshold": -1.1936847035022569e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286594459631065, "compression/movement_sparsity/model_sparsity": 0.8967571371036583, "compression_loss": 105.13123321533203, "distillation_loss": 4.4603986740112305, "epoch": 4.83, "learning_rate": 3.766437069505322e-05, "loss": 109.1355, "step": 5715, "task_loss": 2.530968189239502 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999818795365534, "compression/movement_sparsity/importance_threshold": -1.1760443303625995e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286256528720264, "compression/movement_sparsity/model_sparsity": 0.8967245049092166, "compression_loss": 105.13062286376953, "distillation_loss": 4.510098457336426, "epoch": 4.83, "learning_rate": 3.766123982467126e-05, "loss": 108.907, "step": 5716, "task_loss": 3.4613001346588135 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998214864791746, "compression/movement_sparsity/importance_threshold": -1.1585786129503228e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9286361461395459, "compression/movement_sparsity/model_sparsity": 0.8967346377007158, "compression_loss": 105.13005065917969, "distillation_loss": 3.4357659816741943, "epoch": 4.83, "learning_rate": 3.7658108954289294e-05, "loss": 109.6391, "step": 5717, "task_loss": 1.7554413080215454 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998241508157871, "compression/movement_sparsity/importance_threshold": -1.1412866823294274e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288089988735955, "compression/movement_sparsity/model_sparsity": 0.8969015524115934, "compression_loss": 105.12937927246094, "distillation_loss": 4.820646286010742, "epoch": 4.83, "learning_rate": 3.7654978083907325e-05, "loss": 108.3347, "step": 5718, "task_loss": 2.2108848094940186 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998267885092567, "compression/movement_sparsity/importance_threshold": -1.124167669567383e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288269447458875, "compression/movement_sparsity/model_sparsity": 0.8969188817879642, "compression_loss": 105.12871551513672, "distillation_loss": 3.70078444480896, "epoch": 4.83, "learning_rate": 3.765184721352536e-05, "loss": 108.3261, "step": 5719, "task_loss": 2.2062785625457764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998293996934683, "compression/movement_sparsity/importance_threshold": -1.1072207057290576e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288945190038799, "compression/movement_sparsity/model_sparsity": 0.8969841346623119, "compression_loss": 105.12810516357422, "distillation_loss": 4.789700031280518, "epoch": 4.83, "learning_rate": 3.7648716343143396e-05, "loss": 109.6619, "step": 5720, "task_loss": 3.1604018211364746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998319845023073, "compression/movement_sparsity/importance_threshold": -1.0904449218801862e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9289042372005031, "compression/movement_sparsity/model_sparsity": 0.8969935190089844, "compression_loss": 105.12744140625, "distillation_loss": 2.658677101135254, "epoch": 4.84, "learning_rate": 3.764558547276143e-05, "loss": 108.3436, "step": 5721, "task_loss": 1.555777907371521 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998345430696587, "compression/movement_sparsity/importance_threshold": -1.0738394490882389e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288827617745908, "compression/movement_sparsity/model_sparsity": 0.8969727813300185, "compression_loss": 105.12684631347656, "distillation_loss": 4.045790672302246, "epoch": 4.84, "learning_rate": 3.764245460237946e-05, "loss": 109.1471, "step": 5722, "task_loss": 1.8118047714233398 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998370755294077, "compression/movement_sparsity/importance_threshold": -1.057403418417216e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288844669305628, "compression/movement_sparsity/model_sparsity": 0.8969744279086371, "compression_loss": 105.12631225585938, "distillation_loss": 3.875297784805298, "epoch": 4.84, "learning_rate": 3.76393237319975e-05, "loss": 108.2197, "step": 5723, "task_loss": 1.6815381050109863 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998395820154394, "compression/movement_sparsity/importance_threshold": -1.0411359609354548e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288625503104481, "compression/movement_sparsity/model_sparsity": 0.8969532641918467, "compression_loss": 105.1256332397461, "distillation_loss": 4.167497158050537, "epoch": 4.84, "learning_rate": 3.763619286161553e-05, "loss": 109.6132, "step": 5724, "task_loss": 2.458045721054077 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999842062661639, "compression/movement_sparsity/importance_threshold": -1.0250362077078232e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9288944474588741, "compression/movement_sparsity/model_sparsity": 0.8969840655750971, "compression_loss": 105.12498474121094, "distillation_loss": 2.9875895977020264, "epoch": 4.84, "learning_rate": 3.763306199123356e-05, "loss": 108.5286, "step": 5725, "task_loss": 1.7283819913864136 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998445176018915, "compression/movement_sparsity/importance_threshold": -1.0091032898000563e-06, "compression/movement_sparsity/linear_layer_sparsity": 0.9289046545463703, "compression/movement_sparsity/model_sparsity": 0.8969939220177372, "compression_loss": 105.12431335449219, "distillation_loss": 3.6251533031463623, "epoch": 4.84, "learning_rate": 3.76299311208516e-05, "loss": 109.3802, "step": 5726, "task_loss": 1.9456497430801392 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998469469700823, "compression/movement_sparsity/importance_threshold": -9.933363382787566e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289211098977078, "compression/movement_sparsity/model_sparsity": 0.8970098120771337, "compression_loss": 105.12369537353516, "distillation_loss": 4.362697601318359, "epoch": 4.84, "learning_rate": 3.762680025046963e-05, "loss": 108.9872, "step": 5727, "task_loss": 2.4194047451019287 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998493509000963, "compression/movement_sparsity/importance_threshold": -9.777344842096594e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289619501718607, "compression/movement_sparsity/model_sparsity": 0.89704924936223, "compression_loss": 105.12306213378906, "distillation_loss": 3.7601053714752197, "epoch": 4.84, "learning_rate": 3.762366938008767e-05, "loss": 109.1728, "step": 5728, "task_loss": 2.924100399017334 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998517295258187, "compression/movement_sparsity/importance_threshold": -9.622968586593672e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289169364390353, "compression/movement_sparsity/model_sparsity": 0.8970057819896056, "compression_loss": 105.12232971191406, "distillation_loss": 4.167534828186035, "epoch": 4.84, "learning_rate": 3.76205385097057e-05, "loss": 109.3483, "step": 5729, "task_loss": 2.0552477836608887 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998540829811348, "compression/movement_sparsity/importance_threshold": -9.470225926927478e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289260465031091, "compression/movement_sparsity/model_sparsity": 0.8970145790949526, "compression_loss": 105.12176513671875, "distillation_loss": 5.484816551208496, "epoch": 4.84, "learning_rate": 3.7617407639323734e-05, "loss": 108.7715, "step": 5730, "task_loss": 2.6157913208007812 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998564113999294, "compression/movement_sparsity/importance_threshold": -9.319108173772711e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289890299565617, "compression/movement_sparsity/model_sparsity": 0.8970753988730195, "compression_loss": 105.12110137939453, "distillation_loss": 3.8115243911743164, "epoch": 4.84, "learning_rate": 3.761427676894177e-05, "loss": 108.9376, "step": 5731, "task_loss": 2.3534064292907715 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998587149160879, "compression/movement_sparsity/importance_threshold": -9.169606637786723e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9290152035045224, "compression/movement_sparsity/model_sparsity": 0.8971006732790884, "compression_loss": 105.12053680419922, "distillation_loss": 5.147052764892578, "epoch": 4.84, "learning_rate": 3.7611145898559804e-05, "loss": 109.7836, "step": 5732, "task_loss": 3.0555171966552734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998609936634955, "compression/movement_sparsity/importance_threshold": -9.021712629618192e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289911643825685, "compression/movement_sparsity/model_sparsity": 0.8970774599749266, "compression_loss": 105.11990356445312, "distillation_loss": 5.145715713500977, "epoch": 4.85, "learning_rate": 3.7608015028177836e-05, "loss": 109.3537, "step": 5733, "task_loss": 3.4883439540863037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998632477760372, "compression/movement_sparsity/importance_threshold": -8.875417459933144e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289245917546575, "compression/movement_sparsity/model_sparsity": 0.8970131743215857, "compression_loss": 105.11921691894531, "distillation_loss": 3.1349685192108154, "epoch": 4.85, "learning_rate": 3.760488415779587e-05, "loss": 108.817, "step": 5734, "task_loss": 3.4052882194519043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999865477387598, "compression/movement_sparsity/importance_threshold": -8.730712439397603e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289369928889988, "compression/movement_sparsity/model_sparsity": 0.8970251494388121, "compression_loss": 105.11858367919922, "distillation_loss": 4.993832111358643, "epoch": 4.85, "learning_rate": 3.7601753287413906e-05, "loss": 109.9054, "step": 5735, "task_loss": 3.2055773735046387 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998676826320634, "compression/movement_sparsity/importance_threshold": -8.587588878668923e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289409755609892, "compression/movement_sparsity/model_sparsity": 0.8970289952937675, "compression_loss": 105.11785125732422, "distillation_loss": 4.124655723571777, "epoch": 4.85, "learning_rate": 3.759862241703194e-05, "loss": 109.0483, "step": 5736, "task_loss": 1.6153019666671753 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998698636433183, "compression/movement_sparsity/importance_threshold": -8.446038088404453e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928928800985833, "compression/movement_sparsity/model_sparsity": 0.8970172389527212, "compression_loss": 105.11724853515625, "distillation_loss": 2.836388111114502, "epoch": 4.85, "learning_rate": 3.759549154664997e-05, "loss": 108.8119, "step": 5737, "task_loss": 1.6096569299697876 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998720205552478, "compression/movement_sparsity/importance_threshold": -8.306051379278895e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928865269020669, "compression/movement_sparsity/model_sparsity": 0.8969558895060079, "compression_loss": 105.1166000366211, "distillation_loss": 4.741563320159912, "epoch": 4.85, "learning_rate": 3.759236067626801e-05, "loss": 109.5492, "step": 5738, "task_loss": 3.0475151538848877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998741535017373, "compression/movement_sparsity/importance_threshold": -8.167620061932251e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288698836735442, "compression/movement_sparsity/model_sparsity": 0.8969603456313604, "compression_loss": 105.11592864990234, "distillation_loss": 5.537072658538818, "epoch": 4.85, "learning_rate": 3.758922980588604e-05, "loss": 109.3664, "step": 5739, "task_loss": 2.4839680194854736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998762626166716, "compression/movement_sparsity/importance_threshold": -8.030735447030549e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288854924089794, "compression/movement_sparsity/model_sparsity": 0.8969754181587154, "compression_loss": 105.11534118652344, "distillation_loss": 2.6435699462890625, "epoch": 4.85, "learning_rate": 3.758609893550407e-05, "loss": 108.7142, "step": 5740, "task_loss": 2.391247272491455 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999878348033936, "compression/movement_sparsity/importance_threshold": -7.895388845248485e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288685720151042, "compression/movement_sparsity/model_sparsity": 0.896959079032423, "compression_loss": 105.11473083496094, "distillation_loss": 4.150225639343262, "epoch": 4.85, "learning_rate": 3.7582968065122104e-05, "loss": 109.2325, "step": 5741, "task_loss": 2.6432905197143555 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998804098874158, "compression/movement_sparsity/importance_threshold": -7.761571567226067e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289029255420631, "compression/movement_sparsity/model_sparsity": 0.896992252410047, "compression_loss": 105.11409759521484, "distillation_loss": 4.170088291168213, "epoch": 4.85, "learning_rate": 3.757983719474014e-05, "loss": 109.4872, "step": 5742, "task_loss": 3.0168914794921875 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998824483109959, "compression/movement_sparsity/importance_threshold": -7.629274923637991e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928903223646254, "compression/movement_sparsity/model_sparsity": 0.8969925402734419, "compression_loss": 105.11344146728516, "distillation_loss": 4.674556732177734, "epoch": 4.85, "learning_rate": 3.7576706324358174e-05, "loss": 109.2475, "step": 5743, "task_loss": 1.7718180418014526 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998844634385615, "compression/movement_sparsity/importance_threshold": -7.49849022514161e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288376288000891, "compression/movement_sparsity/model_sparsity": 0.8969291988120361, "compression_loss": 105.11276245117188, "distillation_loss": 5.05418586730957, "epoch": 4.85, "learning_rate": 3.7573575453976206e-05, "loss": 108.7254, "step": 5744, "task_loss": 2.725396156311035 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998864554039979, "compression/movement_sparsity/importance_threshold": -7.369208782385603e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288555985207164, "compression/movement_sparsity/model_sparsity": 0.8969465512174785, "compression_loss": 105.11199951171875, "distillation_loss": 2.597221851348877, "epoch": 4.86, "learning_rate": 3.757044458359424e-05, "loss": 108.2247, "step": 5745, "task_loss": 1.8218852281570435 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99988842434119, "compression/movement_sparsity/importance_threshold": -7.241421906053341e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288097381719889, "compression/movement_sparsity/model_sparsity": 0.8969022663128127, "compression_loss": 105.11126708984375, "distillation_loss": 4.5211639404296875, "epoch": 4.86, "learning_rate": 3.7567313713212276e-05, "loss": 108.642, "step": 5746, "task_loss": 3.855185031890869 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998903703840231, "compression/movement_sparsity/importance_threshold": -7.11512090678483e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928779081136997, "compression/movement_sparsity/model_sparsity": 0.8968726624412849, "compression_loss": 105.11055755615234, "distillation_loss": 5.065799713134766, "epoch": 4.86, "learning_rate": 3.756418284283031e-05, "loss": 108.8801, "step": 5747, "task_loss": 1.449901819229126 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998922936663822, "compression/movement_sparsity/importance_threshold": -6.990297095246095e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288138520098232, "compression/movement_sparsity/model_sparsity": 0.8969062388276619, "compression_loss": 105.10991668701172, "distillation_loss": 4.081058502197266, "epoch": 4.86, "learning_rate": 3.756105197244834e-05, "loss": 108.9332, "step": 5748, "task_loss": 3.5841596126556396 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998941943221527, "compression/movement_sparsity/importance_threshold": -6.866941782103161e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288398586194371, "compression/movement_sparsity/model_sparsity": 0.8969313520302297, "compression_loss": 105.10922241210938, "distillation_loss": 4.641939163208008, "epoch": 4.86, "learning_rate": 3.755792110206637e-05, "loss": 109.0303, "step": 5749, "task_loss": 3.6546993255615234 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998960724852195, "compression/movement_sparsity/importance_threshold": -6.745046278013381e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288025478989044, "compression/movement_sparsity/model_sparsity": 0.8968953230477286, "compression_loss": 105.10852813720703, "distillation_loss": 4.371820449829102, "epoch": 4.86, "learning_rate": 3.755479023168441e-05, "loss": 108.9735, "step": 5750, "task_loss": 2.2587296962738037 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998979282894678, "compression/movement_sparsity/importance_threshold": -6.624601893634105e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928823093239741, "compression/movement_sparsity/model_sparsity": 0.8969151625929026, "compression_loss": 105.10784149169922, "distillation_loss": 3.328397750854492, "epoch": 4.86, "learning_rate": 3.755165936130244e-05, "loss": 109.0704, "step": 5751, "task_loss": 1.6765533685684204 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9998997618687827, "compression/movement_sparsity/importance_threshold": -6.50559993963136e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288364483074932, "compression/movement_sparsity/model_sparsity": 0.8969280588729924, "compression_loss": 105.10711669921875, "distillation_loss": 4.736715316772461, "epoch": 4.86, "learning_rate": 3.7548528490920474e-05, "loss": 110.0682, "step": 5752, "task_loss": 2.4736275672912598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999015733570495, "compression/movement_sparsity/importance_threshold": -6.388031726653823e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288444851964798, "compression/movement_sparsity/model_sparsity": 0.896935819670118, "compression_loss": 105.10639953613281, "distillation_loss": 5.609321594238281, "epoch": 4.86, "learning_rate": 3.754539762053851e-05, "loss": 109.2699, "step": 5753, "task_loss": 2.3937759399414062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999033628881532, "compression/movement_sparsity/importance_threshold": -6.271888565376194e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289154936147513, "compression/movement_sparsity/model_sparsity": 0.8970043887307745, "compression_loss": 105.10568237304688, "distillation_loss": 4.090530872344971, "epoch": 4.86, "learning_rate": 3.7542266750156544e-05, "loss": 108.2874, "step": 5754, "task_loss": 2.1111698150634766 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999905130595979, "compression/movement_sparsity/importance_threshold": -6.157161766455824e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288864821148932, "compression/movement_sparsity/model_sparsity": 0.8969763738651864, "compression_loss": 105.10491943359375, "distillation_loss": 3.9415082931518555, "epoch": 4.86, "learning_rate": 3.7539135879774576e-05, "loss": 109.2202, "step": 5755, "task_loss": 3.248878240585327 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999906876614412, "compression/movement_sparsity/importance_threshold": -6.043842640541391e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288685004700984, "compression/movement_sparsity/model_sparsity": 0.8969590099452082, "compression_loss": 105.10418701171875, "distillation_loss": 3.774620532989502, "epoch": 4.87, "learning_rate": 3.753600500939261e-05, "loss": 108.4395, "step": 5756, "task_loss": 1.9858653545379639 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999086010773374, "compression/movement_sparsity/importance_threshold": -5.931922498307596e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9288881634225299, "compression/movement_sparsity/model_sparsity": 0.8969779974147334, "compression_loss": 105.10344696044922, "distillation_loss": 2.875511884689331, "epoch": 4.87, "learning_rate": 3.7532874139010646e-05, "loss": 109.19, "step": 5757, "task_loss": 1.5887807607650757 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999103041186402, "compression/movement_sparsity/importance_threshold": -5.821392650411789e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928964883517099, "compression/movement_sparsity/model_sparsity": 0.8970520819380354, "compression_loss": 105.10272979736328, "distillation_loss": 3.923691749572754, "epoch": 4.87, "learning_rate": 3.752974326862868e-05, "loss": 108.8235, "step": 5758, "task_loss": 3.617396831512451 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999119858722058, "compression/movement_sparsity/importance_threshold": -5.712244407502648e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289829724794026, "compression/movement_sparsity/model_sparsity": 0.8970695494888358, "compression_loss": 105.1019515991211, "distillation_loss": 3.956963062286377, "epoch": 4.87, "learning_rate": 3.7526612398246716e-05, "loss": 109.7995, "step": 5759, "task_loss": 2.986818313598633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999913646471919, "compression/movement_sparsity/importance_threshold": -5.604469080254873e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289684607673898, "compression/movement_sparsity/model_sparsity": 0.8970555362987739, "compression_loss": 105.10123443603516, "distillation_loss": 4.326057434082031, "epoch": 4.87, "learning_rate": 3.752348152786475e-05, "loss": 109.2446, "step": 5760, "task_loss": 1.8751739263534546 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999152860516652, "compression/movement_sparsity/importance_threshold": -5.498057979325816e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9289656228154924, "compression/movement_sparsity/model_sparsity": 0.8970527958392547, "compression_loss": 105.10054016113281, "distillation_loss": 5.797255992889404, "epoch": 4.87, "learning_rate": 3.752035065748279e-05, "loss": 110.5968, "step": 5761, "task_loss": 3.7239980697631836 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999169047453296, "compression/movement_sparsity/importance_threshold": -5.393002415372827e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9290125682634749, "compression/movement_sparsity/model_sparsity": 0.8970981285666778, "compression_loss": 105.09979248046875, "distillation_loss": 3.190678596496582, "epoch": 4.87, "learning_rate": 3.751721978710082e-05, "loss": 109.0249, "step": 5762, "task_loss": 1.4066734313964844 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999918502686797, "compression/movement_sparsity/importance_threshold": -5.289293699061934e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.928994324286992, "compression/movement_sparsity/model_sparsity": 0.8970805113269122, "compression_loss": 105.09907531738281, "distillation_loss": 3.5723254680633545, "epoch": 4.87, "learning_rate": 3.751408891671885e-05, "loss": 108.6209, "step": 5763, "task_loss": 1.7025810480117798 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999200800099528, "compression/movement_sparsity/importance_threshold": -5.186923141050487e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9290376328638454, "compression/movement_sparsity/model_sparsity": 0.897122332120918, "compression_loss": 105.09833526611328, "distillation_loss": 3.3516628742218018, "epoch": 4.87, "learning_rate": 3.751095804633689e-05, "loss": 108.7405, "step": 5764, "task_loss": 1.2719810009002686 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999216368486822, "compression/movement_sparsity/importance_threshold": -5.085882051987164e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9290725468266833, "compression/movement_sparsity/model_sparsity": 0.8971560466817243, "compression_loss": 105.09757995605469, "distillation_loss": 4.306318283081055, "epoch": 4.87, "learning_rate": 3.750782717595492e-05, "loss": 109.1283, "step": 5765, "task_loss": 2.9801692962646484 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999231733368701, "compression/movement_sparsity/importance_threshold": -4.986161742546666e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9291064949319425, "compression/movement_sparsity/model_sparsity": 0.8971888285651315, "compression_loss": 105.09680938720703, "distillation_loss": 4.873434066772461, "epoch": 4.87, "learning_rate": 3.750469630557295e-05, "loss": 109.2113, "step": 5766, "task_loss": 2.4087841510772705 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999246896084018, "compression/movement_sparsity/importance_threshold": -4.887753523386343e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9291070792161568, "compression/movement_sparsity/model_sparsity": 0.8971893927773854, "compression_loss": 105.09603118896484, "distillation_loss": 4.207455158233643, "epoch": 4.87, "learning_rate": 3.7501565435190984e-05, "loss": 108.8157, "step": 5767, "task_loss": 2.1476244926452637 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999261857971624, "compression/movement_sparsity/importance_threshold": -4.790648705163547e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9291271475902879, "compression/movement_sparsity/model_sparsity": 0.8972087717411276, "compression_loss": 105.09525299072266, "distillation_loss": 4.084499835968018, "epoch": 4.88, "learning_rate": 3.749843456480902e-05, "loss": 108.7705, "step": 5768, "task_loss": 2.635321617126465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999276620370371, "compression/movement_sparsity/importance_threshold": -4.6948385985356306e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929148527622859, "compression/movement_sparsity/model_sparsity": 0.8972294173038072, "compression_loss": 105.09452819824219, "distillation_loss": 4.135584831237793, "epoch": 4.88, "learning_rate": 3.7495303694427055e-05, "loss": 109.277, "step": 5769, "task_loss": 2.4050984382629395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999291184619109, "compression/movement_sparsity/importance_threshold": -4.600314514177292e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9291848247891425, "compression/movement_sparsity/model_sparsity": 0.8972644675507657, "compression_loss": 105.0937271118164, "distillation_loss": 3.8273704051971436, "epoch": 4.88, "learning_rate": 3.7492172824045086e-05, "loss": 108.7908, "step": 5770, "task_loss": 1.7266227006912231 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999305552056691, "compression/movement_sparsity/importance_threshold": -4.50706776273721e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9292123338438785, "compression/movement_sparsity/model_sparsity": 0.8972910315848438, "compression_loss": 105.09294891357422, "distillation_loss": 3.53554105758667, "epoch": 4.88, "learning_rate": 3.748904195366312e-05, "loss": 108.9616, "step": 5771, "task_loss": 1.5998797416687012 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999319724021967, "compression/movement_sparsity/importance_threshold": -4.41508965488141e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9292351805490688, "compression/movement_sparsity/model_sparsity": 0.8973130934354261, "compression_loss": 105.09225463867188, "distillation_loss": 3.103426933288574, "epoch": 4.88, "learning_rate": 3.748591108328116e-05, "loss": 108.9259, "step": 5772, "task_loss": 2.207528829574585 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999933370185379, "compression/movement_sparsity/importance_threshold": -4.3243715012672435e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9292936447429875, "compression/movement_sparsity/model_sparsity": 0.8973695492044268, "compression_loss": 105.09141540527344, "distillation_loss": 3.462790012359619, "epoch": 4.88, "learning_rate": 3.748278021289919e-05, "loss": 108.6414, "step": 5773, "task_loss": 1.6052347421646118 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999347486891009, "compression/movement_sparsity/importance_threshold": -4.2349046125520623e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9292745422264348, "compression/movement_sparsity/model_sparsity": 0.897351102918084, "compression_loss": 105.09064483642578, "distillation_loss": 4.1480712890625, "epoch": 4.88, "learning_rate": 3.747964934251722e-05, "loss": 108.3456, "step": 5774, "task_loss": 1.9174202680587769 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999361080472479, "compression/movement_sparsity/importance_threshold": -4.1466802993932184e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293209749352088, "compression/movement_sparsity/model_sparsity": 0.8973959405204679, "compression_loss": 105.08987426757812, "distillation_loss": 3.3301329612731934, "epoch": 4.88, "learning_rate": 3.747651847213526e-05, "loss": 108.8501, "step": 5775, "task_loss": 1.5561102628707886 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999374483937048, "compression/movement_sparsity/importance_threshold": -4.0596898724654107e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9292768555149562, "compression/movement_sparsity/model_sparsity": 0.8973533367380281, "compression_loss": 105.0890884399414, "distillation_loss": 3.925814628601074, "epoch": 4.88, "learning_rate": 3.747338760175329e-05, "loss": 109.0479, "step": 5776, "task_loss": 2.679823160171509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999387698623569, "compression/movement_sparsity/importance_threshold": -3.973924642425991e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9292543426864597, "compression/movement_sparsity/model_sparsity": 0.897331597294448, "compression_loss": 105.0882339477539, "distillation_loss": 2.61102294921875, "epoch": 4.88, "learning_rate": 3.747025673137132e-05, "loss": 108.4111, "step": 5777, "task_loss": 2.1981050968170166 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999400725870893, "compression/movement_sparsity/importance_threshold": -3.889375919914964e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293216784610994, "compression/movement_sparsity/model_sparsity": 0.8973966198780797, "compression_loss": 105.08746337890625, "distillation_loss": 4.931883811950684, "epoch": 4.88, "learning_rate": 3.7467125860989354e-05, "loss": 109.3806, "step": 5778, "task_loss": 3.32778000831604 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999413567017872, "compression/movement_sparsity/importance_threshold": -3.8060350156157025e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293367744573263, "compression/movement_sparsity/model_sparsity": 0.8974111972803956, "compression_loss": 105.08662414550781, "distillation_loss": 3.8888087272644043, "epoch": 4.88, "learning_rate": 3.746399499060739e-05, "loss": 108.5191, "step": 5779, "task_loss": 1.589375615119934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999426223403356, "compression/movement_sparsity/importance_threshold": -3.723893240185558e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293358324480832, "compression/movement_sparsity/model_sparsity": 0.8974102876320679, "compression_loss": 105.08583068847656, "distillation_loss": 4.3369622230529785, "epoch": 4.89, "learning_rate": 3.7460864120225424e-05, "loss": 109.0256, "step": 5780, "task_loss": 2.55423903465271 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999438696366199, "compression/movement_sparsity/importance_threshold": -3.6429419042732086e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293436070053817, "compression/movement_sparsity/model_sparsity": 0.8974177951094059, "compression_loss": 105.0850601196289, "distillation_loss": 3.21692156791687, "epoch": 4.89, "learning_rate": 3.7457733249843456e-05, "loss": 109.2227, "step": 5781, "task_loss": 1.853084683418274 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999450987245249, "compression/movement_sparsity/importance_threshold": -3.5631723185533537e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293468861514815, "compression/movement_sparsity/model_sparsity": 0.8974209616067493, "compression_loss": 105.08426666259766, "distillation_loss": 4.024066925048828, "epoch": 4.89, "learning_rate": 3.745460237946149e-05, "loss": 109.3727, "step": 5782, "task_loss": 1.6926225423812866 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999946309737936, "compression/movement_sparsity/importance_threshold": -3.4845757936659977e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293161814198192, "compression/movement_sparsity/model_sparsity": 0.8973913116770784, "compression_loss": 105.08343505859375, "distillation_loss": 3.2184667587280273, "epoch": 4.89, "learning_rate": 3.7451471509079527e-05, "loss": 109.5418, "step": 5783, "task_loss": 2.059974431991577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999475028107382, "compression/movement_sparsity/importance_threshold": -3.407143640294513e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929324015597956, "compression/movement_sparsity/model_sparsity": 0.8973988767270954, "compression_loss": 105.0826187133789, "distillation_loss": 3.726748466491699, "epoch": 4.89, "learning_rate": 3.744834063869756e-05, "loss": 108.4402, "step": 5784, "task_loss": 2.5159859657287598 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999486780768168, "compression/movement_sparsity/importance_threshold": -3.3308671690875785e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9293772570064501, "compression/movement_sparsity/model_sparsity": 0.897450289129418, "compression_loss": 105.08187103271484, "distillation_loss": 4.393023490905762, "epoch": 4.89, "learning_rate": 3.744520976831559e-05, "loss": 108.7569, "step": 5785, "task_loss": 2.6508541107177734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999498356700568, "compression/movement_sparsity/importance_threshold": -3.2557376907025454e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9294055530562501, "compression/movement_sparsity/model_sparsity": 0.8974776131228586, "compression_loss": 105.08106994628906, "distillation_loss": 4.559662818908691, "epoch": 4.89, "learning_rate": 3.744207889793363e-05, "loss": 109.6688, "step": 5786, "task_loss": 3.028951406478882 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999509757243433, "compression/movement_sparsity/importance_threshold": -3.181746515814113e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9294513180116365, "compression/movement_sparsity/model_sparsity": 0.897521805911238, "compression_loss": 105.08024597167969, "distillation_loss": 5.316503047943115, "epoch": 4.89, "learning_rate": 3.743894802755166e-05, "loss": 108.8929, "step": 5787, "task_loss": 3.1166834831237793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999520983735616, "compression/movement_sparsity/importance_threshold": -3.108884955070959e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9294896899830887, "compression/movement_sparsity/model_sparsity": 0.8975588596874248, "compression_loss": 105.0794677734375, "distillation_loss": 3.7825193405151367, "epoch": 4.89, "learning_rate": 3.743581715716969e-05, "loss": 108.9189, "step": 5788, "task_loss": 1.889278769493103 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999532037515967, "compression/movement_sparsity/importance_threshold": -3.0371443191391095e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295297313380099, "compression/movement_sparsity/model_sparsity": 0.8975975254986228, "compression_loss": 105.07864379882812, "distillation_loss": 4.679677963256836, "epoch": 4.89, "learning_rate": 3.7432686286787724e-05, "loss": 108.9259, "step": 5789, "task_loss": 1.481547474861145 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999542919923339, "compression/movement_sparsity/importance_threshold": -2.9665159186672424e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295784296386348, "compression/movement_sparsity/model_sparsity": 0.8976445508628078, "compression_loss": 105.07781982421875, "distillation_loss": 3.9746742248535156, "epoch": 4.89, "learning_rate": 3.742955541640576e-05, "loss": 108.9718, "step": 5790, "task_loss": 3.275855302810669 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999553632296582, "compression/movement_sparsity/importance_threshold": -2.896991064321383e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929560412221337, "compression/movement_sparsity/model_sparsity": 0.8976271523992222, "compression_loss": 105.07698822021484, "distillation_loss": 4.912905693054199, "epoch": 4.89, "learning_rate": 3.7426424546023794e-05, "loss": 109.0857, "step": 5791, "task_loss": 2.716425657272339 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999564175974547, "compression/movement_sparsity/importance_threshold": -2.8285610667762306e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295573954069251, "compression/movement_sparsity/model_sparsity": 0.8976242392216662, "compression_loss": 105.076171875, "distillation_loss": 5.949730396270752, "epoch": 4.9, "learning_rate": 3.7423293675641826e-05, "loss": 109.7628, "step": 5792, "task_loss": 3.3814642429351807 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999574552296088, "compression/movement_sparsity/importance_threshold": -2.7612172366717896e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929606224873394, "compression/movement_sparsity/model_sparsity": 0.8976713912457448, "compression_loss": 105.0753402709961, "distillation_loss": 3.7049560546875, "epoch": 4.9, "learning_rate": 3.7420162805259865e-05, "loss": 108.7821, "step": 5793, "task_loss": 2.518496036529541 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999584762600052, "compression/movement_sparsity/importance_threshold": -2.694950884682759e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929585297959193, "compression/movement_sparsity/model_sparsity": 0.8976511832354255, "compression_loss": 105.07455444335938, "distillation_loss": 3.294009208679199, "epoch": 4.9, "learning_rate": 3.7417031934877897e-05, "loss": 109.0502, "step": 5794, "task_loss": 2.564544439315796 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999594808225295, "compression/movement_sparsity/importance_threshold": -2.629753321466491e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295670301343749, "compression/movement_sparsity/model_sparsity": 0.8976335429665881, "compression_loss": 105.07373046875, "distillation_loss": 5.515574932098389, "epoch": 4.9, "learning_rate": 3.7413901064495935e-05, "loss": 109.9283, "step": 5795, "task_loss": 2.9896957874298096 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999604690510666, "compression/movement_sparsity/importance_threshold": -2.565615857671663e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296097425028466, "compression/movement_sparsity/model_sparsity": 0.8976747880338042, "compression_loss": 105.07294464111328, "distillation_loss": 4.799578666687012, "epoch": 4.9, "learning_rate": 3.741077019411397e-05, "loss": 109.6962, "step": 5796, "task_loss": 2.331934928894043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999614410795017, "compression/movement_sparsity/importance_threshold": -2.502529803972975e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295723840856434, "compression/movement_sparsity/model_sparsity": 0.8976387129931599, "compression_loss": 105.07218170166016, "distillation_loss": 4.535719871520996, "epoch": 4.9, "learning_rate": 3.7407639323732e-05, "loss": 108.6176, "step": 5797, "task_loss": 1.8273813724517822 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99996239704172, "compression/movement_sparsity/importance_threshold": -2.440486471027778e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295353118484635, "compression/movement_sparsity/model_sparsity": 0.8976029143013747, "compression_loss": 105.07144165039062, "distillation_loss": 3.792354106903076, "epoch": 4.9, "learning_rate": 3.740450845335004e-05, "loss": 109.1175, "step": 5798, "task_loss": 2.248016834259033 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999633370716063, "compression/movement_sparsity/importance_threshold": -2.3794771694934241e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9294554318494709, "compression/movement_sparsity/model_sparsity": 0.897525778426087, "compression_loss": 105.07067108154297, "distillation_loss": 3.8813838958740234, "epoch": 4.9, "learning_rate": 3.740137758296807e-05, "loss": 109.3079, "step": 5799, "task_loss": 2.1202025413513184 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999642613030463, "compression/movement_sparsity/importance_threshold": -2.3194932100272653e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9294664974770369, "compression/movement_sparsity/model_sparsity": 0.8975364639153044, "compression_loss": 105.06991577148438, "distillation_loss": 3.2871451377868652, "epoch": 4.9, "learning_rate": 3.73982467125861e-05, "loss": 108.4848, "step": 5800, "task_loss": 1.3479297161102295 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999651698699247, "compression/movement_sparsity/importance_threshold": -2.2605259033040004e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929494066152611, "compression/movement_sparsity/model_sparsity": 0.8975630855220614, "compression_loss": 105.06916809082031, "distillation_loss": 4.683316707611084, "epoch": 4.9, "learning_rate": 3.739511584220414e-05, "loss": 109.4346, "step": 5801, "task_loss": 3.404693603515625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999660629061268, "compression/movement_sparsity/importance_threshold": -2.202566559963634e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295200846863925, "compression/movement_sparsity/model_sparsity": 0.8975882102391651, "compression_loss": 105.06843566894531, "distillation_loss": 2.9457459449768066, "epoch": 4.9, "learning_rate": 3.739198497182217e-05, "loss": 109.6814, "step": 5802, "task_loss": 2.7115633487701416 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999669405455378, "compression/movement_sparsity/importance_threshold": -2.1456064906808653e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929537458198638, "compression/movement_sparsity/model_sparsity": 0.8976049869178178, "compression_loss": 105.06761932373047, "distillation_loss": 3.5388832092285156, "epoch": 4.9, "learning_rate": 3.73888541014402e-05, "loss": 108.9372, "step": 5803, "task_loss": 2.2861733436584473 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999678029220427, "compression/movement_sparsity/importance_threshold": -2.089637006113046e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295533292657613, "compression/movement_sparsity/model_sparsity": 0.8976203127649602, "compression_loss": 105.06684875488281, "distillation_loss": 4.335796356201172, "epoch": 4.91, "learning_rate": 3.7385723231058235e-05, "loss": 109.4855, "step": 5804, "task_loss": 3.244521379470825 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999686501695267, "compression/movement_sparsity/importance_threshold": -2.0346494169175278e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295262971777307, "compression/movement_sparsity/model_sparsity": 0.897594209312314, "compression_loss": 105.0660171508789, "distillation_loss": 6.33378791809082, "epoch": 4.91, "learning_rate": 3.738259236067627e-05, "loss": 109.8484, "step": 5805, "task_loss": 3.320375680923462 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999969482421875, "compression/movement_sparsity/importance_threshold": -1.9806350337603362e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295323904273927, "compression/movement_sparsity/model_sparsity": 0.897600093240105, "compression_loss": 105.06522369384766, "distillation_loss": 2.565720558166504, "epoch": 4.91, "learning_rate": 3.7379461490294305e-05, "loss": 108.6511, "step": 5806, "task_loss": 2.683284044265747 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999702998129727, "compression/movement_sparsity/importance_threshold": -1.927585167298823e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295784177144671, "compression/movement_sparsity/model_sparsity": 0.8976445393482719, "compression_loss": 105.0644302368164, "distillation_loss": 3.2452988624572754, "epoch": 4.91, "learning_rate": 3.737633061991234e-05, "loss": 107.9956, "step": 5807, "task_loss": 1.4556864500045776 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999711024767048, "compression/movement_sparsity/importance_threshold": -1.8754911281903403e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295235426950069, "compression/movement_sparsity/model_sparsity": 0.8975915494545454, "compression_loss": 105.06359100341797, "distillation_loss": 3.9798269271850586, "epoch": 4.91, "learning_rate": 3.737319974953037e-05, "loss": 109.2605, "step": 5808, "task_loss": 1.9331167936325073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999718905469568, "compression/movement_sparsity/importance_threshold": -1.8243442270922394e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295282288928878, "compression/movement_sparsity/model_sparsity": 0.8975960746671128, "compression_loss": 105.06273651123047, "distillation_loss": 5.066416263580322, "epoch": 4.91, "learning_rate": 3.737006887914841e-05, "loss": 109.1585, "step": 5809, "task_loss": 3.2637100219726562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999726641576134, "compression/movement_sparsity/importance_threshold": -1.7741357746792197e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295556783267855, "compression/movement_sparsity/model_sparsity": 0.8976225811285118, "compression_loss": 105.06195831298828, "distillation_loss": 4.415713310241699, "epoch": 4.91, "learning_rate": 3.736693800876644e-05, "loss": 109.3685, "step": 5810, "task_loss": 2.7484865188598633 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999734234425601, "compression/movement_sparsity/importance_threshold": -1.7248570815999592e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296163723400521, "compression/movement_sparsity/model_sparsity": 0.897681190115706, "compression_loss": 105.06116485595703, "distillation_loss": 4.238971710205078, "epoch": 4.91, "learning_rate": 3.736380713838447e-05, "loss": 108.9283, "step": 5811, "task_loss": 2.18684720993042 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999741685356818, "compression/movement_sparsity/importance_threshold": -1.6764994585204834e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.92959535003251, "compression/movement_sparsity/model_sparsity": 0.8976608899891002, "compression_loss": 105.0603256225586, "distillation_loss": 3.319906234741211, "epoch": 4.91, "learning_rate": 3.736067626800251e-05, "loss": 108.5104, "step": 5812, "task_loss": 1.6091734170913696 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999748995708638, "compression/movement_sparsity/importance_threshold": -1.6290542160981442e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295729683698576, "compression/movement_sparsity/model_sparsity": 0.8976392772054139, "compression_loss": 105.05948638916016, "distillation_loss": 3.2215921878814697, "epoch": 4.91, "learning_rate": 3.735754539762054e-05, "loss": 108.6719, "step": 5813, "task_loss": 1.8838703632354736 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999756166819911, "compression/movement_sparsity/importance_threshold": -1.5825126649902932e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295750908716968, "compression/movement_sparsity/model_sparsity": 0.8976413267927853, "compression_loss": 105.05870819091797, "distillation_loss": 3.8677010536193848, "epoch": 4.91, "learning_rate": 3.735441452723857e-05, "loss": 108.6031, "step": 5814, "task_loss": 2.5117366313934326 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999976320002949, "compression/movement_sparsity/importance_threshold": -1.536866115862956e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296190791261054, "compression/movement_sparsity/model_sparsity": 0.8976838039153313, "compression_loss": 105.05791473388672, "distillation_loss": 3.194537878036499, "epoch": 4.91, "learning_rate": 3.7351283656856605e-05, "loss": 108.7446, "step": 5815, "task_loss": 1.396252989768982 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999770096676225, "compression/movement_sparsity/importance_threshold": -1.492105879382158e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296694825827022, "compression/movement_sparsity/model_sparsity": 0.8977324758581349, "compression_loss": 105.05705261230469, "distillation_loss": 3.9563002586364746, "epoch": 4.92, "learning_rate": 3.734815278647464e-05, "loss": 108.6688, "step": 5816, "task_loss": 1.7192150354385376 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999776858098969, "compression/movement_sparsity/importance_threshold": -1.4482232661879035e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296816452336908, "compression/movement_sparsity/model_sparsity": 0.8977442206846453, "compression_loss": 105.05619812011719, "distillation_loss": 3.342777967453003, "epoch": 4.92, "learning_rate": 3.7345021916092675e-05, "loss": 108.7127, "step": 5817, "task_loss": 2.0157082080841064 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999783485636571, "compression/movement_sparsity/importance_threshold": -1.405209586954892e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929680130864401, "compression/movement_sparsity/model_sparsity": 0.8977427583385994, "compression_loss": 105.05538177490234, "distillation_loss": 4.1070990562438965, "epoch": 4.92, "learning_rate": 3.734189104571071e-05, "loss": 108.9351, "step": 5818, "task_loss": 1.418753743171692 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999789980627884, "compression/movement_sparsity/importance_threshold": -1.3630561523491486e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296052709399831, "compression/movement_sparsity/model_sparsity": 0.8976704700828813, "compression_loss": 105.0545883178711, "distillation_loss": 4.925287246704102, "epoch": 4.92, "learning_rate": 3.733876017532874e-05, "loss": 109.9352, "step": 5819, "task_loss": 2.953795909881592 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999979634441176, "compression/movement_sparsity/importance_threshold": -1.3217542730193516e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295732307015455, "compression/movement_sparsity/model_sparsity": 0.8976395305252014, "compression_loss": 105.05377197265625, "distillation_loss": 4.291629314422607, "epoch": 4.92, "learning_rate": 3.733562930494678e-05, "loss": 109.4138, "step": 5820, "task_loss": 3.0095012187957764 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999802578327048, "compression/movement_sparsity/importance_threshold": -1.2812952596402e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295272153386387, "compression/movement_sparsity/model_sparsity": 0.8975950959315702, "compression_loss": 105.05302429199219, "distillation_loss": 4.990915775299072, "epoch": 4.92, "learning_rate": 3.733249843456481e-05, "loss": 109.0726, "step": 5821, "task_loss": 3.5502090454101562 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999808683712603, "compression/movement_sparsity/importance_threshold": -1.2416704228430248e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.929551862593142, "compression/movement_sparsity/model_sparsity": 0.8976188964770575, "compression_loss": 105.05221557617188, "distillation_loss": 5.077402591705322, "epoch": 4.92, "learning_rate": 3.732936756418284e-05, "loss": 109.4175, "step": 5822, "task_loss": 3.115377426147461 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999814661907274, "compression/movement_sparsity/importance_threshold": -1.2028710733198722e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295640729408012, "compression/movement_sparsity/model_sparsity": 0.8976306873617111, "compression_loss": 105.05145263671875, "distillation_loss": 4.541425704956055, "epoch": 4.92, "learning_rate": 3.732623669380088e-05, "loss": 108.6568, "step": 5823, "task_loss": 3.6989986896514893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999820514249912, "compression/movement_sparsity/importance_threshold": -1.1648885217194205e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296008709221255, "compression/movement_sparsity/model_sparsity": 0.897666221219173, "compression_loss": 105.05066680908203, "distillation_loss": 4.666583061218262, "epoch": 4.92, "learning_rate": 3.732310582341891e-05, "loss": 108.8762, "step": 5824, "task_loss": 3.015657901763916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999826242079369, "compression/movement_sparsity/importance_threshold": -1.1277140786990214e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295978183352107, "compression/movement_sparsity/model_sparsity": 0.8976632734980097, "compression_loss": 105.0499267578125, "distillation_loss": 5.23804235458374, "epoch": 4.92, "learning_rate": 3.731997495303694e-05, "loss": 109.1425, "step": 5825, "task_loss": 2.889481782913208 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999831846734497, "compression/movement_sparsity/importance_threshold": -1.0913390549247004e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295467113527234, "compression/movement_sparsity/model_sparsity": 0.8976139221975943, "compression_loss": 105.04913330078125, "distillation_loss": 5.273022651672363, "epoch": 4.92, "learning_rate": 3.731684408265498e-05, "loss": 108.9349, "step": 5826, "task_loss": 3.7644948959350586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999837329554148, "compression/movement_sparsity/importance_threshold": -1.0557547610538093e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9295558094926295, "compression/movement_sparsity/model_sparsity": 0.8976227077884056, "compression_loss": 105.04840850830078, "distillation_loss": 3.329468250274658, "epoch": 4.93, "learning_rate": 3.731371321227301e-05, "loss": 109.1443, "step": 5827, "task_loss": 1.8754706382751465 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999842691877171, "compression/movement_sparsity/importance_threshold": -1.0209525077436998e-07, "compression/movement_sparsity/linear_layer_sparsity": 0.9296424981913423, "compression/movement_sparsity/model_sparsity": 0.8977064184636319, "compression_loss": 105.04766082763672, "distillation_loss": 3.6245779991149902, "epoch": 4.93, "learning_rate": 3.731058234189105e-05, "loss": 109.0762, "step": 5828, "task_loss": 1.443039894104004 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999984793504242, "compression/movement_sparsity/importance_threshold": -9.869236056603975e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296031603623115, "compression/movement_sparsity/model_sparsity": 0.8976684320100456, "compression_loss": 105.04692840576172, "distillation_loss": 4.119654655456543, "epoch": 4.93, "learning_rate": 3.7307451471509083e-05, "loss": 109.0485, "step": 5829, "task_loss": 2.481191873550415 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999853060388745, "compression/movement_sparsity/importance_threshold": -9.536593654525805e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296031007414733, "compression/movement_sparsity/model_sparsity": 0.8976683744373667, "compression_loss": 105.04618072509766, "distillation_loss": 5.2365193367004395, "epoch": 4.93, "learning_rate": 3.7304320601127115e-05, "loss": 108.9907, "step": 5830, "task_loss": 3.4590280055999756 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999858069254998, "compression/movement_sparsity/importance_threshold": -9.211510977949477e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296541004064519, "compression/movement_sparsity/model_sparsity": 0.8977176221069599, "compression_loss": 105.04542541503906, "distillation_loss": 6.700359344482422, "epoch": 4.93, "learning_rate": 3.7301189730745154e-05, "loss": 109.5885, "step": 5831, "task_loss": 3.3119540214538574 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999862962980028, "compression/movement_sparsity/importance_threshold": -8.893901133535248e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296359756716455, "compression/movement_sparsity/model_sparsity": 0.8977001200125522, "compression_loss": 105.04462432861328, "distillation_loss": 3.254922389984131, "epoch": 4.93, "learning_rate": 3.7298058860363186e-05, "loss": 108.5004, "step": 5832, "task_loss": 2.543977975845337 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999986774290269, "compression/movement_sparsity/importance_threshold": -8.583677227683162e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295852741108578, "compression/movement_sparsity/model_sparsity": 0.8976511602063538, "compression_loss": 105.04389190673828, "distillation_loss": 2.9491121768951416, "epoch": 4.93, "learning_rate": 3.729492798998122e-05, "loss": 108.8693, "step": 5833, "task_loss": 1.3216373920440674 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999872410361835, "compression/movement_sparsity/importance_threshold": -8.280752366966737e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295728491281812, "compression/movement_sparsity/model_sparsity": 0.8976391620600559, "compression_loss": 105.04313659667969, "distillation_loss": 4.098935127258301, "epoch": 4.93, "learning_rate": 3.729179711959925e-05, "loss": 108.558, "step": 5834, "task_loss": 1.8437068462371826 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999876966696312, "compression/movement_sparsity/importance_threshold": -7.985039658306436e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295820665097637, "compression/movement_sparsity/model_sparsity": 0.8976480627962251, "compression_loss": 105.04236602783203, "distillation_loss": 3.3687472343444824, "epoch": 4.93, "learning_rate": 3.728866624921729e-05, "loss": 109.1985, "step": 5835, "task_loss": 1.9737662076950073 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999881413244974, "compression/movement_sparsity/importance_threshold": -7.69645220801557e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295999766095527, "compression/movement_sparsity/model_sparsity": 0.8976653576289885, "compression_loss": 105.04154205322266, "distillation_loss": 3.747542381286621, "epoch": 4.93, "learning_rate": 3.728553537883532e-05, "loss": 108.6235, "step": 5836, "task_loss": 2.3853402137756348 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999885751346672, "compression/movement_sparsity/importance_threshold": -7.414903122841127e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295208239847859, "compression/movement_sparsity/model_sparsity": 0.8975889241403844, "compression_loss": 105.040771484375, "distillation_loss": 3.137768030166626, "epoch": 4.93, "learning_rate": 3.728240450845335e-05, "loss": 108.6107, "step": 5837, "task_loss": 1.3879072666168213 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999889982340258, "compression/movement_sparsity/importance_threshold": -7.140305509356626e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295384121320488, "compression/movement_sparsity/model_sparsity": 0.8976059080806813, "compression_loss": 105.03995513916016, "distillation_loss": 4.013400554656982, "epoch": 4.93, "learning_rate": 3.727927363807139e-05, "loss": 108.3305, "step": 5838, "task_loss": 2.2190613746643066 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999894107564582, "compression/movement_sparsity/importance_threshold": -6.872572474135585e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295386506154015, "compression/movement_sparsity/model_sparsity": 0.8976061383713971, "compression_loss": 105.03913879394531, "distillation_loss": 4.416202545166016, "epoch": 4.94, "learning_rate": 3.727614276768942e-05, "loss": 109.1782, "step": 5839, "task_loss": 1.944057583808899 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999898128358496, "compression/movement_sparsity/importance_threshold": -6.611617123838259e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296116026729977, "compression/movement_sparsity/model_sparsity": 0.8976765843013882, "compression_loss": 105.0383071899414, "distillation_loss": 5.046707630157471, "epoch": 4.94, "learning_rate": 3.727301189730745e-05, "loss": 108.973, "step": 5840, "task_loss": 2.308835983276367 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999902046060852, "compression/movement_sparsity/importance_threshold": -6.357352565038166e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296147625774213, "compression/movement_sparsity/model_sparsity": 0.8976796356533737, "compression_loss": 105.03749084472656, "distillation_loss": 3.8488473892211914, "epoch": 4.94, "learning_rate": 3.7269881026925485e-05, "loss": 107.8432, "step": 5841, "task_loss": 1.9789059162139893 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999905862010502, "compression/movement_sparsity/importance_threshold": -6.109691904222087e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295614853964242, "compression/movement_sparsity/model_sparsity": 0.8976281887074438, "compression_loss": 105.03675079345703, "distillation_loss": 2.8331284523010254, "epoch": 4.94, "learning_rate": 3.7266750156543524e-05, "loss": 108.5889, "step": 5842, "task_loss": 1.4484412670135498 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999909577546297, "compression/movement_sparsity/importance_threshold": -5.868548248137012e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295891613895071, "compression/movement_sparsity/model_sparsity": 0.8976549139450228, "compression_loss": 105.03593444824219, "distillation_loss": 4.286626815795898, "epoch": 4.94, "learning_rate": 3.7263619286161555e-05, "loss": 108.7465, "step": 5843, "task_loss": 2.4161601066589355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999913194007086, "compression/movement_sparsity/importance_threshold": -5.633834703443197e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9295782388519526, "compression/movement_sparsity/model_sparsity": 0.8976443666302351, "compression_loss": 105.03511810302734, "distillation_loss": 5.048055171966553, "epoch": 4.94, "learning_rate": 3.726048841577959e-05, "loss": 108.8136, "step": 5844, "task_loss": 3.8491601943969727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999916712731723, "compression/movement_sparsity/importance_threshold": -5.4054643766274224e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296191148986084, "compression/movement_sparsity/model_sparsity": 0.8976838384589387, "compression_loss": 105.03425598144531, "distillation_loss": 4.136790752410889, "epoch": 4.94, "learning_rate": 3.725735754539762e-05, "loss": 108.7362, "step": 5845, "task_loss": 1.6181446313858032 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999920135059059, "compression/movement_sparsity/importance_threshold": -5.183350374263207e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.929644537224008, "compression/movement_sparsity/model_sparsity": 0.8977083874492527, "compression_loss": 105.03343200683594, "distillation_loss": 3.3991150856018066, "epoch": 4.94, "learning_rate": 3.725422667501566e-05, "loss": 108.7294, "step": 5846, "task_loss": 2.2250685691833496 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999923462327946, "compression/movement_sparsity/importance_threshold": -4.967405803010805e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296138324923456, "compression/movement_sparsity/model_sparsity": 0.8976787375195817, "compression_loss": 105.03260040283203, "distillation_loss": 3.5926742553710938, "epoch": 4.94, "learning_rate": 3.725109580463369e-05, "loss": 108.4421, "step": 5847, "task_loss": 1.5684707164764404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999926695877234, "compression/movement_sparsity/importance_threshold": -4.75754376953047e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296773525333418, "compression/movement_sparsity/model_sparsity": 0.8977400754517593, "compression_loss": 105.03175354003906, "distillation_loss": 2.9994311332702637, "epoch": 4.94, "learning_rate": 3.724796493425172e-05, "loss": 108.3515, "step": 5848, "task_loss": 1.2715611457824707 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999929837045775, "compression/movement_sparsity/importance_threshold": -4.553677380395721e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296350336624022, "compression/movement_sparsity/model_sparsity": 0.8976992103642245, "compression_loss": 105.03096008300781, "distillation_loss": 3.948836326599121, "epoch": 4.94, "learning_rate": 3.724483406386976e-05, "loss": 108.9885, "step": 5849, "task_loss": 1.8018351793289185 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999993288717242, "compression/movement_sparsity/importance_threshold": -4.355719742093339e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296439529397938, "compression/movement_sparsity/model_sparsity": 0.8977078232369988, "compression_loss": 105.03012084960938, "distillation_loss": 4.354881286621094, "epoch": 4.94, "learning_rate": 3.724170319348779e-05, "loss": 109.3283, "step": 5850, "task_loss": 2.270348310470581 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999935847596021, "compression/movement_sparsity/importance_threshold": -4.163583961370315e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296455507782571, "compression/movement_sparsity/model_sparsity": 0.8977093661847952, "compression_loss": 105.02934265136719, "distillation_loss": 3.1824419498443604, "epoch": 4.95, "learning_rate": 3.723857232310582e-05, "loss": 108.9413, "step": 5851, "task_loss": 2.55712890625 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999938719655429, "compression/movement_sparsity/importance_threshold": -3.977183144800167e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296470293750438, "compression/movement_sparsity/model_sparsity": 0.8977107939872337, "compression_loss": 105.02853393554688, "distillation_loss": 4.624808311462402, "epoch": 4.95, "learning_rate": 3.7235441452723855e-05, "loss": 109.2483, "step": 5852, "task_loss": 2.04655385017395 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999941504689496, "compression/movement_sparsity/importance_threshold": -3.796430398956413e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296435117455913, "compression/movement_sparsity/model_sparsity": 0.8977073971991744, "compression_loss": 105.02774047851562, "distillation_loss": 3.562480926513672, "epoch": 4.95, "learning_rate": 3.7232310582341894e-05, "loss": 108.142, "step": 5853, "task_loss": 1.2160377502441406 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999944204037072, "compression/movement_sparsity/importance_threshold": -3.621238830412571e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296703768952749, "compression/movement_sparsity/model_sparsity": 0.8977333394483195, "compression_loss": 105.02702331542969, "distillation_loss": 5.873029708862305, "epoch": 4.95, "learning_rate": 3.7229179711959925e-05, "loss": 109.7053, "step": 5854, "task_loss": 3.0365538597106934 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999946819037011, "compression/movement_sparsity/importance_threshold": -3.451521545828895e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9296888951276134, "compression/movement_sparsity/model_sparsity": 0.8977512215224084, "compression_loss": 105.02629089355469, "distillation_loss": 4.0027360916137695, "epoch": 4.95, "learning_rate": 3.722604884157796e-05, "loss": 108.3919, "step": 5855, "task_loss": 2.040935754776001 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999949351028162, "compression/movement_sparsity/importance_threshold": -3.2871916517789035e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9297779686598533, "compression/movement_sparsity/model_sparsity": 0.8978372351047936, "compression_loss": 105.02550506591797, "distillation_loss": 5.2696027755737305, "epoch": 4.95, "learning_rate": 3.722291797119599e-05, "loss": 108.7274, "step": 5856, "task_loss": 2.622962236404419 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999951801349377, "compression/movement_sparsity/importance_threshold": -3.1281622550095867e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9298083037423189, "compression/movement_sparsity/model_sparsity": 0.8978665280838549, "compression_loss": 105.0247573852539, "distillation_loss": 3.088512420654297, "epoch": 4.95, "learning_rate": 3.721978710081403e-05, "loss": 109.1465, "step": 5857, "task_loss": 1.5845016241073608 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999954171339508, "compression/movement_sparsity/importance_threshold": -2.974346461834254e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9297543588079343, "compression/movement_sparsity/model_sparsity": 0.8978144363239204, "compression_loss": 105.02397155761719, "distillation_loss": 3.061530590057373, "epoch": 4.95, "learning_rate": 3.721665623043206e-05, "loss": 107.788, "step": 5858, "task_loss": 1.8449136018753052 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999956462337406, "compression/movement_sparsity/importance_threshold": -2.8256573790866324e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9297679404348715, "compression/movement_sparsity/model_sparsity": 0.8978275513801904, "compression_loss": 105.02317810058594, "distillation_loss": 4.876717567443848, "epoch": 4.95, "learning_rate": 3.721352536005009e-05, "loss": 109.1539, "step": 5859, "task_loss": 2.2667832374572754 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999958675681923, "compression/movement_sparsity/importance_threshold": -2.6820081133402396e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9297284952883321, "compression/movement_sparsity/model_sparsity": 0.8977894612957821, "compression_loss": 105.0223617553711, "distillation_loss": 5.415078163146973, "epoch": 4.95, "learning_rate": 3.721039448966813e-05, "loss": 108.5213, "step": 5860, "task_loss": 2.887016773223877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999960812711909, "compression/movement_sparsity/importance_threshold": -2.5433117710818576e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.92979281424856, "compression/movement_sparsity/model_sparsity": 0.8978515707018578, "compression_loss": 105.02149963378906, "distillation_loss": 5.2202324867248535, "epoch": 4.95, "learning_rate": 3.720726361928616e-05, "loss": 109.4125, "step": 5861, "task_loss": 3.3400514125823975 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999962874766216, "compression/movement_sparsity/importance_threshold": -2.4094814590584768e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.929752629803627, "compression/movement_sparsity/model_sparsity": 0.8978127667162302, "compression_loss": 105.02072143554688, "distillation_loss": 4.337552547454834, "epoch": 4.95, "learning_rate": 3.72041327489042e-05, "loss": 108.948, "step": 5862, "task_loss": 2.1340208053588867 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999964863183696, "compression/movement_sparsity/importance_threshold": -2.2804302838436152e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9298198105640876, "compression/movement_sparsity/model_sparsity": 0.8978776396108966, "compression_loss": 105.01993560791016, "distillation_loss": 3.694293975830078, "epoch": 4.96, "learning_rate": 3.720100187852223e-05, "loss": 108.8433, "step": 5863, "task_loss": 2.263529062271118 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99999667793032, "compression/movement_sparsity/importance_threshold": -2.156071352010791e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9298116782817599, "compression/movement_sparsity/model_sparsity": 0.8978697866974847, "compression_loss": 105.01907348632812, "distillation_loss": 5.115645885467529, "epoch": 4.96, "learning_rate": 3.719787100814027e-05, "loss": 108.8686, "step": 5864, "task_loss": 1.9654172658920288 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999996862446358, "compression/movement_sparsity/importance_threshold": -2.0363177701335222e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9298210149050188, "compression/movement_sparsity/model_sparsity": 0.8978788025790119, "compression_loss": 105.01824188232422, "distillation_loss": 3.4674150943756104, "epoch": 4.96, "learning_rate": 3.71947401377583e-05, "loss": 108.7556, "step": 5865, "task_loss": 2.8198800086975098 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999970400003686, "compression/movement_sparsity/importance_threshold": -1.921082644872063e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9298292664290227, "compression/movement_sparsity/model_sparsity": 0.8978867706377818, "compression_loss": 105.01741027832031, "distillation_loss": 5.767710208892822, "epoch": 4.96, "learning_rate": 3.7191609267376334e-05, "loss": 110.4747, "step": 5866, "task_loss": 2.665541648864746 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999972107262372, "compression/movement_sparsity/importance_threshold": -1.8102790827131954e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9298978065445938, "compression/movement_sparsity/model_sparsity": 0.8979529561895288, "compression_loss": 105.01654052734375, "distillation_loss": 4.096896648406982, "epoch": 4.96, "learning_rate": 3.7188478396994366e-05, "loss": 108.9746, "step": 5867, "task_loss": 2.604422092437744 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999973747578486, "compression/movement_sparsity/importance_threshold": -1.7038201904039096e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9299330782324606, "compression/movement_sparsity/model_sparsity": 0.897987016186409, "compression_loss": 105.01566314697266, "distillation_loss": 3.9631705284118652, "epoch": 4.96, "learning_rate": 3.7185347526612404e-05, "loss": 109.6184, "step": 5868, "task_loss": 2.363389730453491 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999975322290882, "compression/movement_sparsity/importance_threshold": -1.601619074517724e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.929982086561444, "compression/movement_sparsity/model_sparsity": 0.8980343409285245, "compression_loss": 105.01478576660156, "distillation_loss": 4.925991058349609, "epoch": 4.96, "learning_rate": 3.7182216656230436e-05, "loss": 108.9358, "step": 5869, "task_loss": 3.003563165664673 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999976832738409, "compression/movement_sparsity/importance_threshold": -1.5035888417148924e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.930090870742786, "compression/movement_sparsity/model_sparsity": 0.8981393880385782, "compression_loss": 105.01395416259766, "distillation_loss": 4.540091514587402, "epoch": 4.96, "learning_rate": 3.717908578584847e-05, "loss": 108.8415, "step": 5870, "task_loss": 2.739344596862793 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999978280259921, "compression/movement_sparsity/importance_threshold": -1.4096425983954608e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.930128658430024, "compression/movement_sparsity/model_sparsity": 0.898175877602511, "compression_loss": 105.01310729980469, "distillation_loss": 3.9631290435791016, "epoch": 4.96, "learning_rate": 3.71759549154665e-05, "loss": 108.6039, "step": 5871, "task_loss": 2.1863090991973877 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999979666194269, "compression/movement_sparsity/importance_threshold": -1.3196934513064196e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9301542596179382, "compression/movement_sparsity/model_sparsity": 0.8982005993108619, "compression_loss": 105.01225280761719, "distillation_loss": 5.819742202758789, "epoch": 4.96, "learning_rate": 3.717282404508454e-05, "loss": 109.5251, "step": 5872, "task_loss": 2.7736802101135254 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999980991880303, "compression/movement_sparsity/importance_threshold": -1.2336545070212868e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9301571333423385, "compression/movement_sparsity/model_sparsity": 0.8982033743139884, "compression_loss": 105.01140594482422, "distillation_loss": 3.3389406204223633, "epoch": 4.96, "learning_rate": 3.716969317470257e-05, "loss": 108.4314, "step": 5873, "task_loss": 1.611488699913025 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999982258656874, "compression/movement_sparsity/importance_threshold": -1.1514388722870528e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.930207644116444, "compression/movement_sparsity/model_sparsity": 0.8982521498876141, "compression_loss": 105.01058959960938, "distillation_loss": 3.3973934650421143, "epoch": 4.96, "learning_rate": 3.71665623043206e-05, "loss": 108.4638, "step": 5874, "task_loss": 2.188612222671509 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999983467862836, "compression/movement_sparsity/importance_threshold": -1.0729596534170271e-08, "compression/movement_sparsity/linear_layer_sparsity": 0.9302423792167672, "compression/movement_sparsity/model_sparsity": 0.8982856917303836, "compression_loss": 105.00977325439453, "distillation_loss": 4.852319240570068, "epoch": 4.97, "learning_rate": 3.716343143393864e-05, "loss": 108.7519, "step": 5875, "task_loss": 3.25911808013916 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999984620837039, "compression/movement_sparsity/importance_threshold": -9.981299572449365e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9302333883943698, "compression/movement_sparsity/model_sparsity": 0.8982770097703945, "compression_loss": 105.00887298583984, "distillation_loss": 3.726454734802246, "epoch": 4.97, "learning_rate": 3.716030056355667e-05, "loss": 108.3286, "step": 5876, "task_loss": 1.6425044536590576 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999985718918334, "compression/movement_sparsity/importance_threshold": -9.268628903442988e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9302012169900883, "compression/movement_sparsity/model_sparsity": 0.8982459435528208, "compression_loss": 105.00801849365234, "distillation_loss": 3.367762804031372, "epoch": 4.97, "learning_rate": 3.7157169693174704e-05, "loss": 108.4263, "step": 5877, "task_loss": 1.4342310428619385 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999986763445573, "compression/movement_sparsity/importance_threshold": -8.590715592886322e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.930189042414932, "compression/movement_sparsity/model_sparsity": 0.8982341872117746, "compression_loss": 105.00724029541016, "distillation_loss": 3.842991352081299, "epoch": 4.97, "learning_rate": 3.7154038822792736e-05, "loss": 109.03, "step": 5878, "task_loss": 1.8248943090438843 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999987755757607, "compression/movement_sparsity/importance_threshold": -7.946690706514548e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.930243404695184, "compression/movement_sparsity/model_sparsity": 0.8982866819804619, "compression_loss": 105.00643157958984, "distillation_loss": 4.234536170959473, "epoch": 4.97, "learning_rate": 3.7150907952410774e-05, "loss": 109.3784, "step": 5879, "task_loss": 3.9255595207214355 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999988697193287, "compression/movement_sparsity/importance_threshold": -7.335685310062845e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.930236524450458, "compression/movement_sparsity/model_sparsity": 0.8982800380933085, "compression_loss": 105.0056381225586, "distillation_loss": 3.182211399078369, "epoch": 4.97, "learning_rate": 3.7147777082028806e-05, "loss": 108.853, "step": 5880, "task_loss": 1.2207002639770508 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999989589091466, "compression/movement_sparsity/importance_threshold": -6.756830470133757e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9302182447014723, "compression/movement_sparsity/model_sparsity": 0.8982623863099355, "compression_loss": 105.00486755371094, "distillation_loss": 4.596044540405273, "epoch": 4.97, "learning_rate": 3.714464621164684e-05, "loss": 108.5877, "step": 5881, "task_loss": 2.9559078216552734 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999990432790993, "compression/movement_sparsity/importance_threshold": -6.209257254197187e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9302880726271479, "compression/movement_sparsity/model_sparsity": 0.8983298154315483, "compression_loss": 105.00404357910156, "distillation_loss": 3.2453160285949707, "epoch": 4.97, "learning_rate": 3.714151534126487e-05, "loss": 108.537, "step": 5882, "task_loss": 1.8550745248794556 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999991229630721, "compression/movement_sparsity/importance_threshold": -5.692096725386231e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9302877029779512, "compression/movement_sparsity/model_sparsity": 0.8983294584809387, "compression_loss": 105.00318145751953, "distillation_loss": 4.434046745300293, "epoch": 4.97, "learning_rate": 3.713838447088291e-05, "loss": 109.4242, "step": 5883, "task_loss": 3.094370126724243 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999991980949503, "compression/movement_sparsity/importance_threshold": -5.204479951170793e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9303363416577378, "compression/movement_sparsity/model_sparsity": 0.8983764262724446, "compression_loss": 105.00244903564453, "distillation_loss": 3.451528310775757, "epoch": 4.97, "learning_rate": 3.713525360050094e-05, "loss": 108.6369, "step": 5884, "task_loss": 2.1146397590637207 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999992688086187, "compression/movement_sparsity/importance_threshold": -4.745537999020777e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9303663428635097, "compression/movement_sparsity/model_sparsity": 0.8984053968445036, "compression_loss": 105.00171661376953, "distillation_loss": 5.24539041519165, "epoch": 4.97, "learning_rate": 3.713212273011897e-05, "loss": 108.9414, "step": 5885, "task_loss": 2.549071788787842 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999993352379626, "compression/movement_sparsity/importance_threshold": -4.31440193293664e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9303665217260242, "compression/movement_sparsity/model_sparsity": 0.8984055695625406, "compression_loss": 105.00091552734375, "distillation_loss": 5.120429992675781, "epoch": 4.97, "learning_rate": 3.712899185973701e-05, "loss": 108.9327, "step": 5886, "task_loss": 2.5347039699554443 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999993975168672, "compression/movement_sparsity/importance_threshold": -3.910202818653563e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304096633645308, "compression/movement_sparsity/model_sparsity": 0.8984472291530452, "compression_loss": 105.00008392333984, "distillation_loss": 3.2169055938720703, "epoch": 4.98, "learning_rate": 3.712586098935504e-05, "loss": 108.2346, "step": 5887, "task_loss": 2.6340842247009277 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999994557792176, "compression/movement_sparsity/importance_threshold": -3.53207172364145e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304491681319084, "compression/movement_sparsity/model_sparsity": 0.8984853768101325, "compression_loss": 104.9992446899414, "distillation_loss": 3.6245455741882324, "epoch": 4.98, "learning_rate": 3.7122730118973074e-05, "loss": 108.8884, "step": 5888, "task_loss": 2.3445160388946533 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999995101588989, "compression/movement_sparsity/importance_threshold": -3.1791397136354815e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304400819161699, "compression/movement_sparsity/model_sparsity": 0.8984766027338571, "compression_loss": 104.99844360351562, "distillation_loss": 4.6158857345581055, "epoch": 4.98, "learning_rate": 3.7119599248591105e-05, "loss": 108.79, "step": 5889, "task_loss": 2.1928961277008057 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999995607897962, "compression/movement_sparsity/importance_threshold": -2.850537854370838e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304706077853176, "compression/movement_sparsity/model_sparsity": 0.8985060799454911, "compression_loss": 104.99752044677734, "distillation_loss": 4.579881191253662, "epoch": 4.98, "learning_rate": 3.7116468378209144e-05, "loss": 109.5875, "step": 5890, "task_loss": 2.7166640758514404 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999996078057948, "compression/movement_sparsity/importance_threshold": -2.5453972124500623e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304795628352122, "compression/movement_sparsity/model_sparsity": 0.8985147273618728, "compression_loss": 104.99668884277344, "distillation_loss": 1.7118351459503174, "epoch": 4.98, "learning_rate": 3.7113337507827176e-05, "loss": 107.8428, "step": 5891, "task_loss": 0.6658555269241333 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999996513407796, "compression/movement_sparsity/importance_threshold": -2.2628488536083347e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9303511276256063, "compression/movement_sparsity/model_sparsity": 0.8983907042968298, "compression_loss": 104.99574279785156, "distillation_loss": 2.422452449798584, "epoch": 4.98, "learning_rate": 3.711020663744521e-05, "loss": 108.7053, "step": 5892, "task_loss": 1.428648591041565 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999999691528636, "compression/movement_sparsity/importance_threshold": -2.0020238435808357e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304045955932855, "compression/movement_sparsity/model_sparsity": 0.8984423354753325, "compression_loss": 104.9948501586914, "distillation_loss": 3.9350972175598145, "epoch": 4.98, "learning_rate": 3.7107075767063246e-05, "loss": 109.0328, "step": 5893, "task_loss": 2.228180408477783 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999999728503249, "compression/movement_sparsity/importance_threshold": -1.7620532481027462e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304284439285573, "compression/movement_sparsity/model_sparsity": 0.8984653645469216, "compression_loss": 104.99398803710938, "distillation_loss": 3.5972797870635986, "epoch": 4.98, "learning_rate": 3.710394489668128e-05, "loss": 108.4251, "step": 5894, "task_loss": 1.651883840560913 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999997623985037, "compression/movement_sparsity/importance_threshold": -1.5420681337766085e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304166390025977, "compression/movement_sparsity/model_sparsity": 0.898453965156485, "compression_loss": 104.99309539794922, "distillation_loss": 4.016494274139404, "epoch": 4.98, "learning_rate": 3.7100814026299316e-05, "loss": 109.0504, "step": 5895, "task_loss": 2.750636100769043 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999997933482855, "compression/movement_sparsity/importance_threshold": -1.341199566337603e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9303668794510533, "compression/movement_sparsity/model_sparsity": 0.8984059149986144, "compression_loss": 104.99220275878906, "distillation_loss": 5.310848236083984, "epoch": 4.98, "learning_rate": 3.709768315591735e-05, "loss": 108.557, "step": 5896, "task_loss": 2.40729022026062 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999998214864791, "compression/movement_sparsity/importance_threshold": -1.1585786132556342e-09, "compression/movement_sparsity/linear_layer_sparsity": 0.9304357176708151, "compression/movement_sparsity/model_sparsity": 0.8984723884137563, "compression_loss": 104.99131774902344, "distillation_loss": 5.400449752807617, "epoch": 4.98, "learning_rate": 3.709455228553538e-05, "loss": 108.9949, "step": 5897, "task_loss": 2.3834540843963623 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999998469469701, "compression/movement_sparsity/importance_threshold": -9.933363385311589e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.930482090758751, "compression/movement_sparsity/model_sparsity": 0.8985171684434613, "compression_loss": 104.990478515625, "distillation_loss": 2.696636915206909, "epoch": 4.99, "learning_rate": 3.709142141515342e-05, "loss": 108.2066, "step": 5898, "task_loss": 1.4422996044158936 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999998698636433, "compression/movement_sparsity/importance_threshold": -8.446038087667196e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9305017298628473, "compression/movement_sparsity/model_sparsity": 0.8985361328839149, "compression_loss": 104.98971557617188, "distillation_loss": 2.885970115661621, "epoch": 4.99, "learning_rate": 3.708829054477145e-05, "loss": 108.3009, "step": 5899, "task_loss": 3.1885762214660645 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999998903703841, "compression/movement_sparsity/importance_threshold": -7.115120905648586e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9304469979333987, "compression/movement_sparsity/model_sparsity": 0.898483281164618, "compression_loss": 104.98881530761719, "distillation_loss": 3.5655529499053955, "epoch": 4.99, "learning_rate": 3.708515967438948e-05, "loss": 108.4088, "step": 5900, "task_loss": 2.8052148818969727 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999086010773, "compression/movement_sparsity/importance_threshold": -5.931922496607567e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.930427144194285, "compression/movement_sparsity/model_sparsity": 0.89846410946252, "compression_loss": 104.98799896240234, "distillation_loss": 2.5799503326416016, "epoch": 4.99, "learning_rate": 3.708202880400752e-05, "loss": 108.0319, "step": 5901, "task_loss": 1.7395423650741577 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999246896084, "compression/movement_sparsity/importance_threshold": -4.88775352656956e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9303764545576649, "compression/movement_sparsity/model_sparsity": 0.8984151611708574, "compression_loss": 104.98712921142578, "distillation_loss": 3.89786434173584, "epoch": 4.99, "learning_rate": 3.707889793362555e-05, "loss": 108.6453, "step": 5902, "task_loss": 2.501769781112671 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999387698624, "compression/movement_sparsity/importance_threshold": -3.9739246442127563e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.930336401278576, "compression/movement_sparsity/model_sparsity": 0.8983764838451236, "compression_loss": 104.98628997802734, "distillation_loss": 5.991863250732422, "epoch": 4.99, "learning_rate": 3.7075767063243584e-05, "loss": 108.9484, "step": 5903, "task_loss": 2.943528652191162 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999509757244, "compression/movement_sparsity/importance_threshold": -3.181746515562578e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9303121952182752, "compression/movement_sparsity/model_sparsity": 0.8983531093374606, "compression_loss": 104.9854507446289, "distillation_loss": 4.563507080078125, "epoch": 4.99, "learning_rate": 3.7072636192861616e-05, "loss": 108.6827, "step": 5904, "task_loss": 2.827897548675537 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999614410795, "compression/movement_sparsity/importance_threshold": -2.502529806644449e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9303490528204377, "compression/movement_sparsity/model_sparsity": 0.8983887007676016, "compression_loss": 104.9845962524414, "distillation_loss": 4.131813049316406, "epoch": 4.99, "learning_rate": 3.7069505322479655e-05, "loss": 108.3062, "step": 5905, "task_loss": 2.0629215240478516 }, { "compression/movement_sparsity/importance_regularization_factor": 0.999999970299813, "compression/movement_sparsity/importance_threshold": -1.9275851661365584e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9303410278556188, "compression/movement_sparsity/model_sparsity": 0.8983809514850118, "compression_loss": 104.9837875366211, "distillation_loss": 4.133953094482422, "epoch": 4.99, "learning_rate": 3.7066374452097686e-05, "loss": 108.9738, "step": 5906, "task_loss": 2.20127010345459 }, { "compression/movement_sparsity/importance_regularization_factor": 0.99999997768581, "compression/movement_sparsity/importance_threshold": -1.4482232600643297e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9303589618037431, "compression/movement_sparsity/model_sparsity": 0.8983982693468469, "compression_loss": 104.9829330444336, "distillation_loss": 6.626056671142578, "epoch": 4.99, "learning_rate": 3.706324358171572e-05, "loss": 109.5947, "step": 5907, "task_loss": 3.99991774559021 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999837329554, "compression/movement_sparsity/importance_threshold": -1.0557547631268038e-10, "compression/movement_sparsity/linear_layer_sparsity": 0.9303832036365468, "compression/movement_sparsity/model_sparsity": 0.8984216783981172, "compression_loss": 104.98204040527344, "distillation_loss": 5.680552005767822, "epoch": 4.99, "learning_rate": 3.706011271133375e-05, "loss": 109.4342, "step": 5908, "task_loss": 3.048506021499634 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999885751346, "compression/movement_sparsity/importance_threshold": -7.414903153285524e-11, "compression/movement_sparsity/linear_layer_sparsity": 0.9303594149221133, "compression/movement_sparsity/model_sparsity": 0.898398706899207, "compression_loss": 104.98112487792969, "distillation_loss": 3.593851089477539, "epoch": 4.99, "learning_rate": 3.705698184095179e-05, "loss": 108.6613, "step": 5909, "task_loss": 2.221275806427002 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999923462328, "compression/movement_sparsity/importance_threshold": -4.967405826949989e-11, "compression/movement_sparsity/linear_layer_sparsity": 0.9304084232510966, "compression/movement_sparsity/model_sparsity": 0.8984460316413225, "compression_loss": 104.980224609375, "distillation_loss": 4.287415027618408, "epoch": 5.0, "learning_rate": 3.705385097056982e-05, "loss": 108.3476, "step": 5910, "task_loss": 2.8877999782562256 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999951801349, "compression/movement_sparsity/importance_threshold": -3.128162225779496e-11, "compression/movement_sparsity/linear_layer_sparsity": 0.9304496212502785, "compression/movement_sparsity/model_sparsity": 0.8984858143624928, "compression_loss": 104.9793701171875, "distillation_loss": 3.685818672180176, "epoch": 5.0, "learning_rate": 3.705072010018785e-05, "loss": 108.7305, "step": 5911, "task_loss": 2.8629908561706543 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999972107262, "compression/movement_sparsity/importance_threshold": -1.8102790967644555e-11, "compression/movement_sparsity/linear_layer_sparsity": 0.9305547685604916, "compression/movement_sparsity/model_sparsity": 0.898587349539129, "compression_loss": 104.978515625, "distillation_loss": 4.127431869506836, "epoch": 5.0, "learning_rate": 3.704758922980589e-05, "loss": 109.2301, "step": 5912, "task_loss": 1.9811686277389526 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999985718918, "compression/movement_sparsity/importance_threshold": -9.268629266867556e-12, "compression/movement_sparsity/linear_layer_sparsity": 0.9305747176929464, "compression/movement_sparsity/model_sparsity": 0.8986066133575132, "compression_loss": 104.97766876220703, "distillation_loss": 3.798346996307373, "epoch": 5.0, "learning_rate": 3.704445835942392e-05, "loss": 108.8374, "step": 5913, "task_loss": 2.233083724975586 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999993975168, "compression/movement_sparsity/importance_threshold": -3.910202890644587e-12, "compression/movement_sparsity/linear_layer_sparsity": 0.9305806678525966, "compression/movement_sparsity/model_sparsity": 0.8986123591108747, "compression_loss": 104.9768295288086, "distillation_loss": 4.093029975891113, "epoch": 5.0, "learning_rate": 3.7041327489041954e-05, "loss": 108.5995, "step": 5914, "task_loss": 1.9554533958435059 }, { "compression/movement_sparsity/importance_regularization_factor": 0.9999999998214865, "compression/movement_sparsity/importance_threshold": -1.15857844151801e-12, "compression/movement_sparsity/linear_layer_sparsity": 0.9305855090646569, "compression/movement_sparsity/model_sparsity": 0.8986170340124073, "compression_loss": 104.97590637207031, "distillation_loss": 3.8832297325134277, "epoch": 5.0, "learning_rate": 3.7038196618659986e-05, "loss": 108.1315, "step": 5915, "task_loss": 2.5057430267333984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 9.000826835632324, "epoch": 5.0, "learning_rate": 3.7035065748278025e-05, "loss": 90.0422, "step": 5916, "task_loss": 3.850513219833374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 5.997817516326904, "epoch": 5.0, "learning_rate": 3.7031934877896056e-05, "loss": 6.9637, "step": 5917, "task_loss": 3.4759128093719482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 6.3032073974609375, "epoch": 5.0, "learning_rate": 3.702880400751409e-05, "loss": 5.9677, "step": 5918, "task_loss": 3.2788095474243164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 5.429514408111572, "epoch": 5.0, "learning_rate": 3.702567313713212e-05, "loss": 4.8695, "step": 5919, "task_loss": 1.5329869985580444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 4.706071376800537, "epoch": 5.0, "learning_rate": 3.702254226675016e-05, "loss": 4.2367, "step": 5920, "task_loss": 2.9755706787109375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 4.352665424346924, "epoch": 5.01, "learning_rate": 3.701941139636819e-05, "loss": 4.5389, "step": 5921, "task_loss": 1.8469088077545166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 4.477771282196045, "epoch": 5.01, "learning_rate": 3.701628052598622e-05, "loss": 3.4231, "step": 5922, "task_loss": 2.3364388942718506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.9568328857421875, "epoch": 5.01, "learning_rate": 3.701314965560426e-05, "loss": 3.3357, "step": 5923, "task_loss": 1.9797382354736328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 4.140225410461426, "epoch": 5.01, "learning_rate": 3.701001878522229e-05, "loss": 3.5332, "step": 5924, "task_loss": 2.8653242588043213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 3.0301475524902344, "epoch": 5.01, "learning_rate": 3.7006887914840324e-05, "loss": 3.0495, "step": 5925, "task_loss": 1.6632204055786133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7874016761779785, "epoch": 5.01, "learning_rate": 3.7003757044458356e-05, "loss": 2.8889, "step": 5926, "task_loss": 0.9640318751335144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.948028087615967, "epoch": 5.01, "learning_rate": 3.7000626174076394e-05, "loss": 2.7324, "step": 5927, "task_loss": 2.2383995056152344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0992584228515625, "epoch": 5.01, "learning_rate": 3.6997495303694426e-05, "loss": 2.4645, "step": 5928, "task_loss": 1.2853058576583862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 3.0800678730010986, "epoch": 5.01, "learning_rate": 3.6994364433312465e-05, "loss": 2.266, "step": 5929, "task_loss": 1.8980129957199097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.3871052265167236, "epoch": 5.01, "learning_rate": 3.6991233562930497e-05, "loss": 2.6891, "step": 5930, "task_loss": 1.5989551544189453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.231372117996216, "epoch": 5.01, "learning_rate": 3.6988102692548535e-05, "loss": 2.6882, "step": 5931, "task_loss": 2.4627528190612793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.4644775390625, "epoch": 5.01, "learning_rate": 3.698497182216657e-05, "loss": 2.5585, "step": 5932, "task_loss": 0.9926791787147522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9256411790847778, "epoch": 5.02, "learning_rate": 3.69818409517846e-05, "loss": 2.2289, "step": 5933, "task_loss": 0.7596468329429626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 3.1811389923095703, "epoch": 5.02, "learning_rate": 3.697871008140264e-05, "loss": 2.5144, "step": 5934, "task_loss": 2.554513454437256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.8120813369750977, "epoch": 5.02, "learning_rate": 3.697557921102067e-05, "loss": 2.1219, "step": 5935, "task_loss": 0.9783168435096741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.399418830871582, "epoch": 5.02, "learning_rate": 3.69724483406387e-05, "loss": 1.9448, "step": 5936, "task_loss": 0.9865663051605225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.3714680671691895, "epoch": 5.02, "learning_rate": 3.696931747025673e-05, "loss": 1.9862, "step": 5937, "task_loss": 1.6002672910690308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.4161133766174316, "epoch": 5.02, "learning_rate": 3.696618659987477e-05, "loss": 2.4183, "step": 5938, "task_loss": 1.8111305236816406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.949773907661438, "epoch": 5.02, "learning_rate": 3.69630557294928e-05, "loss": 2.1504, "step": 5939, "task_loss": 2.6163156032562256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.1158225536346436, "epoch": 5.02, "learning_rate": 3.6959924859110835e-05, "loss": 2.259, "step": 5940, "task_loss": 0.9267263412475586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.5204405784606934, "epoch": 5.02, "learning_rate": 3.6956793988728867e-05, "loss": 2.4099, "step": 5941, "task_loss": 1.7076491117477417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.609302520751953, "epoch": 5.02, "learning_rate": 3.6953663118346905e-05, "loss": 2.2476, "step": 5942, "task_loss": 1.96578049659729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9356658458709717, "epoch": 5.02, "learning_rate": 3.695053224796494e-05, "loss": 2.045, "step": 5943, "task_loss": 0.9752883911132812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9122169017791748, "epoch": 5.02, "learning_rate": 3.694740137758297e-05, "loss": 2.3798, "step": 5944, "task_loss": 1.2333855628967285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.4048056602478027, "epoch": 5.03, "learning_rate": 3.6944270507201e-05, "loss": 2.5255, "step": 5945, "task_loss": 1.2122989892959595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.123961091041565, "epoch": 5.03, "learning_rate": 3.694113963681904e-05, "loss": 1.497, "step": 5946, "task_loss": 0.5156141519546509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7284070253372192, "epoch": 5.03, "learning_rate": 3.693800876643707e-05, "loss": 2.0039, "step": 5947, "task_loss": 0.5543457865715027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.440006732940674, "epoch": 5.03, "learning_rate": 3.69348778960551e-05, "loss": 1.8492, "step": 5948, "task_loss": 1.9144911766052246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.417327880859375, "epoch": 5.03, "learning_rate": 3.693174702567314e-05, "loss": 2.2107, "step": 5949, "task_loss": 0.9075900912284851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 3.22514009475708, "epoch": 5.03, "learning_rate": 3.692861615529117e-05, "loss": 1.9141, "step": 5950, "task_loss": 2.153141975402832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7346683740615845, "epoch": 5.03, "learning_rate": 3.6925485284909205e-05, "loss": 2.0249, "step": 5951, "task_loss": 1.7719463109970093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.2246463298797607, "epoch": 5.03, "learning_rate": 3.6922354414527236e-05, "loss": 1.7283, "step": 5952, "task_loss": 1.9073352813720703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8003265857696533, "epoch": 5.03, "learning_rate": 3.6919223544145275e-05, "loss": 1.6944, "step": 5953, "task_loss": 1.5720432996749878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.1229920387268066, "epoch": 5.03, "learning_rate": 3.691609267376331e-05, "loss": 1.7314, "step": 5954, "task_loss": 1.081828236579895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.222240924835205, "epoch": 5.03, "learning_rate": 3.691296180338134e-05, "loss": 1.8864, "step": 5955, "task_loss": 1.7379320859909058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.202996253967285, "epoch": 5.03, "learning_rate": 3.690983093299937e-05, "loss": 2.1104, "step": 5956, "task_loss": 1.7596657276153564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.306777000427246, "epoch": 5.04, "learning_rate": 3.690670006261741e-05, "loss": 1.7663, "step": 5957, "task_loss": 0.9345565438270569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.605772614479065, "epoch": 5.04, "learning_rate": 3.690356919223544e-05, "loss": 2.1022, "step": 5958, "task_loss": 2.3774261474609375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.173579692840576, "epoch": 5.04, "learning_rate": 3.690043832185347e-05, "loss": 2.6815, "step": 5959, "task_loss": 2.218050003051758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.746980905532837, "epoch": 5.04, "learning_rate": 3.689730745147151e-05, "loss": 1.6944, "step": 5960, "task_loss": 1.2658509016036987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5376797914505005, "epoch": 5.04, "learning_rate": 3.689417658108954e-05, "loss": 1.9987, "step": 5961, "task_loss": 0.4516722559928894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6917290687561035, "epoch": 5.04, "learning_rate": 3.689104571070758e-05, "loss": 2.0703, "step": 5962, "task_loss": 0.6651500463485718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7132363319396973, "epoch": 5.04, "learning_rate": 3.688791484032561e-05, "loss": 2.1628, "step": 5963, "task_loss": 1.764732003211975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9585083723068237, "epoch": 5.04, "learning_rate": 3.6884783969943645e-05, "loss": 1.5841, "step": 5964, "task_loss": 1.1662193536758423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.213394045829773, "epoch": 5.04, "learning_rate": 3.6881653099561683e-05, "loss": 1.5147, "step": 5965, "task_loss": 1.3309000730514526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7660572528839111, "epoch": 5.04, "learning_rate": 3.6878522229179715e-05, "loss": 2.1006, "step": 5966, "task_loss": 0.8854557275772095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7952508926391602, "epoch": 5.04, "learning_rate": 3.687539135879775e-05, "loss": 1.7339, "step": 5967, "task_loss": 1.0924453735351562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.467158317565918, "epoch": 5.04, "learning_rate": 3.6872260488415786e-05, "loss": 1.9528, "step": 5968, "task_loss": 1.750985860824585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.5437986850738525, "epoch": 5.05, "learning_rate": 3.686912961803382e-05, "loss": 1.7642, "step": 5969, "task_loss": 2.318575620651245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8282569646835327, "epoch": 5.05, "learning_rate": 3.686599874765185e-05, "loss": 1.7379, "step": 5970, "task_loss": 1.550215482711792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8546302318572998, "epoch": 5.05, "learning_rate": 3.686286787726989e-05, "loss": 1.653, "step": 5971, "task_loss": 1.4718409776687622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.1863625049591064, "epoch": 5.05, "learning_rate": 3.685973700688792e-05, "loss": 1.9743, "step": 5972, "task_loss": 1.917562484741211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4061282873153687, "epoch": 5.05, "learning_rate": 3.685660613650595e-05, "loss": 1.4838, "step": 5973, "task_loss": 0.7856062054634094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8904027938842773, "epoch": 5.05, "learning_rate": 3.685347526612398e-05, "loss": 2.1134, "step": 5974, "task_loss": 1.411091685295105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.1201744079589844, "epoch": 5.05, "learning_rate": 3.685034439574202e-05, "loss": 1.7033, "step": 5975, "task_loss": 0.8142039179801941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9026157855987549, "epoch": 5.05, "learning_rate": 3.6847213525360053e-05, "loss": 1.5803, "step": 5976, "task_loss": 2.0806117057800293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.2868051528930664, "epoch": 5.05, "learning_rate": 3.6844082654978085e-05, "loss": 1.839, "step": 5977, "task_loss": 2.077691078186035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3912049531936646, "epoch": 5.05, "learning_rate": 3.684095178459612e-05, "loss": 1.7276, "step": 5978, "task_loss": 0.7529770731925964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6635133028030396, "epoch": 5.05, "learning_rate": 3.6837820914214156e-05, "loss": 1.6024, "step": 5979, "task_loss": 0.9368799924850464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1554813385009766, "epoch": 5.05, "learning_rate": 3.683469004383219e-05, "loss": 1.2934, "step": 5980, "task_loss": 0.8032420873641968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.3139145374298096, "epoch": 5.06, "learning_rate": 3.683155917345022e-05, "loss": 1.7192, "step": 5981, "task_loss": 1.5404441356658936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.061272382736206, "epoch": 5.06, "learning_rate": 3.682842830306825e-05, "loss": 1.8044, "step": 5982, "task_loss": 1.0391591787338257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2766306400299072, "epoch": 5.06, "learning_rate": 3.682529743268629e-05, "loss": 1.2582, "step": 5983, "task_loss": 0.45615658164024353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9375619888305664, "epoch": 5.06, "learning_rate": 3.682216656230432e-05, "loss": 1.9791, "step": 5984, "task_loss": 1.0698951482772827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9789780378341675, "epoch": 5.06, "learning_rate": 3.681903569192235e-05, "loss": 1.4356, "step": 5985, "task_loss": 1.067476749420166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6734176874160767, "epoch": 5.06, "learning_rate": 3.681590482154039e-05, "loss": 1.37, "step": 5986, "task_loss": 1.337386131286621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.28910493850708, "epoch": 5.06, "learning_rate": 3.681277395115842e-05, "loss": 1.8298, "step": 5987, "task_loss": 2.060173988342285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2183783054351807, "epoch": 5.06, "learning_rate": 3.6809643080776455e-05, "loss": 1.4865, "step": 5988, "task_loss": 1.1408032178878784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.578479290008545, "epoch": 5.06, "learning_rate": 3.680651221039449e-05, "loss": 1.6176, "step": 5989, "task_loss": 1.6124451160430908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0004606246948242, "epoch": 5.06, "learning_rate": 3.6803381340012525e-05, "loss": 1.5837, "step": 5990, "task_loss": 0.6709921360015869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4867041110992432, "epoch": 5.06, "learning_rate": 3.680025046963056e-05, "loss": 1.8379, "step": 5991, "task_loss": 2.1278719902038574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9536556601524353, "epoch": 5.07, "learning_rate": 3.679711959924859e-05, "loss": 1.4091, "step": 5992, "task_loss": 0.340570867061615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.814920425415039, "epoch": 5.07, "learning_rate": 3.679398872886662e-05, "loss": 1.7588, "step": 5993, "task_loss": 1.7739384174346924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9559295177459717, "epoch": 5.07, "learning_rate": 3.679085785848466e-05, "loss": 1.4324, "step": 5994, "task_loss": 1.4366692304611206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6477056741714478, "epoch": 5.07, "learning_rate": 3.678772698810269e-05, "loss": 1.6622, "step": 5995, "task_loss": 1.6348475217819214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5722053050994873, "epoch": 5.07, "learning_rate": 3.678459611772073e-05, "loss": 1.734, "step": 5996, "task_loss": 1.2362687587738037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7015202045440674, "epoch": 5.07, "learning_rate": 3.678146524733876e-05, "loss": 1.4745, "step": 5997, "task_loss": 1.5282188653945923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7679411172866821, "epoch": 5.07, "learning_rate": 3.67783343769568e-05, "loss": 1.5284, "step": 5998, "task_loss": 0.8445683717727661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.754122257232666, "epoch": 5.07, "learning_rate": 3.677520350657483e-05, "loss": 1.4825, "step": 5999, "task_loss": 1.0538872480392456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9628117084503174, "epoch": 5.07, "learning_rate": 3.6772072636192864e-05, "loss": 1.8028, "step": 6000, "task_loss": 1.0165584087371826 }, { "epoch": 5.07, "eval_accuracy": 0.8178613861386138, "eval_loss": 1.0960134267807007, "eval_runtime": 207.7018, "eval_samples_per_second": 121.569, "eval_steps_per_second": 0.953, "step": 6000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6523845195770264, "epoch": 5.07, "learning_rate": 3.67689417658109e-05, "loss": 1.4472, "step": 6001, "task_loss": 1.808275580406189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.838679552078247, "epoch": 5.07, "learning_rate": 3.6765810895428934e-05, "loss": 2.1218, "step": 6002, "task_loss": 1.855918526649475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.214550018310547, "epoch": 5.07, "learning_rate": 3.6762680025046966e-05, "loss": 1.499, "step": 6003, "task_loss": 1.545353651046753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9415740966796875, "epoch": 5.08, "learning_rate": 3.6759549154665e-05, "loss": 1.6119, "step": 6004, "task_loss": 1.5009706020355225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0637869834899902, "epoch": 5.08, "learning_rate": 3.6756418284283036e-05, "loss": 1.3529, "step": 6005, "task_loss": 1.0170619487762451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5663542747497559, "epoch": 5.08, "learning_rate": 3.675328741390107e-05, "loss": 1.3602, "step": 6006, "task_loss": 1.2244681119918823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.200021505355835, "epoch": 5.08, "learning_rate": 3.67501565435191e-05, "loss": 1.2983, "step": 6007, "task_loss": 1.0255045890808105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6236199140548706, "epoch": 5.08, "learning_rate": 3.674702567313714e-05, "loss": 1.6156, "step": 6008, "task_loss": 1.3306267261505127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0778517723083496, "epoch": 5.08, "learning_rate": 3.674389480275517e-05, "loss": 1.5484, "step": 6009, "task_loss": 1.3374556303024292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0541187524795532, "epoch": 5.08, "learning_rate": 3.67407639323732e-05, "loss": 1.5415, "step": 6010, "task_loss": 0.6332448124885559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0979843139648438, "epoch": 5.08, "learning_rate": 3.6737633061991233e-05, "loss": 1.7573, "step": 6011, "task_loss": 1.933428406715393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7017626762390137, "epoch": 5.08, "learning_rate": 3.673450219160927e-05, "loss": 1.6169, "step": 6012, "task_loss": 1.3782187700271606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.6085729598999023, "epoch": 5.08, "learning_rate": 3.6731371321227304e-05, "loss": 1.7749, "step": 6013, "task_loss": 2.632516860961914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.057474374771118, "epoch": 5.08, "learning_rate": 3.6728240450845336e-05, "loss": 1.5496, "step": 6014, "task_loss": 1.130401611328125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.78172767162323, "epoch": 5.08, "learning_rate": 3.672510958046337e-05, "loss": 1.8263, "step": 6015, "task_loss": 0.8669301271438599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2942357063293457, "epoch": 5.09, "learning_rate": 3.6721978710081406e-05, "loss": 1.5423, "step": 6016, "task_loss": 1.8826055526733398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.2126517295837402, "epoch": 5.09, "learning_rate": 3.671884783969944e-05, "loss": 1.9687, "step": 6017, "task_loss": 1.4537452459335327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.379738211631775, "epoch": 5.09, "learning_rate": 3.671571696931747e-05, "loss": 1.5038, "step": 6018, "task_loss": 1.226047396659851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7885830402374268, "epoch": 5.09, "learning_rate": 3.671258609893551e-05, "loss": 1.3329, "step": 6019, "task_loss": 0.49756959080696106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3430421352386475, "epoch": 5.09, "learning_rate": 3.670945522855354e-05, "loss": 1.5537, "step": 6020, "task_loss": 0.7341465950012207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0137717723846436, "epoch": 5.09, "learning_rate": 3.670632435817157e-05, "loss": 1.6651, "step": 6021, "task_loss": 2.3463075160980225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9903634786605835, "epoch": 5.09, "learning_rate": 3.6703193487789603e-05, "loss": 1.777, "step": 6022, "task_loss": 1.1965830326080322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6544281244277954, "epoch": 5.09, "learning_rate": 3.670006261740764e-05, "loss": 1.5786, "step": 6023, "task_loss": 1.7827749252319336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1703182458877563, "epoch": 5.09, "learning_rate": 3.6696931747025674e-05, "loss": 1.3544, "step": 6024, "task_loss": 0.5278299450874329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0364441871643066, "epoch": 5.09, "learning_rate": 3.6693800876643706e-05, "loss": 1.3531, "step": 6025, "task_loss": 2.153074264526367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0061378479003906, "epoch": 5.09, "learning_rate": 3.669067000626174e-05, "loss": 1.6563, "step": 6026, "task_loss": 1.102484941482544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8499491214752197, "epoch": 5.09, "learning_rate": 3.6687539135879776e-05, "loss": 1.479, "step": 6027, "task_loss": 1.3090089559555054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5169695615768433, "epoch": 5.1, "learning_rate": 3.668440826549781e-05, "loss": 1.5775, "step": 6028, "task_loss": 1.8864103555679321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6625351905822754, "epoch": 5.1, "learning_rate": 3.668127739511584e-05, "loss": 1.7378, "step": 6029, "task_loss": 0.9808271527290344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1402636766433716, "epoch": 5.1, "learning_rate": 3.667814652473388e-05, "loss": 1.4421, "step": 6030, "task_loss": 1.6643121242523193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9662770628929138, "epoch": 5.1, "learning_rate": 3.667501565435191e-05, "loss": 1.1957, "step": 6031, "task_loss": 0.8123306632041931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9677282571792603, "epoch": 5.1, "learning_rate": 3.667188478396995e-05, "loss": 1.4333, "step": 6032, "task_loss": 1.3732589483261108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9190096855163574, "epoch": 5.1, "learning_rate": 3.666875391358798e-05, "loss": 1.6426, "step": 6033, "task_loss": 0.8639971613883972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9879666566848755, "epoch": 5.1, "learning_rate": 3.666562304320602e-05, "loss": 1.0838, "step": 6034, "task_loss": 0.8072118163108826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8992698788642883, "epoch": 5.1, "learning_rate": 3.666249217282405e-05, "loss": 1.4834, "step": 6035, "task_loss": 0.3369576930999756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2346242666244507, "epoch": 5.1, "learning_rate": 3.665936130244208e-05, "loss": 1.3303, "step": 6036, "task_loss": 1.1526074409484863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.700272560119629, "epoch": 5.1, "learning_rate": 3.6656230432060114e-05, "loss": 1.3917, "step": 6037, "task_loss": 0.8953315615653992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.238112211227417, "epoch": 5.1, "learning_rate": 3.665309956167815e-05, "loss": 1.3462, "step": 6038, "task_loss": 1.3071163892745972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9652623534202576, "epoch": 5.1, "learning_rate": 3.6649968691296184e-05, "loss": 1.3393, "step": 6039, "task_loss": 1.3994866609573364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.234447956085205, "epoch": 5.11, "learning_rate": 3.6646837820914216e-05, "loss": 1.0473, "step": 6040, "task_loss": 1.4961812496185303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4927879571914673, "epoch": 5.11, "learning_rate": 3.664370695053225e-05, "loss": 1.5, "step": 6041, "task_loss": 0.7423383593559265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.2933192253112793, "epoch": 5.11, "learning_rate": 3.6640576080150286e-05, "loss": 1.8012, "step": 6042, "task_loss": 1.65071702003479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8601855039596558, "epoch": 5.11, "learning_rate": 3.663744520976832e-05, "loss": 1.6141, "step": 6043, "task_loss": 0.5284629464149475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8594944477081299, "epoch": 5.11, "learning_rate": 3.663431433938635e-05, "loss": 1.8594, "step": 6044, "task_loss": 1.6446025371551514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6238431930541992, "epoch": 5.11, "learning_rate": 3.663118346900439e-05, "loss": 1.5368, "step": 6045, "task_loss": 0.9025294184684753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0207973718643188, "epoch": 5.11, "learning_rate": 3.662805259862242e-05, "loss": 1.4353, "step": 6046, "task_loss": 1.7087682485580444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3068840503692627, "epoch": 5.11, "learning_rate": 3.662492172824045e-05, "loss": 1.5365, "step": 6047, "task_loss": 0.8572789430618286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.44142484664917, "epoch": 5.11, "learning_rate": 3.6621790857858484e-05, "loss": 1.4812, "step": 6048, "task_loss": 0.9783939719200134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4646228551864624, "epoch": 5.11, "learning_rate": 3.661865998747652e-05, "loss": 1.4726, "step": 6049, "task_loss": 1.3746429681777954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5482678413391113, "epoch": 5.11, "learning_rate": 3.6615529117094554e-05, "loss": 1.4632, "step": 6050, "task_loss": 1.2443714141845703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.821457862854004, "epoch": 5.11, "learning_rate": 3.6612398246712586e-05, "loss": 1.6107, "step": 6051, "task_loss": 1.4731169939041138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.14410400390625, "epoch": 5.12, "learning_rate": 3.660926737633062e-05, "loss": 1.3619, "step": 6052, "task_loss": 1.25746750831604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.029381275177002, "epoch": 5.12, "learning_rate": 3.6606136505948656e-05, "loss": 1.2983, "step": 6053, "task_loss": 1.4755830764770508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9186090230941772, "epoch": 5.12, "learning_rate": 3.660300563556669e-05, "loss": 1.3888, "step": 6054, "task_loss": 1.8502864837646484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.962740421295166, "epoch": 5.12, "learning_rate": 3.659987476518472e-05, "loss": 1.5138, "step": 6055, "task_loss": 1.620478868484497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9987142086029053, "epoch": 5.12, "learning_rate": 3.659674389480276e-05, "loss": 1.3749, "step": 6056, "task_loss": 0.7368186712265015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.570122718811035, "epoch": 5.12, "learning_rate": 3.659361302442079e-05, "loss": 1.6741, "step": 6057, "task_loss": 1.9157252311706543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1776857376098633, "epoch": 5.12, "learning_rate": 3.659048215403882e-05, "loss": 1.3756, "step": 6058, "task_loss": 1.2139581441879272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3253580331802368, "epoch": 5.12, "learning_rate": 3.6587351283656854e-05, "loss": 1.3592, "step": 6059, "task_loss": 1.5181820392608643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.517677664756775, "epoch": 5.12, "learning_rate": 3.658422041327489e-05, "loss": 1.5278, "step": 6060, "task_loss": 0.5217088460922241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2481828927993774, "epoch": 5.12, "learning_rate": 3.6581089542892924e-05, "loss": 1.5011, "step": 6061, "task_loss": 1.7071433067321777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1698055267333984, "epoch": 5.12, "learning_rate": 3.6577958672510956e-05, "loss": 0.9664, "step": 6062, "task_loss": 0.6245997548103333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0523216724395752, "epoch": 5.13, "learning_rate": 3.6574827802128995e-05, "loss": 1.1251, "step": 6063, "task_loss": 0.4606989026069641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2589434385299683, "epoch": 5.13, "learning_rate": 3.6571696931747026e-05, "loss": 1.4038, "step": 6064, "task_loss": 0.6535540223121643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.092182159423828, "epoch": 5.13, "learning_rate": 3.6568566061365065e-05, "loss": 1.2733, "step": 6065, "task_loss": 1.222464919090271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.502394199371338, "epoch": 5.13, "learning_rate": 3.65654351909831e-05, "loss": 1.5295, "step": 6066, "task_loss": 1.335970401763916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4826412200927734, "epoch": 5.13, "learning_rate": 3.656230432060113e-05, "loss": 1.6997, "step": 6067, "task_loss": 1.4730424880981445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.640082061290741, "epoch": 5.13, "learning_rate": 3.655917345021917e-05, "loss": 1.2961, "step": 6068, "task_loss": 0.873726487159729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4122364521026611, "epoch": 5.13, "learning_rate": 3.65560425798372e-05, "loss": 1.4451, "step": 6069, "task_loss": 0.9408376216888428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.295285701751709, "epoch": 5.13, "learning_rate": 3.655291170945523e-05, "loss": 1.7018, "step": 6070, "task_loss": 1.5292812585830688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6588457822799683, "epoch": 5.13, "learning_rate": 3.654978083907327e-05, "loss": 1.6356, "step": 6071, "task_loss": 2.3951594829559326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.46422278881073, "epoch": 5.13, "learning_rate": 3.65466499686913e-05, "loss": 1.6475, "step": 6072, "task_loss": 1.712710976600647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.087580919265747, "epoch": 5.13, "learning_rate": 3.654351909830933e-05, "loss": 1.1927, "step": 6073, "task_loss": 0.18028557300567627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5609101057052612, "epoch": 5.13, "learning_rate": 3.6540388227927364e-05, "loss": 1.4121, "step": 6074, "task_loss": 0.674487829208374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7105937004089355, "epoch": 5.14, "learning_rate": 3.65372573575454e-05, "loss": 1.6557, "step": 6075, "task_loss": 1.3429621458053589 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3105881214141846, "epoch": 5.14, "learning_rate": 3.6534126487163435e-05, "loss": 1.4829, "step": 6076, "task_loss": 1.4961820840835571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9168694019317627, "epoch": 5.14, "learning_rate": 3.6530995616781467e-05, "loss": 1.2917, "step": 6077, "task_loss": 2.2581276893615723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.701941728591919, "epoch": 5.14, "learning_rate": 3.65278647463995e-05, "loss": 1.7306, "step": 6078, "task_loss": 1.4787929058074951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.123155951499939, "epoch": 5.14, "learning_rate": 3.652473387601754e-05, "loss": 1.2781, "step": 6079, "task_loss": 1.224151611328125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9255150556564331, "epoch": 5.14, "learning_rate": 3.652160300563557e-05, "loss": 1.1759, "step": 6080, "task_loss": 0.8801817297935486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8573801517486572, "epoch": 5.14, "learning_rate": 3.65184721352536e-05, "loss": 1.3476, "step": 6081, "task_loss": 0.21912872791290283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3999049663543701, "epoch": 5.14, "learning_rate": 3.651534126487164e-05, "loss": 1.5657, "step": 6082, "task_loss": 0.7489321231842041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4704902172088623, "epoch": 5.14, "learning_rate": 3.651221039448967e-05, "loss": 1.4971, "step": 6083, "task_loss": 1.1352827548980713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3071929216384888, "epoch": 5.14, "learning_rate": 3.65090795241077e-05, "loss": 1.2924, "step": 6084, "task_loss": 0.7165845036506653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.633870244026184, "epoch": 5.14, "learning_rate": 3.6505948653725734e-05, "loss": 1.6418, "step": 6085, "task_loss": 1.5376369953155518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.1209726333618164, "epoch": 5.14, "learning_rate": 3.650281778334377e-05, "loss": 1.7705, "step": 6086, "task_loss": 1.0720608234405518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.118126392364502, "epoch": 5.15, "learning_rate": 3.6499686912961805e-05, "loss": 1.5132, "step": 6087, "task_loss": 1.0939449071884155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3653086423873901, "epoch": 5.15, "learning_rate": 3.6496556042579837e-05, "loss": 1.5177, "step": 6088, "task_loss": 0.5802086591720581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7642743587493896, "epoch": 5.15, "learning_rate": 3.649342517219787e-05, "loss": 1.1883, "step": 6089, "task_loss": 1.3531285524368286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0984039306640625, "epoch": 5.15, "learning_rate": 3.649029430181591e-05, "loss": 1.5588, "step": 6090, "task_loss": 1.7439521551132202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4954197406768799, "epoch": 5.15, "learning_rate": 3.648716343143394e-05, "loss": 1.2412, "step": 6091, "task_loss": 1.1622594594955444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.10504412651062, "epoch": 5.15, "learning_rate": 3.648403256105197e-05, "loss": 1.4554, "step": 6092, "task_loss": 1.384721279144287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4515087604522705, "epoch": 5.15, "learning_rate": 3.648090169067001e-05, "loss": 1.4268, "step": 6093, "task_loss": 1.4543139934539795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.143892765045166, "epoch": 5.15, "learning_rate": 3.647777082028804e-05, "loss": 1.2424, "step": 6094, "task_loss": 0.7920064330101013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5636687278747559, "epoch": 5.15, "learning_rate": 3.647463994990607e-05, "loss": 1.3081, "step": 6095, "task_loss": 1.1872886419296265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3452047109603882, "epoch": 5.15, "learning_rate": 3.6471509079524104e-05, "loss": 1.6726, "step": 6096, "task_loss": 1.7042359113693237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.365617275238037, "epoch": 5.15, "learning_rate": 3.646837820914214e-05, "loss": 1.2937, "step": 6097, "task_loss": 1.023611068725586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1313706636428833, "epoch": 5.15, "learning_rate": 3.6465247338760175e-05, "loss": 1.3105, "step": 6098, "task_loss": 1.8356813192367554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8160462379455566, "epoch": 5.16, "learning_rate": 3.646211646837821e-05, "loss": 1.4275, "step": 6099, "task_loss": 1.5652230978012085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.457406759262085, "epoch": 5.16, "learning_rate": 3.6458985597996245e-05, "loss": 1.6588, "step": 6100, "task_loss": 0.5557131171226501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6661267280578613, "epoch": 5.16, "learning_rate": 3.6455854727614284e-05, "loss": 1.5028, "step": 6101, "task_loss": 3.6507017612457275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6910134553909302, "epoch": 5.16, "learning_rate": 3.6452723857232315e-05, "loss": 1.1237, "step": 6102, "task_loss": 0.19734066724777222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.771165370941162, "epoch": 5.16, "learning_rate": 3.644959298685035e-05, "loss": 1.3202, "step": 6103, "task_loss": 0.7856285572052002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.580336570739746, "epoch": 5.16, "learning_rate": 3.644646211646838e-05, "loss": 1.2018, "step": 6104, "task_loss": 0.8775936365127563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1935255527496338, "epoch": 5.16, "learning_rate": 3.644333124608642e-05, "loss": 1.2751, "step": 6105, "task_loss": 0.5245146155357361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6143630743026733, "epoch": 5.16, "learning_rate": 3.644020037570445e-05, "loss": 1.1912, "step": 6106, "task_loss": 1.21666419506073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7414246201515198, "epoch": 5.16, "learning_rate": 3.643706950532248e-05, "loss": 1.3746, "step": 6107, "task_loss": 1.0092713832855225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7172352075576782, "epoch": 5.16, "learning_rate": 3.643393863494052e-05, "loss": 1.1246, "step": 6108, "task_loss": 0.6863077282905579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9939789772033691, "epoch": 5.16, "learning_rate": 3.643080776455855e-05, "loss": 1.1449, "step": 6109, "task_loss": 1.1528769731521606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0707781314849854, "epoch": 5.16, "learning_rate": 3.642767689417658e-05, "loss": 1.5155, "step": 6110, "task_loss": 1.2563642263412476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.156301498413086, "epoch": 5.17, "learning_rate": 3.6424546023794615e-05, "loss": 1.6419, "step": 6111, "task_loss": 1.617692470550537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.679142951965332, "epoch": 5.17, "learning_rate": 3.6421415153412653e-05, "loss": 1.4122, "step": 6112, "task_loss": 1.101264238357544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.34159517288208, "epoch": 5.17, "learning_rate": 3.6418284283030685e-05, "loss": 1.3815, "step": 6113, "task_loss": 1.0210744142532349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2611744403839111, "epoch": 5.17, "learning_rate": 3.641515341264872e-05, "loss": 1.549, "step": 6114, "task_loss": 0.9049245715141296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3253756761550903, "epoch": 5.17, "learning_rate": 3.641202254226675e-05, "loss": 1.5636, "step": 6115, "task_loss": 1.0267633199691772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0691472291946411, "epoch": 5.17, "learning_rate": 3.640889167188479e-05, "loss": 1.3828, "step": 6116, "task_loss": 0.6659682393074036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9444923400878906, "epoch": 5.17, "learning_rate": 3.640576080150282e-05, "loss": 1.1508, "step": 6117, "task_loss": 0.22833162546157837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9052455425262451, "epoch": 5.17, "learning_rate": 3.640262993112085e-05, "loss": 1.5262, "step": 6118, "task_loss": 1.7748912572860718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.3852038383483887, "epoch": 5.17, "learning_rate": 3.639949906073889e-05, "loss": 1.4357, "step": 6119, "task_loss": 1.3489131927490234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.828330397605896, "epoch": 5.17, "learning_rate": 3.639636819035692e-05, "loss": 1.1421, "step": 6120, "task_loss": 0.6403317451477051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.872029423713684, "epoch": 5.17, "learning_rate": 3.639323731997495e-05, "loss": 1.5739, "step": 6121, "task_loss": 1.9415736198425293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7881554961204529, "epoch": 5.17, "learning_rate": 3.6390106449592985e-05, "loss": 1.2446, "step": 6122, "task_loss": 0.6851750016212463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2687032222747803, "epoch": 5.18, "learning_rate": 3.638697557921102e-05, "loss": 1.3473, "step": 6123, "task_loss": 1.1219964027404785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7864205241203308, "epoch": 5.18, "learning_rate": 3.6383844708829055e-05, "loss": 1.1139, "step": 6124, "task_loss": 1.0692418813705444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.466882586479187, "epoch": 5.18, "learning_rate": 3.638071383844709e-05, "loss": 1.3676, "step": 6125, "task_loss": 0.9591256380081177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.3839774131774902, "epoch": 5.18, "learning_rate": 3.637758296806512e-05, "loss": 1.4479, "step": 6126, "task_loss": 1.3639037609100342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6754485368728638, "epoch": 5.18, "learning_rate": 3.637445209768316e-05, "loss": 1.5076, "step": 6127, "task_loss": 0.5004161596298218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7314077615737915, "epoch": 5.18, "learning_rate": 3.637132122730119e-05, "loss": 1.8356, "step": 6128, "task_loss": 0.8442422151565552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8581748008728027, "epoch": 5.18, "learning_rate": 3.636819035691922e-05, "loss": 1.3479, "step": 6129, "task_loss": 0.5146194100379944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2280867099761963, "epoch": 5.18, "learning_rate": 3.636505948653726e-05, "loss": 1.2499, "step": 6130, "task_loss": 1.0800772905349731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9497032761573792, "epoch": 5.18, "learning_rate": 3.636192861615529e-05, "loss": 1.3204, "step": 6131, "task_loss": 1.0427738428115845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2686192989349365, "epoch": 5.18, "learning_rate": 3.635879774577333e-05, "loss": 1.164, "step": 6132, "task_loss": 1.0744878053665161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0135504007339478, "epoch": 5.18, "learning_rate": 3.635566687539136e-05, "loss": 1.0919, "step": 6133, "task_loss": 0.5231415033340454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0223939418792725, "epoch": 5.19, "learning_rate": 3.635253600500939e-05, "loss": 1.3904, "step": 6134, "task_loss": 1.8189412355422974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2494484186172485, "epoch": 5.19, "learning_rate": 3.634940513462743e-05, "loss": 1.2841, "step": 6135, "task_loss": 0.5491918325424194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.098093032836914, "epoch": 5.19, "learning_rate": 3.6346274264245464e-05, "loss": 1.0822, "step": 6136, "task_loss": 1.1762385368347168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.858355164527893, "epoch": 5.19, "learning_rate": 3.6343143393863495e-05, "loss": 1.4037, "step": 6137, "task_loss": 2.118990898132324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7735649347305298, "epoch": 5.19, "learning_rate": 3.6340012523481534e-05, "loss": 1.3351, "step": 6138, "task_loss": 1.5209455490112305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.850955605506897, "epoch": 5.19, "learning_rate": 3.6336881653099566e-05, "loss": 1.4592, "step": 6139, "task_loss": 0.640375554561615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.386589765548706, "epoch": 5.19, "learning_rate": 3.63337507827176e-05, "loss": 1.2074, "step": 6140, "task_loss": 0.6136994361877441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7672364711761475, "epoch": 5.19, "learning_rate": 3.633061991233563e-05, "loss": 1.1906, "step": 6141, "task_loss": 1.8067741394042969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7080194354057312, "epoch": 5.19, "learning_rate": 3.632748904195367e-05, "loss": 1.5208, "step": 6142, "task_loss": 0.43782708048820496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0584958791732788, "epoch": 5.19, "learning_rate": 3.63243581715717e-05, "loss": 1.1251, "step": 6143, "task_loss": 0.6611096262931824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4944480657577515, "epoch": 5.19, "learning_rate": 3.632122730118973e-05, "loss": 1.2877, "step": 6144, "task_loss": 0.8358311653137207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7758561372756958, "epoch": 5.19, "learning_rate": 3.631809643080777e-05, "loss": 1.4819, "step": 6145, "task_loss": 1.4481488466262817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6393752098083496, "epoch": 5.2, "learning_rate": 3.63149655604258e-05, "loss": 1.2756, "step": 6146, "task_loss": 1.2702738046646118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3999230861663818, "epoch": 5.2, "learning_rate": 3.6311834690043834e-05, "loss": 1.1689, "step": 6147, "task_loss": 0.9565707445144653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8039963841438293, "epoch": 5.2, "learning_rate": 3.6308703819661865e-05, "loss": 1.1445, "step": 6148, "task_loss": 1.36081862449646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1678038835525513, "epoch": 5.2, "learning_rate": 3.6305572949279904e-05, "loss": 1.2504, "step": 6149, "task_loss": 0.4888961911201477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1501085758209229, "epoch": 5.2, "learning_rate": 3.6302442078897936e-05, "loss": 1.0488, "step": 6150, "task_loss": 1.2206778526306152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8758236169815063, "epoch": 5.2, "learning_rate": 3.629931120851597e-05, "loss": 1.1267, "step": 6151, "task_loss": 1.0384737253189087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9882737398147583, "epoch": 5.2, "learning_rate": 3.6296180338134e-05, "loss": 1.2901, "step": 6152, "task_loss": 0.36473548412323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7602756023406982, "epoch": 5.2, "learning_rate": 3.629304946775204e-05, "loss": 1.1889, "step": 6153, "task_loss": 2.3698301315307617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8622870445251465, "epoch": 5.2, "learning_rate": 3.628991859737007e-05, "loss": 1.4203, "step": 6154, "task_loss": 1.0741602182388306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9766383171081543, "epoch": 5.2, "learning_rate": 3.62867877269881e-05, "loss": 1.3054, "step": 6155, "task_loss": 1.9052064418792725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5276590585708618, "epoch": 5.2, "learning_rate": 3.628365685660614e-05, "loss": 1.3961, "step": 6156, "task_loss": 1.5221526622772217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8133926391601562, "epoch": 5.2, "learning_rate": 3.628052598622417e-05, "loss": 1.3844, "step": 6157, "task_loss": 1.5077011585235596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.326931118965149, "epoch": 5.21, "learning_rate": 3.6277395115842203e-05, "loss": 1.2565, "step": 6158, "task_loss": 1.3363896608352661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1007499694824219, "epoch": 5.21, "learning_rate": 3.6274264245460235e-05, "loss": 1.18, "step": 6159, "task_loss": 1.4532089233398438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0409033298492432, "epoch": 5.21, "learning_rate": 3.6271133375078274e-05, "loss": 1.2359, "step": 6160, "task_loss": 0.9478939771652222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5100560188293457, "epoch": 5.21, "learning_rate": 3.6268002504696306e-05, "loss": 1.8266, "step": 6161, "task_loss": 1.0317871570587158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.405659556388855, "epoch": 5.21, "learning_rate": 3.626487163431434e-05, "loss": 1.2703, "step": 6162, "task_loss": 1.5752323865890503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5741112232208252, "epoch": 5.21, "learning_rate": 3.626174076393237e-05, "loss": 1.3322, "step": 6163, "task_loss": 1.156861424446106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0359060764312744, "epoch": 5.21, "learning_rate": 3.625860989355041e-05, "loss": 1.019, "step": 6164, "task_loss": 1.1130077838897705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3286919593811035, "epoch": 5.21, "learning_rate": 3.625547902316844e-05, "loss": 1.3137, "step": 6165, "task_loss": 1.2317736148834229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1322071552276611, "epoch": 5.21, "learning_rate": 3.625234815278648e-05, "loss": 0.9877, "step": 6166, "task_loss": 0.9415155053138733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3308329582214355, "epoch": 5.21, "learning_rate": 3.624921728240451e-05, "loss": 1.7299, "step": 6167, "task_loss": 0.8722330331802368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5354074239730835, "epoch": 5.21, "learning_rate": 3.624608641202255e-05, "loss": 1.1808, "step": 6168, "task_loss": 1.3692717552185059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1281094551086426, "epoch": 5.21, "learning_rate": 3.624295554164058e-05, "loss": 1.3149, "step": 6169, "task_loss": 0.6713020205497742 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8326303958892822, "epoch": 5.22, "learning_rate": 3.623982467125861e-05, "loss": 1.0737, "step": 6170, "task_loss": 0.8451918959617615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3216361999511719, "epoch": 5.22, "learning_rate": 3.623669380087665e-05, "loss": 1.4845, "step": 6171, "task_loss": 1.2981711626052856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0450704097747803, "epoch": 5.22, "learning_rate": 3.623356293049468e-05, "loss": 1.1097, "step": 6172, "task_loss": 0.9734954237937927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2120904922485352, "epoch": 5.22, "learning_rate": 3.6230432060112714e-05, "loss": 1.4359, "step": 6173, "task_loss": 0.8063766360282898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8598896265029907, "epoch": 5.22, "learning_rate": 3.6227301189730746e-05, "loss": 1.2058, "step": 6174, "task_loss": 0.8407964706420898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6155974864959717, "epoch": 5.22, "learning_rate": 3.6224170319348784e-05, "loss": 1.1441, "step": 6175, "task_loss": 0.9151532649993896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.141890525817871, "epoch": 5.22, "learning_rate": 3.6221039448966816e-05, "loss": 1.2833, "step": 6176, "task_loss": 0.7445327043533325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0537046194076538, "epoch": 5.22, "learning_rate": 3.621790857858485e-05, "loss": 1.3738, "step": 6177, "task_loss": 1.4287636280059814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.106567621231079, "epoch": 5.22, "learning_rate": 3.621477770820288e-05, "loss": 0.8955, "step": 6178, "task_loss": 0.7679278254508972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0007920265197754, "epoch": 5.22, "learning_rate": 3.621164683782092e-05, "loss": 1.3328, "step": 6179, "task_loss": 1.1425368785858154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9006073474884033, "epoch": 5.22, "learning_rate": 3.620851596743895e-05, "loss": 0.9469, "step": 6180, "task_loss": 0.34007006883621216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3671388626098633, "epoch": 5.22, "learning_rate": 3.620538509705698e-05, "loss": 1.2467, "step": 6181, "task_loss": 1.2151801586151123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1344201564788818, "epoch": 5.23, "learning_rate": 3.620225422667502e-05, "loss": 1.1177, "step": 6182, "task_loss": 1.3448373079299927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3712456226348877, "epoch": 5.23, "learning_rate": 3.619912335629305e-05, "loss": 1.2516, "step": 6183, "task_loss": 2.25657320022583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.352975845336914, "epoch": 5.23, "learning_rate": 3.6195992485911084e-05, "loss": 1.396, "step": 6184, "task_loss": 0.9102721810340881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9172277450561523, "epoch": 5.23, "learning_rate": 3.6192861615529116e-05, "loss": 1.2921, "step": 6185, "task_loss": 1.6864967346191406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9732224941253662, "epoch": 5.23, "learning_rate": 3.6189730745147154e-05, "loss": 1.1938, "step": 6186, "task_loss": 0.8411593437194824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5920690298080444, "epoch": 5.23, "learning_rate": 3.6186599874765186e-05, "loss": 1.8995, "step": 6187, "task_loss": 0.9648462533950806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.611713171005249, "epoch": 5.23, "learning_rate": 3.618346900438322e-05, "loss": 1.2203, "step": 6188, "task_loss": 1.072453260421753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1223783493041992, "epoch": 5.23, "learning_rate": 3.618033813400125e-05, "loss": 1.2553, "step": 6189, "task_loss": 1.2973381280899048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5223734378814697, "epoch": 5.23, "learning_rate": 3.617720726361929e-05, "loss": 1.2773, "step": 6190, "task_loss": 0.9167593121528625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9306957125663757, "epoch": 5.23, "learning_rate": 3.617407639323732e-05, "loss": 1.2337, "step": 6191, "task_loss": 1.3778960704803467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1775541305541992, "epoch": 5.23, "learning_rate": 3.617094552285535e-05, "loss": 1.4347, "step": 6192, "task_loss": 0.3971376419067383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.549649953842163, "epoch": 5.23, "learning_rate": 3.616781465247339e-05, "loss": 1.1527, "step": 6193, "task_loss": 1.1758184432983398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2982876300811768, "epoch": 5.24, "learning_rate": 3.616468378209142e-05, "loss": 1.7341, "step": 6194, "task_loss": 0.9413337707519531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8556355237960815, "epoch": 5.24, "learning_rate": 3.6161552911709454e-05, "loss": 1.3709, "step": 6195, "task_loss": 1.6363489627838135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2900898456573486, "epoch": 5.24, "learning_rate": 3.6158422041327486e-05, "loss": 1.0099, "step": 6196, "task_loss": 1.6437559127807617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1785508394241333, "epoch": 5.24, "learning_rate": 3.6155291170945524e-05, "loss": 1.2567, "step": 6197, "task_loss": 0.9648005962371826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5606000423431396, "epoch": 5.24, "learning_rate": 3.6152160300563556e-05, "loss": 1.1303, "step": 6198, "task_loss": 0.8449077010154724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8706614375114441, "epoch": 5.24, "learning_rate": 3.6149029430181595e-05, "loss": 1.2058, "step": 6199, "task_loss": 0.5568203926086426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9356739521026611, "epoch": 5.24, "learning_rate": 3.6145898559799626e-05, "loss": 1.2413, "step": 6200, "task_loss": 0.9641901254653931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7556639313697815, "epoch": 5.24, "learning_rate": 3.614276768941766e-05, "loss": 1.6428, "step": 6201, "task_loss": 0.21536152064800262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2612335681915283, "epoch": 5.24, "learning_rate": 3.61396368190357e-05, "loss": 1.2452, "step": 6202, "task_loss": 1.8081003427505493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0268927812576294, "epoch": 5.24, "learning_rate": 3.613650594865373e-05, "loss": 1.0507, "step": 6203, "task_loss": 0.9136358499526978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7713882923126221, "epoch": 5.24, "learning_rate": 3.613337507827177e-05, "loss": 1.2237, "step": 6204, "task_loss": 0.505316436290741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.525716245174408, "epoch": 5.24, "learning_rate": 3.61302442078898e-05, "loss": 1.1287, "step": 6205, "task_loss": 0.12472329288721085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.245796799659729, "epoch": 5.25, "learning_rate": 3.612711333750783e-05, "loss": 1.2819, "step": 6206, "task_loss": 1.44695246219635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.390151023864746, "epoch": 5.25, "learning_rate": 3.612398246712586e-05, "loss": 1.5801, "step": 6207, "task_loss": 1.7847180366516113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9552649855613708, "epoch": 5.25, "learning_rate": 3.61208515967439e-05, "loss": 1.2043, "step": 6208, "task_loss": 1.207979440689087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.539603590965271, "epoch": 5.25, "learning_rate": 3.611772072636193e-05, "loss": 1.5071, "step": 6209, "task_loss": 0.7965636849403381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.191713809967041, "epoch": 5.25, "learning_rate": 3.6114589855979965e-05, "loss": 1.2003, "step": 6210, "task_loss": 1.4402698278427124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1984834671020508, "epoch": 5.25, "learning_rate": 3.6111458985597996e-05, "loss": 1.2898, "step": 6211, "task_loss": 0.6795680522918701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.4219377040863037, "epoch": 5.25, "learning_rate": 3.6108328115216035e-05, "loss": 1.5149, "step": 6212, "task_loss": 1.4195173978805542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2217317819595337, "epoch": 5.25, "learning_rate": 3.610519724483407e-05, "loss": 1.1041, "step": 6213, "task_loss": 0.5472096800804138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4428704977035522, "epoch": 5.25, "learning_rate": 3.61020663744521e-05, "loss": 1.0039, "step": 6214, "task_loss": 1.793933391571045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0193586349487305, "epoch": 5.25, "learning_rate": 3.609893550407013e-05, "loss": 1.4585, "step": 6215, "task_loss": 1.9177181720733643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0318212509155273, "epoch": 5.25, "learning_rate": 3.609580463368817e-05, "loss": 1.4637, "step": 6216, "task_loss": 1.4566283226013184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3394819498062134, "epoch": 5.26, "learning_rate": 3.60926737633062e-05, "loss": 1.1165, "step": 6217, "task_loss": 0.9192404747009277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9425513744354248, "epoch": 5.26, "learning_rate": 3.608954289292423e-05, "loss": 1.1225, "step": 6218, "task_loss": 0.5411958694458008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9517233371734619, "epoch": 5.26, "learning_rate": 3.608641202254227e-05, "loss": 1.0776, "step": 6219, "task_loss": 0.6669647097587585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2584975957870483, "epoch": 5.26, "learning_rate": 3.60832811521603e-05, "loss": 1.4045, "step": 6220, "task_loss": 1.6150532960891724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6709139347076416, "epoch": 5.26, "learning_rate": 3.6080150281778334e-05, "loss": 1.3562, "step": 6221, "task_loss": 0.5958110690116882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.613500952720642, "epoch": 5.26, "learning_rate": 3.6077019411396366e-05, "loss": 1.4133, "step": 6222, "task_loss": 0.8025984168052673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9763256311416626, "epoch": 5.26, "learning_rate": 3.6073888541014405e-05, "loss": 1.0701, "step": 6223, "task_loss": 0.4776271879673004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9193191528320312, "epoch": 5.26, "learning_rate": 3.6070757670632437e-05, "loss": 1.4081, "step": 6224, "task_loss": 0.6744579672813416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3776895999908447, "epoch": 5.26, "learning_rate": 3.606762680025047e-05, "loss": 1.1994, "step": 6225, "task_loss": 1.5977023839950562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2887613773345947, "epoch": 5.26, "learning_rate": 3.60644959298685e-05, "loss": 1.0312, "step": 6226, "task_loss": 0.4873543381690979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7067044377326965, "epoch": 5.26, "learning_rate": 3.606136505948654e-05, "loss": 1.1992, "step": 6227, "task_loss": 0.5076286792755127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.779624581336975, "epoch": 5.26, "learning_rate": 3.605823418910457e-05, "loss": 1.2636, "step": 6228, "task_loss": 2.627349853515625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.945124626159668, "epoch": 5.27, "learning_rate": 3.60551033187226e-05, "loss": 1.0519, "step": 6229, "task_loss": 0.29511189460754395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9006686210632324, "epoch": 5.27, "learning_rate": 3.605197244834064e-05, "loss": 1.7011, "step": 6230, "task_loss": 1.250108003616333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3282780647277832, "epoch": 5.27, "learning_rate": 3.604884157795867e-05, "loss": 1.2749, "step": 6231, "task_loss": 1.2262325286865234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49828970432281494, "epoch": 5.27, "learning_rate": 3.6045710707576704e-05, "loss": 1.2552, "step": 6232, "task_loss": 0.34046992659568787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9939165711402893, "epoch": 5.27, "learning_rate": 3.604257983719474e-05, "loss": 1.0476, "step": 6233, "task_loss": 1.9087615013122559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1989895105361938, "epoch": 5.27, "learning_rate": 3.6039448966812775e-05, "loss": 1.3507, "step": 6234, "task_loss": 1.420727014541626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0952825546264648, "epoch": 5.27, "learning_rate": 3.603631809643081e-05, "loss": 1.2517, "step": 6235, "task_loss": 0.924236536026001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3956170082092285, "epoch": 5.27, "learning_rate": 3.6033187226048845e-05, "loss": 1.2136, "step": 6236, "task_loss": 0.8001537322998047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3870691061019897, "epoch": 5.27, "learning_rate": 3.603005635566688e-05, "loss": 1.2432, "step": 6237, "task_loss": 0.9522228240966797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.262291431427002, "epoch": 5.27, "learning_rate": 3.6026925485284915e-05, "loss": 0.9847, "step": 6238, "task_loss": 1.0076707601547241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2623631954193115, "epoch": 5.27, "learning_rate": 3.602379461490295e-05, "loss": 0.8977, "step": 6239, "task_loss": 0.833356499671936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4916409254074097, "epoch": 5.27, "learning_rate": 3.602066374452098e-05, "loss": 1.0619, "step": 6240, "task_loss": 1.7494628429412842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1034795045852661, "epoch": 5.28, "learning_rate": 3.601753287413902e-05, "loss": 1.0261, "step": 6241, "task_loss": 2.4069488048553467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5387364625930786, "epoch": 5.28, "learning_rate": 3.601440200375705e-05, "loss": 1.2394, "step": 6242, "task_loss": 1.878438949584961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4658160209655762, "epoch": 5.28, "learning_rate": 3.601127113337508e-05, "loss": 1.124, "step": 6243, "task_loss": 2.680520534515381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7502906322479248, "epoch": 5.28, "learning_rate": 3.600814026299311e-05, "loss": 1.19, "step": 6244, "task_loss": 0.8298736810684204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9995706081390381, "epoch": 5.28, "learning_rate": 3.600500939261115e-05, "loss": 1.1607, "step": 6245, "task_loss": 1.811976671218872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0733139514923096, "epoch": 5.28, "learning_rate": 3.600187852222918e-05, "loss": 1.2795, "step": 6246, "task_loss": 1.2633274793624878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.919567883014679, "epoch": 5.28, "learning_rate": 3.5998747651847215e-05, "loss": 0.9386, "step": 6247, "task_loss": 0.7086263298988342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6525590419769287, "epoch": 5.28, "learning_rate": 3.599561678146525e-05, "loss": 1.0796, "step": 6248, "task_loss": 0.8634302020072937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.03059983253479, "epoch": 5.28, "learning_rate": 3.5992485911083285e-05, "loss": 0.9951, "step": 6249, "task_loss": 1.7619150876998901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5451719760894775, "epoch": 5.28, "learning_rate": 3.598935504070132e-05, "loss": 1.2567, "step": 6250, "task_loss": 1.5727558135986328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.131095051765442, "epoch": 5.28, "learning_rate": 3.598622417031935e-05, "loss": 1.3278, "step": 6251, "task_loss": 1.2390565872192383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.691896915435791, "epoch": 5.28, "learning_rate": 3.598309329993739e-05, "loss": 1.5055, "step": 6252, "task_loss": 1.719545602798462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.983845591545105, "epoch": 5.29, "learning_rate": 3.597996242955542e-05, "loss": 0.8996, "step": 6253, "task_loss": 0.5710669755935669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6599743962287903, "epoch": 5.29, "learning_rate": 3.597683155917345e-05, "loss": 1.0419, "step": 6254, "task_loss": 0.12073215842247009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6826252937316895, "epoch": 5.29, "learning_rate": 3.597370068879148e-05, "loss": 1.3066, "step": 6255, "task_loss": 1.2276685237884521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1574227809906006, "epoch": 5.29, "learning_rate": 3.597056981840952e-05, "loss": 1.1045, "step": 6256, "task_loss": 1.1261646747589111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0992143154144287, "epoch": 5.29, "learning_rate": 3.596743894802755e-05, "loss": 1.1046, "step": 6257, "task_loss": 0.516495943069458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.451198935508728, "epoch": 5.29, "learning_rate": 3.5964308077645585e-05, "loss": 1.5348, "step": 6258, "task_loss": 1.7652535438537598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.01911461353302, "epoch": 5.29, "learning_rate": 3.596117720726362e-05, "loss": 1.6055, "step": 6259, "task_loss": 0.7690885066986084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.043402075767517, "epoch": 5.29, "learning_rate": 3.5958046336881655e-05, "loss": 1.265, "step": 6260, "task_loss": 1.1561319828033447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.389328956604004, "epoch": 5.29, "learning_rate": 3.595491546649969e-05, "loss": 1.1467, "step": 6261, "task_loss": 1.0841835737228394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6979016065597534, "epoch": 5.29, "learning_rate": 3.595178459611772e-05, "loss": 1.1823, "step": 6262, "task_loss": 1.0601513385772705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0005974769592285, "epoch": 5.29, "learning_rate": 3.594865372573575e-05, "loss": 1.1059, "step": 6263, "task_loss": 0.27517393231391907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1660971641540527, "epoch": 5.29, "learning_rate": 3.594552285535379e-05, "loss": 1.0664, "step": 6264, "task_loss": 1.0236659049987793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8910622596740723, "epoch": 5.3, "learning_rate": 3.594239198497182e-05, "loss": 1.2196, "step": 6265, "task_loss": 0.7757294774055481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9431942105293274, "epoch": 5.3, "learning_rate": 3.593926111458986e-05, "loss": 1.1225, "step": 6266, "task_loss": 0.9701443314552307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7034191489219666, "epoch": 5.3, "learning_rate": 3.593613024420789e-05, "loss": 1.1329, "step": 6267, "task_loss": 0.5820156335830688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.607806921005249, "epoch": 5.3, "learning_rate": 3.593299937382592e-05, "loss": 1.1514, "step": 6268, "task_loss": 0.9594535827636719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8514241576194763, "epoch": 5.3, "learning_rate": 3.592986850344396e-05, "loss": 1.1042, "step": 6269, "task_loss": 0.8006724119186401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5595362186431885, "epoch": 5.3, "learning_rate": 3.592673763306199e-05, "loss": 1.4088, "step": 6270, "task_loss": 0.5105048418045044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7528355121612549, "epoch": 5.3, "learning_rate": 3.592360676268003e-05, "loss": 1.1431, "step": 6271, "task_loss": 0.5234667658805847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6754131317138672, "epoch": 5.3, "learning_rate": 3.5920475892298064e-05, "loss": 1.299, "step": 6272, "task_loss": 1.3239189386367798 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8233328461647034, "epoch": 5.3, "learning_rate": 3.5917345021916096e-05, "loss": 1.1261, "step": 6273, "task_loss": 1.0144625902175903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0327129364013672, "epoch": 5.3, "learning_rate": 3.591421415153413e-05, "loss": 0.9113, "step": 6274, "task_loss": 1.0259311199188232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3241872787475586, "epoch": 5.3, "learning_rate": 3.5911083281152166e-05, "loss": 1.2151, "step": 6275, "task_loss": 0.34086257219314575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6495931148529053, "epoch": 5.3, "learning_rate": 3.59079524107702e-05, "loss": 1.1068, "step": 6276, "task_loss": 1.1500986814498901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.516337275505066, "epoch": 5.31, "learning_rate": 3.590482154038823e-05, "loss": 1.1458, "step": 6277, "task_loss": 1.2977194786071777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9136282205581665, "epoch": 5.31, "learning_rate": 3.590169067000627e-05, "loss": 1.1392, "step": 6278, "task_loss": 0.5148177742958069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8950405120849609, "epoch": 5.31, "learning_rate": 3.58985597996243e-05, "loss": 1.189, "step": 6279, "task_loss": 0.5856208801269531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3142120838165283, "epoch": 5.31, "learning_rate": 3.589542892924233e-05, "loss": 1.2594, "step": 6280, "task_loss": 1.6574043035507202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1934990882873535, "epoch": 5.31, "learning_rate": 3.589229805886036e-05, "loss": 1.065, "step": 6281, "task_loss": 1.315049648284912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4491824209690094, "epoch": 5.31, "learning_rate": 3.58891671884784e-05, "loss": 1.1674, "step": 6282, "task_loss": 0.3966837227344513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1138911247253418, "epoch": 5.31, "learning_rate": 3.5886036318096434e-05, "loss": 1.1012, "step": 6283, "task_loss": 0.7716041803359985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8432117104530334, "epoch": 5.31, "learning_rate": 3.5882905447714465e-05, "loss": 1.2124, "step": 6284, "task_loss": 1.0378227233886719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9759465456008911, "epoch": 5.31, "learning_rate": 3.58797745773325e-05, "loss": 1.1672, "step": 6285, "task_loss": 2.0205843448638916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8995105624198914, "epoch": 5.31, "learning_rate": 3.5876643706950536e-05, "loss": 1.2336, "step": 6286, "task_loss": 0.7382641434669495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3993706703186035, "epoch": 5.31, "learning_rate": 3.587351283656857e-05, "loss": 1.2666, "step": 6287, "task_loss": 1.065693974494934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6898131370544434, "epoch": 5.32, "learning_rate": 3.58703819661866e-05, "loss": 1.4274, "step": 6288, "task_loss": 1.394351601600647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7496863603591919, "epoch": 5.32, "learning_rate": 3.586725109580464e-05, "loss": 1.0168, "step": 6289, "task_loss": 0.7797940373420715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1748729944229126, "epoch": 5.32, "learning_rate": 3.586412022542267e-05, "loss": 1.5156, "step": 6290, "task_loss": 1.2034296989440918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4179853200912476, "epoch": 5.32, "learning_rate": 3.58609893550407e-05, "loss": 1.253, "step": 6291, "task_loss": 1.2203385829925537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1170263290405273, "epoch": 5.32, "learning_rate": 3.585785848465873e-05, "loss": 1.0557, "step": 6292, "task_loss": 1.8319255113601685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0441277027130127, "epoch": 5.32, "learning_rate": 3.585472761427677e-05, "loss": 0.9215, "step": 6293, "task_loss": 1.3326600790023804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0859233140945435, "epoch": 5.32, "learning_rate": 3.5851596743894804e-05, "loss": 1.4722, "step": 6294, "task_loss": 0.40048739314079285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3667182922363281, "epoch": 5.32, "learning_rate": 3.5848465873512835e-05, "loss": 1.2951, "step": 6295, "task_loss": 0.8666419386863708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2466959953308105, "epoch": 5.32, "learning_rate": 3.584533500313087e-05, "loss": 1.3288, "step": 6296, "task_loss": 2.1017820835113525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1239523887634277, "epoch": 5.32, "learning_rate": 3.5842204132748906e-05, "loss": 1.3147, "step": 6297, "task_loss": 1.0198177099227905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2820825576782227, "epoch": 5.32, "learning_rate": 3.583907326236694e-05, "loss": 1.5593, "step": 6298, "task_loss": 0.68439781665802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.302155017852783, "epoch": 5.32, "learning_rate": 3.583594239198497e-05, "loss": 1.6169, "step": 6299, "task_loss": 1.9808474779129028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.569717288017273, "epoch": 5.33, "learning_rate": 3.583281152160301e-05, "loss": 1.192, "step": 6300, "task_loss": 1.3193671703338623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0674583911895752, "epoch": 5.33, "learning_rate": 3.582968065122104e-05, "loss": 1.1247, "step": 6301, "task_loss": 2.083862543106079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1681983470916748, "epoch": 5.33, "learning_rate": 3.582654978083908e-05, "loss": 1.3053, "step": 6302, "task_loss": 0.9377240538597107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7429990768432617, "epoch": 5.33, "learning_rate": 3.582341891045711e-05, "loss": 1.0943, "step": 6303, "task_loss": 1.5056195259094238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2799397706985474, "epoch": 5.33, "learning_rate": 3.582028804007514e-05, "loss": 1.2975, "step": 6304, "task_loss": 1.3007200956344604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4593722820281982, "epoch": 5.33, "learning_rate": 3.581715716969318e-05, "loss": 1.3099, "step": 6305, "task_loss": 2.1906564235687256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4877761602401733, "epoch": 5.33, "learning_rate": 3.581402629931121e-05, "loss": 1.1171, "step": 6306, "task_loss": 1.3435848951339722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9973902702331543, "epoch": 5.33, "learning_rate": 3.5810895428929244e-05, "loss": 1.2002, "step": 6307, "task_loss": 0.5877852439880371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.983462929725647, "epoch": 5.33, "learning_rate": 3.580776455854728e-05, "loss": 1.1606, "step": 6308, "task_loss": 0.90736985206604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5594786405563354, "epoch": 5.33, "learning_rate": 3.5804633688165314e-05, "loss": 1.0041, "step": 6309, "task_loss": 1.230464220046997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.544875979423523, "epoch": 5.33, "learning_rate": 3.5801502817783346e-05, "loss": 1.1159, "step": 6310, "task_loss": 0.07882226258516312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7256287336349487, "epoch": 5.33, "learning_rate": 3.579837194740138e-05, "loss": 1.4004, "step": 6311, "task_loss": 0.8579704761505127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1282432079315186, "epoch": 5.34, "learning_rate": 3.5795241077019416e-05, "loss": 0.8974, "step": 6312, "task_loss": 0.6022537350654602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9551079273223877, "epoch": 5.34, "learning_rate": 3.579211020663745e-05, "loss": 1.2989, "step": 6313, "task_loss": 1.8056633472442627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4981790781021118, "epoch": 5.34, "learning_rate": 3.578897933625548e-05, "loss": 0.9995, "step": 6314, "task_loss": 0.7812796831130981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0608830451965332, "epoch": 5.34, "learning_rate": 3.578584846587352e-05, "loss": 1.0272, "step": 6315, "task_loss": 1.5571300983428955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.223242163658142, "epoch": 5.34, "learning_rate": 3.578271759549155e-05, "loss": 0.8772, "step": 6316, "task_loss": 0.4586220979690552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.6829257011413574, "epoch": 5.34, "learning_rate": 3.577958672510958e-05, "loss": 1.6339, "step": 6317, "task_loss": 2.0465681552886963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6966932415962219, "epoch": 5.34, "learning_rate": 3.5776455854727614e-05, "loss": 1.149, "step": 6318, "task_loss": 1.202324390411377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2541537284851074, "epoch": 5.34, "learning_rate": 3.577332498434565e-05, "loss": 1.2464, "step": 6319, "task_loss": 1.6774958372116089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7508423328399658, "epoch": 5.34, "learning_rate": 3.5770194113963684e-05, "loss": 0.7625, "step": 6320, "task_loss": 1.2577707767486572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4945155382156372, "epoch": 5.34, "learning_rate": 3.5767063243581716e-05, "loss": 1.1809, "step": 6321, "task_loss": 1.0388693809509277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0273234844207764, "epoch": 5.34, "learning_rate": 3.576393237319975e-05, "loss": 1.0099, "step": 6322, "task_loss": 0.9408475756645203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6688100099563599, "epoch": 5.34, "learning_rate": 3.5760801502817786e-05, "loss": 1.2255, "step": 6323, "task_loss": 1.1637934446334839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7642556428909302, "epoch": 5.35, "learning_rate": 3.575767063243582e-05, "loss": 0.9268, "step": 6324, "task_loss": 0.742929220199585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7998132705688477, "epoch": 5.35, "learning_rate": 3.575453976205385e-05, "loss": 1.1431, "step": 6325, "task_loss": 1.5905362367630005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0970619916915894, "epoch": 5.35, "learning_rate": 3.575140889167189e-05, "loss": 0.9937, "step": 6326, "task_loss": 0.6389658451080322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.500705361366272, "epoch": 5.35, "learning_rate": 3.574827802128992e-05, "loss": 1.1243, "step": 6327, "task_loss": 1.5148837566375732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9230647087097168, "epoch": 5.35, "learning_rate": 3.574514715090795e-05, "loss": 1.4079, "step": 6328, "task_loss": 1.416494607925415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2466554641723633, "epoch": 5.35, "learning_rate": 3.5742016280525984e-05, "loss": 1.2219, "step": 6329, "task_loss": 1.6794705390930176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2301051616668701, "epoch": 5.35, "learning_rate": 3.573888541014402e-05, "loss": 1.1792, "step": 6330, "task_loss": 1.4451097249984741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2690112590789795, "epoch": 5.35, "learning_rate": 3.5735754539762054e-05, "loss": 1.3094, "step": 6331, "task_loss": 1.3645142316818237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.07229483127594, "epoch": 5.35, "learning_rate": 3.5732623669380086e-05, "loss": 0.9248, "step": 6332, "task_loss": 1.2890040874481201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.034593939781189, "epoch": 5.35, "learning_rate": 3.5729492798998124e-05, "loss": 0.855, "step": 6333, "task_loss": 0.7116720676422119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6890979409217834, "epoch": 5.35, "learning_rate": 3.5726361928616156e-05, "loss": 0.9103, "step": 6334, "task_loss": 0.43140602111816406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5750360488891602, "epoch": 5.35, "learning_rate": 3.572323105823419e-05, "loss": 1.098, "step": 6335, "task_loss": 1.0778305530548096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3475818634033203, "epoch": 5.36, "learning_rate": 3.5720100187852226e-05, "loss": 1.1821, "step": 6336, "task_loss": 0.974353015422821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.771288275718689, "epoch": 5.36, "learning_rate": 3.571696931747026e-05, "loss": 1.3666, "step": 6337, "task_loss": 1.535383939743042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7002061605453491, "epoch": 5.36, "learning_rate": 3.57138384470883e-05, "loss": 1.1179, "step": 6338, "task_loss": 0.5249250531196594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1172676086425781, "epoch": 5.36, "learning_rate": 3.571070757670633e-05, "loss": 1.0367, "step": 6339, "task_loss": 1.5754286050796509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2037923336029053, "epoch": 5.36, "learning_rate": 3.570757670632436e-05, "loss": 1.0825, "step": 6340, "task_loss": 0.6013635993003845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9423974752426147, "epoch": 5.36, "learning_rate": 3.57044458359424e-05, "loss": 0.8051, "step": 6341, "task_loss": 1.152550458908081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9960110187530518, "epoch": 5.36, "learning_rate": 3.570131496556043e-05, "loss": 1.1765, "step": 6342, "task_loss": 0.6412350535392761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7078251838684082, "epoch": 5.36, "learning_rate": 3.569818409517846e-05, "loss": 1.29, "step": 6343, "task_loss": 0.5662834048271179 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6569747924804688, "epoch": 5.36, "learning_rate": 3.5695053224796494e-05, "loss": 1.2253, "step": 6344, "task_loss": 1.451343297958374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0257487297058105, "epoch": 5.36, "learning_rate": 3.569192235441453e-05, "loss": 1.2365, "step": 6345, "task_loss": 0.5275859236717224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7797366380691528, "epoch": 5.36, "learning_rate": 3.5688791484032565e-05, "loss": 0.8036, "step": 6346, "task_loss": 0.8363023400306702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3881083726882935, "epoch": 5.36, "learning_rate": 3.5685660613650596e-05, "loss": 1.3853, "step": 6347, "task_loss": 1.6423602104187012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.246205449104309, "epoch": 5.37, "learning_rate": 3.568252974326863e-05, "loss": 1.0554, "step": 6348, "task_loss": 1.3366527557373047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9609673619270325, "epoch": 5.37, "learning_rate": 3.567939887288667e-05, "loss": 1.0369, "step": 6349, "task_loss": 1.138465166091919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.052157163619995, "epoch": 5.37, "learning_rate": 3.56762680025047e-05, "loss": 1.7519, "step": 6350, "task_loss": 1.9160293340682983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9949177503585815, "epoch": 5.37, "learning_rate": 3.567313713212273e-05, "loss": 0.9273, "step": 6351, "task_loss": 0.9635921716690063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.503780484199524, "epoch": 5.37, "learning_rate": 3.567000626174077e-05, "loss": 1.0295, "step": 6352, "task_loss": 1.3605378866195679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3764536380767822, "epoch": 5.37, "learning_rate": 3.56668753913588e-05, "loss": 0.8984, "step": 6353, "task_loss": 1.7645039558410645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8513151407241821, "epoch": 5.37, "learning_rate": 3.566374452097683e-05, "loss": 1.1512, "step": 6354, "task_loss": 0.9969727993011475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4360969066619873, "epoch": 5.37, "learning_rate": 3.5660613650594864e-05, "loss": 1.5363, "step": 6355, "task_loss": 1.790378451347351 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.181695818901062, "epoch": 5.37, "learning_rate": 3.56574827802129e-05, "loss": 1.0548, "step": 6356, "task_loss": 0.8222810626029968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.156247615814209, "epoch": 5.37, "learning_rate": 3.5654351909830935e-05, "loss": 1.2353, "step": 6357, "task_loss": 0.6886773705482483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2429468631744385, "epoch": 5.37, "learning_rate": 3.5651221039448966e-05, "loss": 1.2727, "step": 6358, "task_loss": 1.0146561861038208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4982025623321533, "epoch": 5.38, "learning_rate": 3.5648090169067e-05, "loss": 1.305, "step": 6359, "task_loss": 1.711761474609375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0249230861663818, "epoch": 5.38, "learning_rate": 3.564495929868504e-05, "loss": 1.1283, "step": 6360, "task_loss": 2.098328113555908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5623283386230469, "epoch": 5.38, "learning_rate": 3.564182842830307e-05, "loss": 1.2086, "step": 6361, "task_loss": 1.0880765914916992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.193929672241211, "epoch": 5.38, "learning_rate": 3.56386975579211e-05, "loss": 1.1459, "step": 6362, "task_loss": 1.1182104349136353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8567476868629456, "epoch": 5.38, "learning_rate": 3.563556668753914e-05, "loss": 1.0078, "step": 6363, "task_loss": 0.784233808517456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9374117851257324, "epoch": 5.38, "learning_rate": 3.563243581715717e-05, "loss": 1.3421, "step": 6364, "task_loss": 0.764230489730835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6206719875335693, "epoch": 5.38, "learning_rate": 3.56293049467752e-05, "loss": 0.9732, "step": 6365, "task_loss": 0.462668776512146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2882137298583984, "epoch": 5.38, "learning_rate": 3.5626174076393234e-05, "loss": 1.1524, "step": 6366, "task_loss": 1.1057997941970825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.961618959903717, "epoch": 5.38, "learning_rate": 3.562304320601127e-05, "loss": 1.3049, "step": 6367, "task_loss": 1.2743057012557983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8405143022537231, "epoch": 5.38, "learning_rate": 3.5619912335629304e-05, "loss": 1.2529, "step": 6368, "task_loss": 0.2548726499080658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0309228897094727, "epoch": 5.38, "learning_rate": 3.561678146524734e-05, "loss": 1.1484, "step": 6369, "task_loss": 0.8863463401794434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7609108686447144, "epoch": 5.38, "learning_rate": 3.5613650594865375e-05, "loss": 1.0004, "step": 6370, "task_loss": 0.4051814377307892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6196961402893066, "epoch": 5.39, "learning_rate": 3.5610519724483407e-05, "loss": 1.3514, "step": 6371, "task_loss": 1.7118240594863892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6942396759986877, "epoch": 5.39, "learning_rate": 3.5607388854101445e-05, "loss": 0.9908, "step": 6372, "task_loss": 0.45530927181243896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6215879917144775, "epoch": 5.39, "learning_rate": 3.560425798371948e-05, "loss": 1.1249, "step": 6373, "task_loss": 0.7154276371002197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3944449424743652, "epoch": 5.39, "learning_rate": 3.560112711333751e-05, "loss": 1.575, "step": 6374, "task_loss": 1.3891637325286865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9033657312393188, "epoch": 5.39, "learning_rate": 3.559799624295555e-05, "loss": 1.0874, "step": 6375, "task_loss": 0.4665800631046295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0801172256469727, "epoch": 5.39, "learning_rate": 3.559486537257358e-05, "loss": 1.0858, "step": 6376, "task_loss": 0.9120462536811829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.862079381942749, "epoch": 5.39, "learning_rate": 3.559173450219161e-05, "loss": 1.0358, "step": 6377, "task_loss": 0.6100090742111206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7412952780723572, "epoch": 5.39, "learning_rate": 3.558860363180965e-05, "loss": 1.0374, "step": 6378, "task_loss": 0.7172748446464539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7286319732666016, "epoch": 5.39, "learning_rate": 3.558547276142768e-05, "loss": 1.3149, "step": 6379, "task_loss": 1.7833409309387207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2088083028793335, "epoch": 5.39, "learning_rate": 3.558234189104571e-05, "loss": 1.2655, "step": 6380, "task_loss": 1.489200234413147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2064722776412964, "epoch": 5.39, "learning_rate": 3.5579211020663745e-05, "loss": 1.0708, "step": 6381, "task_loss": 1.0547040700912476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.455769658088684, "epoch": 5.39, "learning_rate": 3.557608015028178e-05, "loss": 1.2345, "step": 6382, "task_loss": 0.7194487452507019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3606934547424316, "epoch": 5.4, "learning_rate": 3.5572949279899815e-05, "loss": 1.0516, "step": 6383, "task_loss": 1.1281007528305054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0855116844177246, "epoch": 5.4, "learning_rate": 3.556981840951785e-05, "loss": 1.2387, "step": 6384, "task_loss": 1.7949036359786987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8361400961875916, "epoch": 5.4, "learning_rate": 3.556668753913588e-05, "loss": 1.0641, "step": 6385, "task_loss": 1.1838756799697876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0090287923812866, "epoch": 5.4, "learning_rate": 3.556355666875392e-05, "loss": 1.1008, "step": 6386, "task_loss": 1.0565695762634277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2167842388153076, "epoch": 5.4, "learning_rate": 3.556042579837195e-05, "loss": 1.4077, "step": 6387, "task_loss": 0.8098171949386597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2641569375991821, "epoch": 5.4, "learning_rate": 3.555729492798998e-05, "loss": 1.048, "step": 6388, "task_loss": 0.684532105922699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48422402143478394, "epoch": 5.4, "learning_rate": 3.555416405760802e-05, "loss": 0.8169, "step": 6389, "task_loss": 0.03458156809210777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0898631811141968, "epoch": 5.4, "learning_rate": 3.555103318722605e-05, "loss": 1.249, "step": 6390, "task_loss": 1.5383977890014648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5663740634918213, "epoch": 5.4, "learning_rate": 3.554790231684408e-05, "loss": 1.1384, "step": 6391, "task_loss": 1.6336698532104492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3373264074325562, "epoch": 5.4, "learning_rate": 3.5544771446462115e-05, "loss": 1.1336, "step": 6392, "task_loss": 1.1092519760131836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9765065908432007, "epoch": 5.4, "learning_rate": 3.554164057608015e-05, "loss": 1.2489, "step": 6393, "task_loss": 0.5267802476882935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3448505401611328, "epoch": 5.4, "learning_rate": 3.5538509705698185e-05, "loss": 1.4271, "step": 6394, "task_loss": 0.9652625918388367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9564717411994934, "epoch": 5.41, "learning_rate": 3.553537883531622e-05, "loss": 0.9623, "step": 6395, "task_loss": 0.8883202075958252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.029958724975586, "epoch": 5.41, "learning_rate": 3.553224796493425e-05, "loss": 1.0507, "step": 6396, "task_loss": 0.8890610933303833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.024905800819397, "epoch": 5.41, "learning_rate": 3.552911709455229e-05, "loss": 1.2492, "step": 6397, "task_loss": 1.311061978340149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2415674924850464, "epoch": 5.41, "learning_rate": 3.552598622417032e-05, "loss": 1.2157, "step": 6398, "task_loss": 0.5983930230140686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4239780902862549, "epoch": 5.41, "learning_rate": 3.552285535378835e-05, "loss": 1.1966, "step": 6399, "task_loss": 1.8789446353912354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.14497709274292, "epoch": 5.41, "learning_rate": 3.551972448340639e-05, "loss": 1.1273, "step": 6400, "task_loss": 1.1997102499008179 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6814292073249817, "epoch": 5.41, "learning_rate": 3.551659361302442e-05, "loss": 0.8111, "step": 6401, "task_loss": 1.1293572187423706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3477864265441895, "epoch": 5.41, "learning_rate": 3.551346274264245e-05, "loss": 1.2736, "step": 6402, "task_loss": 1.0622754096984863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0369722843170166, "epoch": 5.41, "learning_rate": 3.551033187226049e-05, "loss": 1.0143, "step": 6403, "task_loss": 1.410489559173584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9915076494216919, "epoch": 5.41, "learning_rate": 3.550720100187852e-05, "loss": 1.1132, "step": 6404, "task_loss": 0.8849968314170837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6930464506149292, "epoch": 5.41, "learning_rate": 3.550407013149656e-05, "loss": 0.9041, "step": 6405, "task_loss": 0.8408554792404175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.805408239364624, "epoch": 5.41, "learning_rate": 3.5500939261114593e-05, "loss": 0.9924, "step": 6406, "task_loss": 1.2217357158660889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.043248176574707, "epoch": 5.42, "learning_rate": 3.5497808390732625e-05, "loss": 0.8891, "step": 6407, "task_loss": 0.760205090045929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6690831184387207, "epoch": 5.42, "learning_rate": 3.5494677520350664e-05, "loss": 0.6997, "step": 6408, "task_loss": 0.6362285614013672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.378875494003296, "epoch": 5.42, "learning_rate": 3.5491546649968696e-05, "loss": 1.1929, "step": 6409, "task_loss": 0.9241880774497986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0608899593353271, "epoch": 5.42, "learning_rate": 3.548841577958673e-05, "loss": 0.9236, "step": 6410, "task_loss": 1.4076604843139648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.894716739654541, "epoch": 5.42, "learning_rate": 3.548528490920476e-05, "loss": 1.1169, "step": 6411, "task_loss": 2.19657826423645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7610044479370117, "epoch": 5.42, "learning_rate": 3.54821540388228e-05, "loss": 1.5881, "step": 6412, "task_loss": 1.3982621431350708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8727271556854248, "epoch": 5.42, "learning_rate": 3.547902316844083e-05, "loss": 0.7673, "step": 6413, "task_loss": 1.2706485986709595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0862258672714233, "epoch": 5.42, "learning_rate": 3.547589229805886e-05, "loss": 1.173, "step": 6414, "task_loss": 1.9012936353683472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0984699726104736, "epoch": 5.42, "learning_rate": 3.54727614276769e-05, "loss": 1.0339, "step": 6415, "task_loss": 0.5531238317489624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9166451692581177, "epoch": 5.42, "learning_rate": 3.546963055729493e-05, "loss": 0.9783, "step": 6416, "task_loss": 1.6301275491714478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4365322589874268, "epoch": 5.42, "learning_rate": 3.546649968691296e-05, "loss": 1.368, "step": 6417, "task_loss": 2.148007392883301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2964650392532349, "epoch": 5.42, "learning_rate": 3.5463368816530995e-05, "loss": 1.0093, "step": 6418, "task_loss": 0.8783218264579773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2808705568313599, "epoch": 5.43, "learning_rate": 3.5460237946149034e-05, "loss": 1.1724, "step": 6419, "task_loss": 1.403984785079956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0268559455871582, "epoch": 5.43, "learning_rate": 3.5457107075767065e-05, "loss": 1.1939, "step": 6420, "task_loss": 0.5887287855148315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6717723608016968, "epoch": 5.43, "learning_rate": 3.54539762053851e-05, "loss": 0.7753, "step": 6421, "task_loss": 0.8569459915161133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7973546385765076, "epoch": 5.43, "learning_rate": 3.545084533500313e-05, "loss": 1.1117, "step": 6422, "task_loss": 0.4460127055644989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9489736557006836, "epoch": 5.43, "learning_rate": 3.544771446462117e-05, "loss": 1.0671, "step": 6423, "task_loss": 0.7028795480728149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5548537969589233, "epoch": 5.43, "learning_rate": 3.54445835942392e-05, "loss": 1.025, "step": 6424, "task_loss": 0.37687283754348755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7320155501365662, "epoch": 5.43, "learning_rate": 3.544145272385723e-05, "loss": 0.8699, "step": 6425, "task_loss": 0.21868453919887543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2764171361923218, "epoch": 5.43, "learning_rate": 3.543832185347527e-05, "loss": 1.056, "step": 6426, "task_loss": 1.0537376403808594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.382278323173523, "epoch": 5.43, "learning_rate": 3.54351909830933e-05, "loss": 1.0163, "step": 6427, "task_loss": 1.5680608749389648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9561434388160706, "epoch": 5.43, "learning_rate": 3.543206011271133e-05, "loss": 0.8619, "step": 6428, "task_loss": 0.91252601146698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8495367765426636, "epoch": 5.43, "learning_rate": 3.5428929242329365e-05, "loss": 0.9401, "step": 6429, "task_loss": 0.4400724470615387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8920392990112305, "epoch": 5.44, "learning_rate": 3.5425798371947404e-05, "loss": 0.8833, "step": 6430, "task_loss": 0.1780775785446167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.16690993309021, "epoch": 5.44, "learning_rate": 3.5422667501565435e-05, "loss": 1.2268, "step": 6431, "task_loss": 0.5418632626533508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6882870197296143, "epoch": 5.44, "learning_rate": 3.541953663118347e-05, "loss": 0.9025, "step": 6432, "task_loss": 0.4388028681278229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7205920219421387, "epoch": 5.44, "learning_rate": 3.54164057608015e-05, "loss": 1.1449, "step": 6433, "task_loss": 1.246321201324463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0943174362182617, "epoch": 5.44, "learning_rate": 3.541327489041954e-05, "loss": 1.1633, "step": 6434, "task_loss": 1.2947598695755005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1731948852539062, "epoch": 5.44, "learning_rate": 3.541014402003757e-05, "loss": 0.9418, "step": 6435, "task_loss": 0.850745677947998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1761341094970703, "epoch": 5.44, "learning_rate": 3.540701314965561e-05, "loss": 0.826, "step": 6436, "task_loss": 0.4950915277004242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0872869491577148, "epoch": 5.44, "learning_rate": 3.540388227927364e-05, "loss": 1.2235, "step": 6437, "task_loss": 1.366835117340088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8429986238479614, "epoch": 5.44, "learning_rate": 3.540075140889167e-05, "loss": 0.7758, "step": 6438, "task_loss": 0.5511539578437805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49486634135246277, "epoch": 5.44, "learning_rate": 3.539762053850971e-05, "loss": 0.7729, "step": 6439, "task_loss": 0.9223983883857727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1231046915054321, "epoch": 5.44, "learning_rate": 3.539448966812774e-05, "loss": 0.9615, "step": 6440, "task_loss": 1.4532215595245361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1274126768112183, "epoch": 5.44, "learning_rate": 3.539135879774578e-05, "loss": 1.03, "step": 6441, "task_loss": 1.2757904529571533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0556986331939697, "epoch": 5.45, "learning_rate": 3.538822792736381e-05, "loss": 1.2491, "step": 6442, "task_loss": 1.6777384281158447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9318726658821106, "epoch": 5.45, "learning_rate": 3.5385097056981844e-05, "loss": 0.8945, "step": 6443, "task_loss": 0.8508906960487366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7701947689056396, "epoch": 5.45, "learning_rate": 3.5381966186599876e-05, "loss": 1.0406, "step": 6444, "task_loss": 1.7601364850997925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4998117685317993, "epoch": 5.45, "learning_rate": 3.5378835316217914e-05, "loss": 1.5004, "step": 6445, "task_loss": 0.7875202894210815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4848902225494385, "epoch": 5.45, "learning_rate": 3.5375704445835946e-05, "loss": 1.4052, "step": 6446, "task_loss": 1.9354093074798584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0804474353790283, "epoch": 5.45, "learning_rate": 3.537257357545398e-05, "loss": 0.8909, "step": 6447, "task_loss": 0.6980280876159668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8370321989059448, "epoch": 5.45, "learning_rate": 3.536944270507201e-05, "loss": 1.071, "step": 6448, "task_loss": 1.1746385097503662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5208892822265625, "epoch": 5.45, "learning_rate": 3.536631183469005e-05, "loss": 1.1156, "step": 6449, "task_loss": 1.6435749530792236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.302030563354492, "epoch": 5.45, "learning_rate": 3.536318096430808e-05, "loss": 1.2344, "step": 6450, "task_loss": 0.99325031042099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0033084154129028, "epoch": 5.45, "learning_rate": 3.536005009392611e-05, "loss": 0.9529, "step": 6451, "task_loss": 1.1598353385925293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5826972723007202, "epoch": 5.45, "learning_rate": 3.535691922354415e-05, "loss": 1.1298, "step": 6452, "task_loss": 0.46119174361228943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6729243993759155, "epoch": 5.45, "learning_rate": 3.535378835316218e-05, "loss": 0.9164, "step": 6453, "task_loss": 0.9528642892837524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4008054733276367, "epoch": 5.46, "learning_rate": 3.5350657482780214e-05, "loss": 1.4418, "step": 6454, "task_loss": 1.3965712785720825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2289025783538818, "epoch": 5.46, "learning_rate": 3.5347526612398246e-05, "loss": 1.0125, "step": 6455, "task_loss": 0.8698993921279907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8941463232040405, "epoch": 5.46, "learning_rate": 3.5344395742016284e-05, "loss": 1.1504, "step": 6456, "task_loss": 0.8088862895965576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8513845205307007, "epoch": 5.46, "learning_rate": 3.5341264871634316e-05, "loss": 0.9844, "step": 6457, "task_loss": 0.7091092467308044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2404515743255615, "epoch": 5.46, "learning_rate": 3.533813400125235e-05, "loss": 1.0635, "step": 6458, "task_loss": 0.7728650569915771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7164535522460938, "epoch": 5.46, "learning_rate": 3.533500313087038e-05, "loss": 0.8865, "step": 6459, "task_loss": 0.5366823077201843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7448195219039917, "epoch": 5.46, "learning_rate": 3.533187226048842e-05, "loss": 1.2805, "step": 6460, "task_loss": 1.6417980194091797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3164863586425781, "epoch": 5.46, "learning_rate": 3.532874139010645e-05, "loss": 1.18, "step": 6461, "task_loss": 2.3099007606506348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0012719631195068, "epoch": 5.46, "learning_rate": 3.532561051972448e-05, "loss": 0.9963, "step": 6462, "task_loss": 0.9332010746002197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7868875861167908, "epoch": 5.46, "learning_rate": 3.532247964934252e-05, "loss": 1.1501, "step": 6463, "task_loss": 0.4712570905685425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2785568237304688, "epoch": 5.46, "learning_rate": 3.531934877896055e-05, "loss": 1.104, "step": 6464, "task_loss": 0.8090829253196716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9002989530563354, "epoch": 5.46, "learning_rate": 3.5316217908578584e-05, "loss": 0.9781, "step": 6465, "task_loss": 1.4962294101715088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.257664442062378, "epoch": 5.47, "learning_rate": 3.5313087038196616e-05, "loss": 0.9879, "step": 6466, "task_loss": 0.604256272315979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0811975002288818, "epoch": 5.47, "learning_rate": 3.5309956167814654e-05, "loss": 0.9147, "step": 6467, "task_loss": 1.7866227626800537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.10722017288208, "epoch": 5.47, "learning_rate": 3.5306825297432686e-05, "loss": 1.236, "step": 6468, "task_loss": 0.868015706539154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7373790740966797, "epoch": 5.47, "learning_rate": 3.530369442705072e-05, "loss": 1.278, "step": 6469, "task_loss": 2.4480528831481934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5817369222640991, "epoch": 5.47, "learning_rate": 3.5300563556668756e-05, "loss": 0.9688, "step": 6470, "task_loss": 0.4144582152366638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7592877149581909, "epoch": 5.47, "learning_rate": 3.529743268628679e-05, "loss": 0.9319, "step": 6471, "task_loss": 0.8928740620613098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6836357116699219, "epoch": 5.47, "learning_rate": 3.5294301815904827e-05, "loss": 0.9617, "step": 6472, "task_loss": 0.4807351231575012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4256367683410645, "epoch": 5.47, "learning_rate": 3.529117094552286e-05, "loss": 1.1211, "step": 6473, "task_loss": 1.6320196390151978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1116664409637451, "epoch": 5.47, "learning_rate": 3.52880400751409e-05, "loss": 1.1754, "step": 6474, "task_loss": 1.462411880493164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2853585481643677, "epoch": 5.47, "learning_rate": 3.528490920475893e-05, "loss": 0.9127, "step": 6475, "task_loss": 0.8428748846054077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7872532606124878, "epoch": 5.47, "learning_rate": 3.528177833437696e-05, "loss": 0.864, "step": 6476, "task_loss": 1.4581531286239624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4191910028457642, "epoch": 5.47, "learning_rate": 3.527864746399499e-05, "loss": 1.0942, "step": 6477, "task_loss": 1.3727918863296509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.983662486076355, "epoch": 5.48, "learning_rate": 3.527551659361303e-05, "loss": 1.0297, "step": 6478, "task_loss": 1.3408384323120117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6133733987808228, "epoch": 5.48, "learning_rate": 3.527238572323106e-05, "loss": 1.0162, "step": 6479, "task_loss": 0.6349960565567017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7707383036613464, "epoch": 5.48, "learning_rate": 3.5269254852849094e-05, "loss": 1.037, "step": 6480, "task_loss": 0.9542657136917114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0545612573623657, "epoch": 5.48, "learning_rate": 3.5266123982467126e-05, "loss": 0.8118, "step": 6481, "task_loss": 0.502586305141449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5072686672210693, "epoch": 5.48, "learning_rate": 3.5262993112085165e-05, "loss": 1.4636, "step": 6482, "task_loss": 1.1138575077056885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1153607368469238, "epoch": 5.48, "learning_rate": 3.5259862241703196e-05, "loss": 1.0795, "step": 6483, "task_loss": 0.6796573996543884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4949363470077515, "epoch": 5.48, "learning_rate": 3.525673137132123e-05, "loss": 1.0714, "step": 6484, "task_loss": 1.1958705186843872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9910521507263184, "epoch": 5.48, "learning_rate": 3.525360050093926e-05, "loss": 1.0868, "step": 6485, "task_loss": 1.100682020187378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3897485733032227, "epoch": 5.48, "learning_rate": 3.52504696305573e-05, "loss": 1.1183, "step": 6486, "task_loss": 1.3168612718582153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7732568979263306, "epoch": 5.48, "learning_rate": 3.524733876017533e-05, "loss": 1.1708, "step": 6487, "task_loss": 0.911601722240448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.383826732635498, "epoch": 5.48, "learning_rate": 3.524420788979336e-05, "loss": 1.0695, "step": 6488, "task_loss": 0.7365452647209167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5717445611953735, "epoch": 5.48, "learning_rate": 3.52410770194114e-05, "loss": 1.2136, "step": 6489, "task_loss": 0.9712956547737122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4830011129379272, "epoch": 5.49, "learning_rate": 3.523794614902943e-05, "loss": 1.0446, "step": 6490, "task_loss": 1.6223558187484741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.969512939453125, "epoch": 5.49, "learning_rate": 3.5234815278647464e-05, "loss": 1.085, "step": 6491, "task_loss": 2.7358968257904053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9703763127326965, "epoch": 5.49, "learning_rate": 3.5231684408265496e-05, "loss": 0.7666, "step": 6492, "task_loss": 0.28145283460617065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5131440162658691, "epoch": 5.49, "learning_rate": 3.5228553537883535e-05, "loss": 1.0463, "step": 6493, "task_loss": 1.3012194633483887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2698523998260498, "epoch": 5.49, "learning_rate": 3.5225422667501566e-05, "loss": 0.9708, "step": 6494, "task_loss": 0.6339731812477112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7602527141571045, "epoch": 5.49, "learning_rate": 3.52222917971196e-05, "loss": 1.2085, "step": 6495, "task_loss": 0.9359627962112427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5748775005340576, "epoch": 5.49, "learning_rate": 3.521916092673763e-05, "loss": 1.3184, "step": 6496, "task_loss": 1.90810227394104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6938300132751465, "epoch": 5.49, "learning_rate": 3.521603005635567e-05, "loss": 0.9497, "step": 6497, "task_loss": 1.8549816608428955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1808998584747314, "epoch": 5.49, "learning_rate": 3.52128991859737e-05, "loss": 0.9969, "step": 6498, "task_loss": 1.2452946901321411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9807001948356628, "epoch": 5.49, "learning_rate": 3.520976831559173e-05, "loss": 0.8768, "step": 6499, "task_loss": 0.9641425609588623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3169949054718018, "epoch": 5.49, "learning_rate": 3.520663744520977e-05, "loss": 1.2549, "step": 6500, "task_loss": 0.8817269802093506 }, { "epoch": 5.49, "eval_accuracy": 0.8695049504950495, "eval_loss": 0.6559610962867737, "eval_runtime": 207.402, "eval_samples_per_second": 121.744, "eval_steps_per_second": 0.955, "step": 6500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7670607566833496, "epoch": 5.5, "learning_rate": 3.52035065748278e-05, "loss": 0.8803, "step": 6501, "task_loss": 0.7151564359664917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7319363951683044, "epoch": 5.5, "learning_rate": 3.5200375704445834e-05, "loss": 0.9675, "step": 6502, "task_loss": 0.9163309335708618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8991597890853882, "epoch": 5.5, "learning_rate": 3.519724483406387e-05, "loss": 0.8684, "step": 6503, "task_loss": 0.6093701720237732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5415489673614502, "epoch": 5.5, "learning_rate": 3.5194113963681905e-05, "loss": 1.1494, "step": 6504, "task_loss": 1.149862289428711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3731272220611572, "epoch": 5.5, "learning_rate": 3.5190983093299936e-05, "loss": 0.9227, "step": 6505, "task_loss": 1.0957105159759521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.348618745803833, "epoch": 5.5, "learning_rate": 3.5187852222917975e-05, "loss": 1.1458, "step": 6506, "task_loss": 1.3253443241119385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7232856750488281, "epoch": 5.5, "learning_rate": 3.518472135253601e-05, "loss": 0.8952, "step": 6507, "task_loss": 1.0806362628936768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9598774909973145, "epoch": 5.5, "learning_rate": 3.5181590482154045e-05, "loss": 1.0199, "step": 6508, "task_loss": 0.6651568412780762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0539076328277588, "epoch": 5.5, "learning_rate": 3.517845961177208e-05, "loss": 0.9468, "step": 6509, "task_loss": 0.9811800122261047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7784957885742188, "epoch": 5.5, "learning_rate": 3.517532874139011e-05, "loss": 1.0631, "step": 6510, "task_loss": 0.7335864901542664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5568876266479492, "epoch": 5.5, "learning_rate": 3.517219787100815e-05, "loss": 1.2281, "step": 6511, "task_loss": 1.3201439380645752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3929016590118408, "epoch": 5.5, "learning_rate": 3.516906700062618e-05, "loss": 1.2498, "step": 6512, "task_loss": 1.6513925790786743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1124606132507324, "epoch": 5.51, "learning_rate": 3.516593613024421e-05, "loss": 0.9565, "step": 6513, "task_loss": 1.105628252029419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8921550512313843, "epoch": 5.51, "learning_rate": 3.516280525986224e-05, "loss": 1.1308, "step": 6514, "task_loss": 1.1509788036346436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5551445484161377, "epoch": 5.51, "learning_rate": 3.515967438948028e-05, "loss": 0.9553, "step": 6515, "task_loss": 1.257402777671814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9353476762771606, "epoch": 5.51, "learning_rate": 3.515654351909831e-05, "loss": 1.0768, "step": 6516, "task_loss": 1.0882173776626587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1788185834884644, "epoch": 5.51, "learning_rate": 3.5153412648716345e-05, "loss": 1.0226, "step": 6517, "task_loss": 0.7314975261688232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3897722959518433, "epoch": 5.51, "learning_rate": 3.5150281778334377e-05, "loss": 0.9851, "step": 6518, "task_loss": 1.3822243213653564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1716434955596924, "epoch": 5.51, "learning_rate": 3.5147150907952415e-05, "loss": 1.1943, "step": 6519, "task_loss": 1.1102166175842285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1207398176193237, "epoch": 5.51, "learning_rate": 3.514402003757045e-05, "loss": 1.0405, "step": 6520, "task_loss": 0.8855395913124084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3094619512557983, "epoch": 5.51, "learning_rate": 3.514088916718848e-05, "loss": 1.2825, "step": 6521, "task_loss": 0.7750612497329712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8269303441047668, "epoch": 5.51, "learning_rate": 3.513775829680652e-05, "loss": 0.7399, "step": 6522, "task_loss": 1.052710771560669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3154716491699219, "epoch": 5.51, "learning_rate": 3.513462742642455e-05, "loss": 1.1467, "step": 6523, "task_loss": 1.5675305128097534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6334990859031677, "epoch": 5.51, "learning_rate": 3.513149655604258e-05, "loss": 0.8651, "step": 6524, "task_loss": 0.12033046782016754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2443420886993408, "epoch": 5.52, "learning_rate": 3.512836568566061e-05, "loss": 1.1491, "step": 6525, "task_loss": 1.225780725479126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6865392327308655, "epoch": 5.52, "learning_rate": 3.512523481527865e-05, "loss": 1.2938, "step": 6526, "task_loss": 0.28426221013069153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8913328051567078, "epoch": 5.52, "learning_rate": 3.512210394489668e-05, "loss": 0.7977, "step": 6527, "task_loss": 0.49008986353874207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1251304149627686, "epoch": 5.52, "learning_rate": 3.5118973074514715e-05, "loss": 1.3192, "step": 6528, "task_loss": 1.0135204792022705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0962631702423096, "epoch": 5.52, "learning_rate": 3.5115842204132746e-05, "loss": 0.9487, "step": 6529, "task_loss": 1.0452762842178345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9359259605407715, "epoch": 5.52, "learning_rate": 3.5112711333750785e-05, "loss": 1.1064, "step": 6530, "task_loss": 1.3839014768600464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9288924932479858, "epoch": 5.52, "learning_rate": 3.510958046336882e-05, "loss": 1.0674, "step": 6531, "task_loss": 1.4705004692077637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4037395715713501, "epoch": 5.52, "learning_rate": 3.510644959298685e-05, "loss": 0.7232, "step": 6532, "task_loss": 0.4161471724510193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1633098125457764, "epoch": 5.52, "learning_rate": 3.510331872260488e-05, "loss": 1.1275, "step": 6533, "task_loss": 1.3454216718673706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5863195657730103, "epoch": 5.52, "learning_rate": 3.510018785222292e-05, "loss": 1.0411, "step": 6534, "task_loss": 0.6057096719741821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4624860286712646, "epoch": 5.52, "learning_rate": 3.509705698184095e-05, "loss": 1.0456, "step": 6535, "task_loss": 2.005662202835083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.645957350730896, "epoch": 5.52, "learning_rate": 3.509392611145898e-05, "loss": 1.326, "step": 6536, "task_loss": 0.5554060935974121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5946741104125977, "epoch": 5.53, "learning_rate": 3.509079524107702e-05, "loss": 1.3208, "step": 6537, "task_loss": 1.5686017274856567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8540177941322327, "epoch": 5.53, "learning_rate": 3.508766437069505e-05, "loss": 1.0841, "step": 6538, "task_loss": 0.5751433372497559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2663617134094238, "epoch": 5.53, "learning_rate": 3.508453350031309e-05, "loss": 1.1079, "step": 6539, "task_loss": 1.0886826515197754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3480929136276245, "epoch": 5.53, "learning_rate": 3.508140262993112e-05, "loss": 1.0562, "step": 6540, "task_loss": 1.168260931968689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4563570022583008, "epoch": 5.53, "learning_rate": 3.5078271759549155e-05, "loss": 1.0836, "step": 6541, "task_loss": 0.7335141897201538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0891141891479492, "epoch": 5.53, "learning_rate": 3.5075140889167194e-05, "loss": 1.0403, "step": 6542, "task_loss": 0.5779906511306763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0670543909072876, "epoch": 5.53, "learning_rate": 3.5072010018785225e-05, "loss": 1.2481, "step": 6543, "task_loss": 1.1764647960662842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2829759120941162, "epoch": 5.53, "learning_rate": 3.506887914840326e-05, "loss": 1.0864, "step": 6544, "task_loss": 1.3481394052505493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5074011087417603, "epoch": 5.53, "learning_rate": 3.5065748278021296e-05, "loss": 0.7245, "step": 6545, "task_loss": 0.26960498094558716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7399095296859741, "epoch": 5.53, "learning_rate": 3.506261740763933e-05, "loss": 0.8705, "step": 6546, "task_loss": 1.1187796592712402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6609180569648743, "epoch": 5.53, "learning_rate": 3.505948653725736e-05, "loss": 1.2716, "step": 6547, "task_loss": 0.061413299292325974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.049589991569519, "epoch": 5.53, "learning_rate": 3.50563556668754e-05, "loss": 0.8063, "step": 6548, "task_loss": 0.41330164670944214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.926830530166626, "epoch": 5.54, "learning_rate": 3.505322479649343e-05, "loss": 0.9592, "step": 6549, "task_loss": 1.2271808385849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6236917972564697, "epoch": 5.54, "learning_rate": 3.505009392611146e-05, "loss": 1.0733, "step": 6550, "task_loss": 1.2977533340454102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8228510618209839, "epoch": 5.54, "learning_rate": 3.504696305572949e-05, "loss": 0.7877, "step": 6551, "task_loss": 0.7246243357658386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.691398561000824, "epoch": 5.54, "learning_rate": 3.504383218534753e-05, "loss": 1.0765, "step": 6552, "task_loss": 1.2479419708251953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6601076722145081, "epoch": 5.54, "learning_rate": 3.5040701314965563e-05, "loss": 0.9668, "step": 6553, "task_loss": 0.6847500801086426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7214894890785217, "epoch": 5.54, "learning_rate": 3.5037570444583595e-05, "loss": 1.1393, "step": 6554, "task_loss": 0.8153467774391174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7717337012290955, "epoch": 5.54, "learning_rate": 3.503443957420163e-05, "loss": 0.7879, "step": 6555, "task_loss": 0.7754324078559875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5205346345901489, "epoch": 5.54, "learning_rate": 3.5031308703819666e-05, "loss": 0.9855, "step": 6556, "task_loss": 0.44444555044174194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0046558380126953, "epoch": 5.54, "learning_rate": 3.50281778334377e-05, "loss": 0.8146, "step": 6557, "task_loss": 1.6004341840744019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2292954921722412, "epoch": 5.54, "learning_rate": 3.502504696305573e-05, "loss": 1.0806, "step": 6558, "task_loss": 1.984984040260315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0625687837600708, "epoch": 5.54, "learning_rate": 3.502191609267377e-05, "loss": 0.9596, "step": 6559, "task_loss": 0.5812765955924988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7357698082923889, "epoch": 5.54, "learning_rate": 3.50187852222918e-05, "loss": 0.8694, "step": 6560, "task_loss": 0.7406274080276489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1158151626586914, "epoch": 5.55, "learning_rate": 3.501565435190983e-05, "loss": 1.1039, "step": 6561, "task_loss": 1.0075116157531738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3593974113464355, "epoch": 5.55, "learning_rate": 3.501252348152786e-05, "loss": 1.1365, "step": 6562, "task_loss": 1.1892555952072144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8201116323471069, "epoch": 5.55, "learning_rate": 3.50093926111459e-05, "loss": 0.7729, "step": 6563, "task_loss": 0.49186307191848755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7462435364723206, "epoch": 5.55, "learning_rate": 3.500626174076393e-05, "loss": 1.1225, "step": 6564, "task_loss": 1.2363544702529907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9348537921905518, "epoch": 5.55, "learning_rate": 3.5003130870381965e-05, "loss": 0.8363, "step": 6565, "task_loss": 2.0546114444732666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9056026339530945, "epoch": 5.55, "learning_rate": 3.5e-05, "loss": 1.001, "step": 6566, "task_loss": 1.2001638412475586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9921395778656006, "epoch": 5.55, "learning_rate": 3.4996869129618035e-05, "loss": 1.1643, "step": 6567, "task_loss": 1.0145716667175293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7653767466545105, "epoch": 5.55, "learning_rate": 3.499373825923607e-05, "loss": 0.8758, "step": 6568, "task_loss": 1.3037340641021729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.338374137878418, "epoch": 5.55, "learning_rate": 3.49906073888541e-05, "loss": 1.3433, "step": 6569, "task_loss": 0.7541374564170837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2296284437179565, "epoch": 5.55, "learning_rate": 3.498747651847214e-05, "loss": 0.8261, "step": 6570, "task_loss": 1.3932433128356934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4434834718704224, "epoch": 5.55, "learning_rate": 3.498434564809017e-05, "loss": 1.0627, "step": 6571, "task_loss": 1.5315967798233032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.097566843032837, "epoch": 5.56, "learning_rate": 3.49812147777082e-05, "loss": 0.9811, "step": 6572, "task_loss": 0.6884433627128601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8986769914627075, "epoch": 5.56, "learning_rate": 3.497808390732624e-05, "loss": 1.0605, "step": 6573, "task_loss": 0.27439364790916443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8275444507598877, "epoch": 5.56, "learning_rate": 3.497495303694427e-05, "loss": 1.022, "step": 6574, "task_loss": 1.1512285470962524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0177621841430664, "epoch": 5.56, "learning_rate": 3.497182216656231e-05, "loss": 0.9234, "step": 6575, "task_loss": 0.9589892029762268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.804730236530304, "epoch": 5.56, "learning_rate": 3.496869129618034e-05, "loss": 1.3831, "step": 6576, "task_loss": 0.4312915503978729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3546161651611328, "epoch": 5.56, "learning_rate": 3.4965560425798374e-05, "loss": 0.8985, "step": 6577, "task_loss": 1.315101981163025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3033215999603271, "epoch": 5.56, "learning_rate": 3.496242955541641e-05, "loss": 1.3394, "step": 6578, "task_loss": 2.160012722015381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5016425848007202, "epoch": 5.56, "learning_rate": 3.4959298685034444e-05, "loss": 0.6612, "step": 6579, "task_loss": 0.2701767086982727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1882069110870361, "epoch": 5.56, "learning_rate": 3.4956167814652476e-05, "loss": 0.7864, "step": 6580, "task_loss": 1.39484441280365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9984457492828369, "epoch": 5.56, "learning_rate": 3.495303694427051e-05, "loss": 0.8067, "step": 6581, "task_loss": 0.733373761177063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1925852298736572, "epoch": 5.56, "learning_rate": 3.4949906073888546e-05, "loss": 1.0737, "step": 6582, "task_loss": 2.019429922103882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3052089214324951, "epoch": 5.56, "learning_rate": 3.494677520350658e-05, "loss": 1.1183, "step": 6583, "task_loss": 1.0307042598724365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6474347710609436, "epoch": 5.57, "learning_rate": 3.494364433312461e-05, "loss": 0.8668, "step": 6584, "task_loss": 0.6333401203155518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7065171003341675, "epoch": 5.57, "learning_rate": 3.494051346274265e-05, "loss": 0.6682, "step": 6585, "task_loss": 0.8855287432670593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9136041402816772, "epoch": 5.57, "learning_rate": 3.493738259236068e-05, "loss": 1.0847, "step": 6586, "task_loss": 0.986397922039032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1949249505996704, "epoch": 5.57, "learning_rate": 3.493425172197871e-05, "loss": 0.955, "step": 6587, "task_loss": 1.2919868230819702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7149330377578735, "epoch": 5.57, "learning_rate": 3.4931120851596744e-05, "loss": 0.7936, "step": 6588, "task_loss": 0.43637216091156006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0504767894744873, "epoch": 5.57, "learning_rate": 3.492798998121478e-05, "loss": 0.9961, "step": 6589, "task_loss": 0.951395571231842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8962199687957764, "epoch": 5.57, "learning_rate": 3.4924859110832814e-05, "loss": 1.0348, "step": 6590, "task_loss": 1.1037795543670654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.402076005935669, "epoch": 5.57, "learning_rate": 3.4921728240450846e-05, "loss": 1.1394, "step": 6591, "task_loss": 1.0215140581130981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8759152889251709, "epoch": 5.57, "learning_rate": 3.491859737006888e-05, "loss": 0.9754, "step": 6592, "task_loss": 1.1429678201675415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2350777387619019, "epoch": 5.57, "learning_rate": 3.4915466499686916e-05, "loss": 1.0464, "step": 6593, "task_loss": 0.7531865835189819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2301359176635742, "epoch": 5.57, "learning_rate": 3.491233562930495e-05, "loss": 1.1181, "step": 6594, "task_loss": 1.2469125986099243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6570467948913574, "epoch": 5.57, "learning_rate": 3.490920475892298e-05, "loss": 1.3735, "step": 6595, "task_loss": 1.5098928213119507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0289840698242188, "epoch": 5.58, "learning_rate": 3.490607388854102e-05, "loss": 1.2972, "step": 6596, "task_loss": 1.584557294845581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8213797211647034, "epoch": 5.58, "learning_rate": 3.490294301815905e-05, "loss": 0.9932, "step": 6597, "task_loss": 0.20164215564727783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.704422116279602, "epoch": 5.58, "learning_rate": 3.489981214777708e-05, "loss": 0.8943, "step": 6598, "task_loss": 0.4473611116409302 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.005910038948059, "epoch": 5.58, "learning_rate": 3.4896681277395113e-05, "loss": 1.0494, "step": 6599, "task_loss": 0.9249246120452881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7101396322250366, "epoch": 5.58, "learning_rate": 3.489355040701315e-05, "loss": 0.8846, "step": 6600, "task_loss": 1.1577047109603882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1190091371536255, "epoch": 5.58, "learning_rate": 3.4890419536631184e-05, "loss": 1.1515, "step": 6601, "task_loss": 1.59475576877594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.817979633808136, "epoch": 5.58, "learning_rate": 3.4887288666249216e-05, "loss": 1.0701, "step": 6602, "task_loss": 0.7703238725662231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0304832458496094, "epoch": 5.58, "learning_rate": 3.488415779586725e-05, "loss": 0.9782, "step": 6603, "task_loss": 1.2944889068603516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0177359580993652, "epoch": 5.58, "learning_rate": 3.4881026925485286e-05, "loss": 1.0029, "step": 6604, "task_loss": 0.5953258872032166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0797600746154785, "epoch": 5.58, "learning_rate": 3.487789605510332e-05, "loss": 0.8184, "step": 6605, "task_loss": 0.7229672074317932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7886568307876587, "epoch": 5.58, "learning_rate": 3.4874765184721356e-05, "loss": 0.8895, "step": 6606, "task_loss": 0.7107439637184143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0582889318466187, "epoch": 5.58, "learning_rate": 3.487163431433939e-05, "loss": 1.0689, "step": 6607, "task_loss": 1.2527813911437988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0404630899429321, "epoch": 5.59, "learning_rate": 3.486850344395742e-05, "loss": 1.0704, "step": 6608, "task_loss": 1.0951716899871826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9468684196472168, "epoch": 5.59, "learning_rate": 3.486537257357546e-05, "loss": 1.0025, "step": 6609, "task_loss": 0.9522283673286438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6813836693763733, "epoch": 5.59, "learning_rate": 3.486224170319349e-05, "loss": 1.114, "step": 6610, "task_loss": 1.2223799228668213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6244620680809021, "epoch": 5.59, "learning_rate": 3.485911083281153e-05, "loss": 0.8879, "step": 6611, "task_loss": 0.7821725606918335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9846162796020508, "epoch": 5.59, "learning_rate": 3.485597996242956e-05, "loss": 0.9457, "step": 6612, "task_loss": 1.5154057741165161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8304262161254883, "epoch": 5.59, "learning_rate": 3.485284909204759e-05, "loss": 1.0244, "step": 6613, "task_loss": 0.36280563473701477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1766122579574585, "epoch": 5.59, "learning_rate": 3.4849718221665624e-05, "loss": 1.015, "step": 6614, "task_loss": 0.7364667654037476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6606078147888184, "epoch": 5.59, "learning_rate": 3.484658735128366e-05, "loss": 0.9244, "step": 6615, "task_loss": 1.0123035907745361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1629263162612915, "epoch": 5.59, "learning_rate": 3.4843456480901694e-05, "loss": 1.1467, "step": 6616, "task_loss": 1.5898222923278809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.167676329612732, "epoch": 5.59, "learning_rate": 3.4840325610519726e-05, "loss": 0.9543, "step": 6617, "task_loss": 1.167914628982544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0045502185821533, "epoch": 5.59, "learning_rate": 3.483719474013776e-05, "loss": 1.0782, "step": 6618, "task_loss": 1.3196532726287842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7791489362716675, "epoch": 5.59, "learning_rate": 3.4834063869755797e-05, "loss": 0.794, "step": 6619, "task_loss": 1.628294825553894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6722187995910645, "epoch": 5.6, "learning_rate": 3.483093299937383e-05, "loss": 0.8854, "step": 6620, "task_loss": 0.8673161864280701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9036926031112671, "epoch": 5.6, "learning_rate": 3.482780212899186e-05, "loss": 0.9781, "step": 6621, "task_loss": 0.5865060687065125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8528828620910645, "epoch": 5.6, "learning_rate": 3.48246712586099e-05, "loss": 0.9903, "step": 6622, "task_loss": 1.3568795919418335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9477301836013794, "epoch": 5.6, "learning_rate": 3.482154038822793e-05, "loss": 0.9029, "step": 6623, "task_loss": 1.4101219177246094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7083646059036255, "epoch": 5.6, "learning_rate": 3.481840951784596e-05, "loss": 0.8989, "step": 6624, "task_loss": 0.9004839062690735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5503624677658081, "epoch": 5.6, "learning_rate": 3.4815278647463994e-05, "loss": 0.8147, "step": 6625, "task_loss": 0.1812506765127182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6054623126983643, "epoch": 5.6, "learning_rate": 3.481214777708203e-05, "loss": 1.4559, "step": 6626, "task_loss": 1.633237600326538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2052398920059204, "epoch": 5.6, "learning_rate": 3.4809016906700064e-05, "loss": 0.7694, "step": 6627, "task_loss": 1.1953551769256592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.959985613822937, "epoch": 5.6, "learning_rate": 3.4805886036318096e-05, "loss": 1.0205, "step": 6628, "task_loss": 0.7583950757980347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6330899000167847, "epoch": 5.6, "learning_rate": 3.480275516593613e-05, "loss": 0.8541, "step": 6629, "task_loss": 0.8433346748352051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7329245209693909, "epoch": 5.6, "learning_rate": 3.4799624295554166e-05, "loss": 0.7923, "step": 6630, "task_loss": 0.4547787308692932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43918758630752563, "epoch": 5.6, "learning_rate": 3.47964934251722e-05, "loss": 0.8019, "step": 6631, "task_loss": 1.0181941986083984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8630151152610779, "epoch": 5.61, "learning_rate": 3.479336255479023e-05, "loss": 0.7777, "step": 6632, "task_loss": 0.6050834059715271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.54451984167099, "epoch": 5.61, "learning_rate": 3.479023168440827e-05, "loss": 1.1746, "step": 6633, "task_loss": 0.8030322194099426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1862462759017944, "epoch": 5.61, "learning_rate": 3.47871008140263e-05, "loss": 1.2328, "step": 6634, "task_loss": 0.7088916897773743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8252637982368469, "epoch": 5.61, "learning_rate": 3.478396994364433e-05, "loss": 1.0052, "step": 6635, "task_loss": 0.5815935730934143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.0381293296813965, "epoch": 5.61, "learning_rate": 3.4780839073262364e-05, "loss": 1.2364, "step": 6636, "task_loss": 1.6578774452209473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8960502743721008, "epoch": 5.61, "learning_rate": 3.47777082028804e-05, "loss": 0.9219, "step": 6637, "task_loss": 0.6694778203964233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6250662803649902, "epoch": 5.61, "learning_rate": 3.4774577332498434e-05, "loss": 1.2718, "step": 6638, "task_loss": 1.4790406227111816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3996056020259857, "epoch": 5.61, "learning_rate": 3.4771446462116466e-05, "loss": 0.9881, "step": 6639, "task_loss": 0.19963154196739197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5398781299591064, "epoch": 5.61, "learning_rate": 3.4768315591734505e-05, "loss": 0.8348, "step": 6640, "task_loss": 0.8020597696304321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9610475301742554, "epoch": 5.61, "learning_rate": 3.4765184721352536e-05, "loss": 1.032, "step": 6641, "task_loss": 1.1197824478149414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5821443796157837, "epoch": 5.61, "learning_rate": 3.4762053850970575e-05, "loss": 0.8396, "step": 6642, "task_loss": 0.3511286973953247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.670217752456665, "epoch": 5.61, "learning_rate": 3.475892298058861e-05, "loss": 0.9005, "step": 6643, "task_loss": 0.9970346689224243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8406213521957397, "epoch": 5.62, "learning_rate": 3.475579211020664e-05, "loss": 1.0056, "step": 6644, "task_loss": 0.6775112152099609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9372668266296387, "epoch": 5.62, "learning_rate": 3.475266123982468e-05, "loss": 0.9848, "step": 6645, "task_loss": 0.7128216624259949 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6106367111206055, "epoch": 5.62, "learning_rate": 3.474953036944271e-05, "loss": 0.8289, "step": 6646, "task_loss": 0.24855920672416687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6629154682159424, "epoch": 5.62, "learning_rate": 3.474639949906074e-05, "loss": 0.8446, "step": 6647, "task_loss": 0.8605557680130005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.009009838104248, "epoch": 5.62, "learning_rate": 3.474326862867878e-05, "loss": 1.057, "step": 6648, "task_loss": 0.37451669573783875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5989046096801758, "epoch": 5.62, "learning_rate": 3.474013775829681e-05, "loss": 0.9386, "step": 6649, "task_loss": 0.9155551195144653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2369914054870605, "epoch": 5.62, "learning_rate": 3.473700688791484e-05, "loss": 1.0902, "step": 6650, "task_loss": 1.4583839178085327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6337558627128601, "epoch": 5.62, "learning_rate": 3.4733876017532875e-05, "loss": 0.787, "step": 6651, "task_loss": 0.9837329387664795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1450562477111816, "epoch": 5.62, "learning_rate": 3.473074514715091e-05, "loss": 0.8243, "step": 6652, "task_loss": 1.4322789907455444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.310030221939087, "epoch": 5.62, "learning_rate": 3.4727614276768945e-05, "loss": 0.9842, "step": 6653, "task_loss": 1.0932469367980957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9735713005065918, "epoch": 5.62, "learning_rate": 3.472448340638698e-05, "loss": 0.7375, "step": 6654, "task_loss": 0.852595329284668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.18473482131958, "epoch": 5.63, "learning_rate": 3.472135253600501e-05, "loss": 0.8874, "step": 6655, "task_loss": 1.451383113861084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8809065818786621, "epoch": 5.63, "learning_rate": 3.471822166562305e-05, "loss": 0.8827, "step": 6656, "task_loss": 0.9997420907020569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8551783561706543, "epoch": 5.63, "learning_rate": 3.471509079524108e-05, "loss": 0.9863, "step": 6657, "task_loss": 1.2487871646881104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5097219944000244, "epoch": 5.63, "learning_rate": 3.471195992485911e-05, "loss": 1.319, "step": 6658, "task_loss": 1.2756050825119019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8988037109375, "epoch": 5.63, "learning_rate": 3.470882905447715e-05, "loss": 0.9607, "step": 6659, "task_loss": 0.5559248328208923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.183026671409607, "epoch": 5.63, "learning_rate": 3.470569818409518e-05, "loss": 1.0504, "step": 6660, "task_loss": 0.8332744240760803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7614070177078247, "epoch": 5.63, "learning_rate": 3.470256731371321e-05, "loss": 0.9079, "step": 6661, "task_loss": 1.282742977142334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6625787019729614, "epoch": 5.63, "learning_rate": 3.4699436443331244e-05, "loss": 1.0238, "step": 6662, "task_loss": 1.4337284564971924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9705732464790344, "epoch": 5.63, "learning_rate": 3.469630557294928e-05, "loss": 0.7699, "step": 6663, "task_loss": 0.6357478499412537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1060912609100342, "epoch": 5.63, "learning_rate": 3.4693174702567315e-05, "loss": 0.8702, "step": 6664, "task_loss": 1.5996599197387695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9897202253341675, "epoch": 5.63, "learning_rate": 3.4690043832185347e-05, "loss": 1.0373, "step": 6665, "task_loss": 0.981027364730835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.418851375579834, "epoch": 5.63, "learning_rate": 3.468691296180338e-05, "loss": 0.7385, "step": 6666, "task_loss": 0.42243802547454834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8067449927330017, "epoch": 5.64, "learning_rate": 3.468378209142142e-05, "loss": 0.9232, "step": 6667, "task_loss": 1.1512974500656128 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9117462635040283, "epoch": 5.64, "learning_rate": 3.468065122103945e-05, "loss": 1.1029, "step": 6668, "task_loss": 0.4552113115787506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9768902063369751, "epoch": 5.64, "learning_rate": 3.467752035065748e-05, "loss": 0.8788, "step": 6669, "task_loss": 0.506569504737854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.092519998550415, "epoch": 5.64, "learning_rate": 3.467438948027552e-05, "loss": 1.2538, "step": 6670, "task_loss": 0.7843132019042969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48467138409614563, "epoch": 5.64, "learning_rate": 3.467125860989355e-05, "loss": 0.6688, "step": 6671, "task_loss": 0.34746992588043213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4690340757369995, "epoch": 5.64, "learning_rate": 3.466812773951158e-05, "loss": 1.1113, "step": 6672, "task_loss": 0.8301376700401306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0578720569610596, "epoch": 5.64, "learning_rate": 3.466499686912962e-05, "loss": 1.19, "step": 6673, "task_loss": 1.4110254049301147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8043971657752991, "epoch": 5.64, "learning_rate": 3.466186599874765e-05, "loss": 0.8827, "step": 6674, "task_loss": 1.0772510766983032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8103055953979492, "epoch": 5.64, "learning_rate": 3.4658735128365685e-05, "loss": 0.6972, "step": 6675, "task_loss": 0.31249502301216125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6011784672737122, "epoch": 5.64, "learning_rate": 3.465560425798372e-05, "loss": 0.8945, "step": 6676, "task_loss": 0.43088817596435547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2279930114746094, "epoch": 5.64, "learning_rate": 3.4652473387601755e-05, "loss": 1.0387, "step": 6677, "task_loss": 1.0060359239578247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45848512649536133, "epoch": 5.64, "learning_rate": 3.4649342517219794e-05, "loss": 0.9765, "step": 6678, "task_loss": 0.26078349351882935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0192776918411255, "epoch": 5.65, "learning_rate": 3.4646211646837825e-05, "loss": 1.0007, "step": 6679, "task_loss": 0.879000723361969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.373227596282959, "epoch": 5.65, "learning_rate": 3.464308077645586e-05, "loss": 1.2727, "step": 6680, "task_loss": 1.3427433967590332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6780997514724731, "epoch": 5.65, "learning_rate": 3.463994990607389e-05, "loss": 0.7272, "step": 6681, "task_loss": 0.11943206936120987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6115241050720215, "epoch": 5.65, "learning_rate": 3.463681903569193e-05, "loss": 0.7534, "step": 6682, "task_loss": 0.21717725694179535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8050803542137146, "epoch": 5.65, "learning_rate": 3.463368816530996e-05, "loss": 1.0544, "step": 6683, "task_loss": 0.7581562995910645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9570859670639038, "epoch": 5.65, "learning_rate": 3.463055729492799e-05, "loss": 1.3808, "step": 6684, "task_loss": 1.4510577917099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9066624641418457, "epoch": 5.65, "learning_rate": 3.462742642454603e-05, "loss": 0.845, "step": 6685, "task_loss": 0.6439868807792664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1185392141342163, "epoch": 5.65, "learning_rate": 3.462429555416406e-05, "loss": 0.9986, "step": 6686, "task_loss": 1.6982088088989258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34580814838409424, "epoch": 5.65, "learning_rate": 3.462116468378209e-05, "loss": 0.6314, "step": 6687, "task_loss": 0.153678297996521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8500189781188965, "epoch": 5.65, "learning_rate": 3.4618033813400125e-05, "loss": 1.2165, "step": 6688, "task_loss": 1.3705567121505737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6247624158859253, "epoch": 5.65, "learning_rate": 3.4614902943018164e-05, "loss": 0.7919, "step": 6689, "task_loss": 1.3475784063339233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7604828476905823, "epoch": 5.65, "learning_rate": 3.4611772072636195e-05, "loss": 0.7229, "step": 6690, "task_loss": 1.2115044593811035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1588345766067505, "epoch": 5.66, "learning_rate": 3.460864120225423e-05, "loss": 1.3864, "step": 6691, "task_loss": 1.0025798082351685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2367191314697266, "epoch": 5.66, "learning_rate": 3.460551033187226e-05, "loss": 1.2027, "step": 6692, "task_loss": 2.144637107849121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5100722312927246, "epoch": 5.66, "learning_rate": 3.46023794614903e-05, "loss": 1.2789, "step": 6693, "task_loss": 1.251057744026184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1362872123718262, "epoch": 5.66, "learning_rate": 3.459924859110833e-05, "loss": 1.0496, "step": 6694, "task_loss": 0.4880228638648987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1559550762176514, "epoch": 5.66, "learning_rate": 3.459611772072636e-05, "loss": 1.2143, "step": 6695, "task_loss": 0.67025226354599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.277501106262207, "epoch": 5.66, "learning_rate": 3.45929868503444e-05, "loss": 0.9974, "step": 6696, "task_loss": 0.9419430494308472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9824845790863037, "epoch": 5.66, "learning_rate": 3.458985597996243e-05, "loss": 0.8467, "step": 6697, "task_loss": 0.7860819101333618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8685312271118164, "epoch": 5.66, "learning_rate": 3.458672510958046e-05, "loss": 0.7902, "step": 6698, "task_loss": 0.15061281621456146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6604938507080078, "epoch": 5.66, "learning_rate": 3.4583594239198495e-05, "loss": 0.7521, "step": 6699, "task_loss": 0.9795024991035461 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.131030797958374, "epoch": 5.66, "learning_rate": 3.4580463368816533e-05, "loss": 0.9853, "step": 6700, "task_loss": 1.6943680047988892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6858046650886536, "epoch": 5.66, "learning_rate": 3.4577332498434565e-05, "loss": 0.7411, "step": 6701, "task_loss": 0.3604755103588104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.141845464706421, "epoch": 5.66, "learning_rate": 3.45742016280526e-05, "loss": 0.8195, "step": 6702, "task_loss": 0.7092632055282593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7573987245559692, "epoch": 5.67, "learning_rate": 3.457107075767063e-05, "loss": 0.7938, "step": 6703, "task_loss": 0.7890477180480957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1855480670928955, "epoch": 5.67, "learning_rate": 3.456793988728867e-05, "loss": 1.0844, "step": 6704, "task_loss": 1.561013102531433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9001875519752502, "epoch": 5.67, "learning_rate": 3.45648090169067e-05, "loss": 1.0923, "step": 6705, "task_loss": 1.6436747312545776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3021154403686523, "epoch": 5.67, "learning_rate": 3.456167814652473e-05, "loss": 0.8436, "step": 6706, "task_loss": 0.8426536321640015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9432727098464966, "epoch": 5.67, "learning_rate": 3.455854727614277e-05, "loss": 0.7594, "step": 6707, "task_loss": 1.3524242639541626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1389557123184204, "epoch": 5.67, "learning_rate": 3.45554164057608e-05, "loss": 0.9254, "step": 6708, "task_loss": 0.8182578086853027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7380849719047546, "epoch": 5.67, "learning_rate": 3.455228553537884e-05, "loss": 0.9191, "step": 6709, "task_loss": 0.4796901345252991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9076193571090698, "epoch": 5.67, "learning_rate": 3.454915466499687e-05, "loss": 0.9743, "step": 6710, "task_loss": 0.7789208292961121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3289003372192383, "epoch": 5.67, "learning_rate": 3.454602379461491e-05, "loss": 1.2903, "step": 6711, "task_loss": 1.1539658308029175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1768125295639038, "epoch": 5.67, "learning_rate": 3.454289292423294e-05, "loss": 0.943, "step": 6712, "task_loss": 0.5002838373184204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.699855625629425, "epoch": 5.67, "learning_rate": 3.4539762053850974e-05, "loss": 0.9047, "step": 6713, "task_loss": 0.6357236504554749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6390883922576904, "epoch": 5.67, "learning_rate": 3.4536631183469005e-05, "loss": 1.2245, "step": 6714, "task_loss": 1.7679835557937622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7382948398590088, "epoch": 5.68, "learning_rate": 3.4533500313087044e-05, "loss": 0.961, "step": 6715, "task_loss": 0.8896480798721313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.883655309677124, "epoch": 5.68, "learning_rate": 3.4530369442705076e-05, "loss": 0.9251, "step": 6716, "task_loss": 0.2188754826784134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0567823648452759, "epoch": 5.68, "learning_rate": 3.452723857232311e-05, "loss": 1.1319, "step": 6717, "task_loss": 0.8278278112411499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3290064334869385, "epoch": 5.68, "learning_rate": 3.452410770194114e-05, "loss": 0.9571, "step": 6718, "task_loss": 2.5940897464752197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5678396224975586, "epoch": 5.68, "learning_rate": 3.452097683155918e-05, "loss": 1.364, "step": 6719, "task_loss": 0.6950979232788086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1593945026397705, "epoch": 5.68, "learning_rate": 3.451784596117721e-05, "loss": 0.9418, "step": 6720, "task_loss": 1.13798189163208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6492642164230347, "epoch": 5.68, "learning_rate": 3.451471509079524e-05, "loss": 1.112, "step": 6721, "task_loss": 1.0465567111968994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9119036197662354, "epoch": 5.68, "learning_rate": 3.451158422041328e-05, "loss": 1.0119, "step": 6722, "task_loss": 0.5283597111701965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2325292825698853, "epoch": 5.68, "learning_rate": 3.450845335003131e-05, "loss": 1.062, "step": 6723, "task_loss": 1.808332920074463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8311161994934082, "epoch": 5.68, "learning_rate": 3.4505322479649344e-05, "loss": 1.0115, "step": 6724, "task_loss": 1.0877290964126587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0311212539672852, "epoch": 5.68, "learning_rate": 3.4502191609267375e-05, "loss": 1.1273, "step": 6725, "task_loss": 0.7874642014503479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.984528124332428, "epoch": 5.69, "learning_rate": 3.4499060738885414e-05, "loss": 0.8936, "step": 6726, "task_loss": 1.2714612483978271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0900547504425049, "epoch": 5.69, "learning_rate": 3.4495929868503446e-05, "loss": 1.1614, "step": 6727, "task_loss": 1.4324233531951904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.212727665901184, "epoch": 5.69, "learning_rate": 3.449279899812148e-05, "loss": 1.1282, "step": 6728, "task_loss": 1.9263601303100586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5216874480247498, "epoch": 5.69, "learning_rate": 3.448966812773951e-05, "loss": 1.0535, "step": 6729, "task_loss": 0.41171255707740784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6727341413497925, "epoch": 5.69, "learning_rate": 3.448653725735755e-05, "loss": 0.6916, "step": 6730, "task_loss": 1.1546810865402222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7868494391441345, "epoch": 5.69, "learning_rate": 3.448340638697558e-05, "loss": 1.0024, "step": 6731, "task_loss": 1.934863805770874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6388569474220276, "epoch": 5.69, "learning_rate": 3.448027551659361e-05, "loss": 1.079, "step": 6732, "task_loss": 0.8772174119949341 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7990570068359375, "epoch": 5.69, "learning_rate": 3.447714464621165e-05, "loss": 0.7655, "step": 6733, "task_loss": 0.41090333461761475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.191469430923462, "epoch": 5.69, "learning_rate": 3.447401377582968e-05, "loss": 0.9045, "step": 6734, "task_loss": 1.1147217750549316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2628508806228638, "epoch": 5.69, "learning_rate": 3.4470882905447714e-05, "loss": 1.0841, "step": 6735, "task_loss": 0.6997637748718262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6626905798912048, "epoch": 5.69, "learning_rate": 3.4467752035065745e-05, "loss": 0.8143, "step": 6736, "task_loss": 0.8942087292671204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.845964789390564, "epoch": 5.69, "learning_rate": 3.4464621164683784e-05, "loss": 0.7713, "step": 6737, "task_loss": 0.7430846095085144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8494870066642761, "epoch": 5.7, "learning_rate": 3.4461490294301816e-05, "loss": 0.8474, "step": 6738, "task_loss": 1.3855009078979492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9008998870849609, "epoch": 5.7, "learning_rate": 3.445835942391985e-05, "loss": 1.0558, "step": 6739, "task_loss": 0.3017881512641907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9143382906913757, "epoch": 5.7, "learning_rate": 3.4455228553537886e-05, "loss": 0.8752, "step": 6740, "task_loss": 0.43693313002586365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9610604047775269, "epoch": 5.7, "learning_rate": 3.445209768315592e-05, "loss": 1.0613, "step": 6741, "task_loss": 1.3319991827011108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9544972777366638, "epoch": 5.7, "learning_rate": 3.444896681277395e-05, "loss": 1.2444, "step": 6742, "task_loss": 1.1779448986053467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8036787509918213, "epoch": 5.7, "learning_rate": 3.444583594239199e-05, "loss": 0.9511, "step": 6743, "task_loss": 1.376995325088501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.871783971786499, "epoch": 5.7, "learning_rate": 3.444270507201002e-05, "loss": 0.8101, "step": 6744, "task_loss": 1.0454481840133667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4121613502502441, "epoch": 5.7, "learning_rate": 3.443957420162806e-05, "loss": 1.1028, "step": 6745, "task_loss": 1.3023372888565063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6915987133979797, "epoch": 5.7, "learning_rate": 3.443644333124609e-05, "loss": 1.0403, "step": 6746, "task_loss": 1.2590889930725098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0600472688674927, "epoch": 5.7, "learning_rate": 3.443331246086412e-05, "loss": 1.0058, "step": 6747, "task_loss": 0.29208144545555115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5021612644195557, "epoch": 5.7, "learning_rate": 3.443018159048216e-05, "loss": 0.8986, "step": 6748, "task_loss": 0.7213618159294128 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9333989024162292, "epoch": 5.7, "learning_rate": 3.442705072010019e-05, "loss": 0.7386, "step": 6749, "task_loss": 1.1914031505584717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7156679630279541, "epoch": 5.71, "learning_rate": 3.4423919849718224e-05, "loss": 0.7004, "step": 6750, "task_loss": 0.502121090888977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0100656747817993, "epoch": 5.71, "learning_rate": 3.4420788979336256e-05, "loss": 0.8241, "step": 6751, "task_loss": 1.21210777759552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8718386888504028, "epoch": 5.71, "learning_rate": 3.4417658108954294e-05, "loss": 0.7163, "step": 6752, "task_loss": 0.9056766629219055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.704569399356842, "epoch": 5.71, "learning_rate": 3.4414527238572326e-05, "loss": 0.7974, "step": 6753, "task_loss": 0.7914525866508484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3627746105194092, "epoch": 5.71, "learning_rate": 3.441139636819036e-05, "loss": 1.1851, "step": 6754, "task_loss": 2.196011543273926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7290357351303101, "epoch": 5.71, "learning_rate": 3.4408265497808397e-05, "loss": 1.131, "step": 6755, "task_loss": 1.1840615272521973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6945778131484985, "epoch": 5.71, "learning_rate": 3.440513462742643e-05, "loss": 0.8901, "step": 6756, "task_loss": 0.761883020401001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2775368690490723, "epoch": 5.71, "learning_rate": 3.440200375704446e-05, "loss": 1.0485, "step": 6757, "task_loss": 1.0773380994796753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5838445425033569, "epoch": 5.71, "learning_rate": 3.439887288666249e-05, "loss": 0.7277, "step": 6758, "task_loss": 0.997978687286377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7496052980422974, "epoch": 5.71, "learning_rate": 3.439574201628053e-05, "loss": 0.8001, "step": 6759, "task_loss": 0.7876948714256287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0787291526794434, "epoch": 5.71, "learning_rate": 3.439261114589856e-05, "loss": 0.9067, "step": 6760, "task_loss": 1.2232123613357544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9356867074966431, "epoch": 5.71, "learning_rate": 3.4389480275516594e-05, "loss": 1.1712, "step": 6761, "task_loss": 0.5709046125411987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8666655421257019, "epoch": 5.72, "learning_rate": 3.4386349405134626e-05, "loss": 0.8273, "step": 6762, "task_loss": 0.7747504115104675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9656802415847778, "epoch": 5.72, "learning_rate": 3.4383218534752664e-05, "loss": 0.8993, "step": 6763, "task_loss": 0.5539913773536682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42221879959106445, "epoch": 5.72, "learning_rate": 3.4380087664370696e-05, "loss": 0.8943, "step": 6764, "task_loss": 0.9233754873275757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9743028879165649, "epoch": 5.72, "learning_rate": 3.437695679398873e-05, "loss": 1.1533, "step": 6765, "task_loss": 1.0506963729858398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1443817615509033, "epoch": 5.72, "learning_rate": 3.437382592360676e-05, "loss": 1.3417, "step": 6766, "task_loss": 1.5527923107147217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0372629165649414, "epoch": 5.72, "learning_rate": 3.43706950532248e-05, "loss": 1.1166, "step": 6767, "task_loss": 1.0547542572021484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9608932733535767, "epoch": 5.72, "learning_rate": 3.436756418284283e-05, "loss": 1.1414, "step": 6768, "task_loss": 0.945680558681488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9359352588653564, "epoch": 5.72, "learning_rate": 3.436443331246086e-05, "loss": 1.2666, "step": 6769, "task_loss": 1.7312777042388916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.700814962387085, "epoch": 5.72, "learning_rate": 3.43613024420789e-05, "loss": 0.7805, "step": 6770, "task_loss": 0.48391878604888916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1375958919525146, "epoch": 5.72, "learning_rate": 3.435817157169693e-05, "loss": 1.1343, "step": 6771, "task_loss": 1.1330835819244385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0507416725158691, "epoch": 5.72, "learning_rate": 3.4355040701314964e-05, "loss": 0.9541, "step": 6772, "task_loss": 1.1874196529388428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6646072864532471, "epoch": 5.72, "learning_rate": 3.4351909830932996e-05, "loss": 0.9866, "step": 6773, "task_loss": 1.9092590808868408 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7793500423431396, "epoch": 5.73, "learning_rate": 3.4348778960551034e-05, "loss": 0.854, "step": 6774, "task_loss": 0.45860111713409424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2328341007232666, "epoch": 5.73, "learning_rate": 3.4345648090169066e-05, "loss": 0.9777, "step": 6775, "task_loss": 1.2648451328277588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9653269052505493, "epoch": 5.73, "learning_rate": 3.4342517219787105e-05, "loss": 0.8803, "step": 6776, "task_loss": 1.1812392473220825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4324069321155548, "epoch": 5.73, "learning_rate": 3.4339386349405136e-05, "loss": 0.5746, "step": 6777, "task_loss": 0.08025415986776352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7819256782531738, "epoch": 5.73, "learning_rate": 3.4336255479023175e-05, "loss": 0.7542, "step": 6778, "task_loss": 0.5272654294967651 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.923900842666626, "epoch": 5.73, "learning_rate": 3.433312460864121e-05, "loss": 0.9243, "step": 6779, "task_loss": 1.3074673414230347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8262261152267456, "epoch": 5.73, "learning_rate": 3.432999373825924e-05, "loss": 0.7815, "step": 6780, "task_loss": 0.5164614319801331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7516949772834778, "epoch": 5.73, "learning_rate": 3.432686286787728e-05, "loss": 0.9954, "step": 6781, "task_loss": 0.8873908519744873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6196455955505371, "epoch": 5.73, "learning_rate": 3.432373199749531e-05, "loss": 1.0193, "step": 6782, "task_loss": 0.4144763946533203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6983757019042969, "epoch": 5.73, "learning_rate": 3.432060112711334e-05, "loss": 0.8538, "step": 6783, "task_loss": 0.5245763063430786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0143908262252808, "epoch": 5.73, "learning_rate": 3.431747025673137e-05, "loss": 0.8338, "step": 6784, "task_loss": 0.7498934864997864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9918961524963379, "epoch": 5.73, "learning_rate": 3.431433938634941e-05, "loss": 0.9881, "step": 6785, "task_loss": 0.7099630832672119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.322962999343872, "epoch": 5.74, "learning_rate": 3.431120851596744e-05, "loss": 0.9564, "step": 6786, "task_loss": 1.3148505687713623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48060452938079834, "epoch": 5.74, "learning_rate": 3.4308077645585475e-05, "loss": 0.8622, "step": 6787, "task_loss": 0.5323296189308167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9131132364273071, "epoch": 5.74, "learning_rate": 3.4304946775203506e-05, "loss": 1.0465, "step": 6788, "task_loss": 0.8599462509155273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47590377926826477, "epoch": 5.74, "learning_rate": 3.4301815904821545e-05, "loss": 0.8143, "step": 6789, "task_loss": 0.3248344957828522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9946514368057251, "epoch": 5.74, "learning_rate": 3.429868503443958e-05, "loss": 0.7961, "step": 6790, "task_loss": 1.1581246852874756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7642863988876343, "epoch": 5.74, "learning_rate": 3.429555416405761e-05, "loss": 1.1268, "step": 6791, "task_loss": 1.0411489009857178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0984902381896973, "epoch": 5.74, "learning_rate": 3.429242329367565e-05, "loss": 1.3219, "step": 6792, "task_loss": 0.9363031983375549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8410823941230774, "epoch": 5.74, "learning_rate": 3.428929242329368e-05, "loss": 0.7645, "step": 6793, "task_loss": 1.7503619194030762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1450940370559692, "epoch": 5.74, "learning_rate": 3.428616155291171e-05, "loss": 0.9637, "step": 6794, "task_loss": 0.45714008808135986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.041182518005371, "epoch": 5.74, "learning_rate": 3.428303068252974e-05, "loss": 1.0094, "step": 6795, "task_loss": 1.1956734657287598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1495860815048218, "epoch": 5.74, "learning_rate": 3.427989981214778e-05, "loss": 1.1834, "step": 6796, "task_loss": 0.811334490776062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9604222178459167, "epoch": 5.75, "learning_rate": 3.427676894176581e-05, "loss": 1.0361, "step": 6797, "task_loss": 0.7395707964897156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8974707126617432, "epoch": 5.75, "learning_rate": 3.4273638071383845e-05, "loss": 0.7757, "step": 6798, "task_loss": 0.5009844303131104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4465599060058594, "epoch": 5.75, "learning_rate": 3.4270507201001876e-05, "loss": 1.0679, "step": 6799, "task_loss": 0.7421238422393799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6422151923179626, "epoch": 5.75, "learning_rate": 3.4267376330619915e-05, "loss": 0.693, "step": 6800, "task_loss": 0.6435941457748413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8973758220672607, "epoch": 5.75, "learning_rate": 3.426424546023795e-05, "loss": 1.0538, "step": 6801, "task_loss": 1.587386131286621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6058131456375122, "epoch": 5.75, "learning_rate": 3.426111458985598e-05, "loss": 1.0564, "step": 6802, "task_loss": 1.6023509502410889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6504706740379333, "epoch": 5.75, "learning_rate": 3.425798371947401e-05, "loss": 0.9166, "step": 6803, "task_loss": 0.9015833735466003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.413360357284546, "epoch": 5.75, "learning_rate": 3.425485284909205e-05, "loss": 0.9355, "step": 6804, "task_loss": 1.0692315101623535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.954505205154419, "epoch": 5.75, "learning_rate": 3.425172197871008e-05, "loss": 0.9616, "step": 6805, "task_loss": 1.0117506980895996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9432027339935303, "epoch": 5.75, "learning_rate": 3.424859110832811e-05, "loss": 0.9077, "step": 6806, "task_loss": 1.0189402103424072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6112250089645386, "epoch": 5.75, "learning_rate": 3.424546023794615e-05, "loss": 0.7925, "step": 6807, "task_loss": 0.5768846869468689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.840883493423462, "epoch": 5.75, "learning_rate": 3.424232936756418e-05, "loss": 1.1773, "step": 6808, "task_loss": 1.0960079431533813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5906684994697571, "epoch": 5.76, "learning_rate": 3.4239198497182214e-05, "loss": 0.7356, "step": 6809, "task_loss": 1.5447579622268677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0513132810592651, "epoch": 5.76, "learning_rate": 3.423606762680025e-05, "loss": 0.9145, "step": 6810, "task_loss": 0.35399138927459717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6116812825202942, "epoch": 5.76, "learning_rate": 3.4232936756418285e-05, "loss": 0.8618, "step": 6811, "task_loss": 1.0080676078796387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6084984540939331, "epoch": 5.76, "learning_rate": 3.422980588603632e-05, "loss": 0.8087, "step": 6812, "task_loss": 1.2621889114379883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6198509335517883, "epoch": 5.76, "learning_rate": 3.4226675015654355e-05, "loss": 0.8435, "step": 6813, "task_loss": 0.7218862771987915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9866610765457153, "epoch": 5.76, "learning_rate": 3.422354414527239e-05, "loss": 0.7426, "step": 6814, "task_loss": 0.6811528205871582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0939302444458008, "epoch": 5.76, "learning_rate": 3.4220413274890425e-05, "loss": 0.8814, "step": 6815, "task_loss": 0.7771010398864746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5856218338012695, "epoch": 5.76, "learning_rate": 3.421728240450846e-05, "loss": 0.7695, "step": 6816, "task_loss": 0.9932563304901123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.111897349357605, "epoch": 5.76, "learning_rate": 3.421415153412649e-05, "loss": 1.101, "step": 6817, "task_loss": 1.9630908966064453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9088292121887207, "epoch": 5.76, "learning_rate": 3.421102066374453e-05, "loss": 1.0288, "step": 6818, "task_loss": 1.4543638229370117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9090243577957153, "epoch": 5.76, "learning_rate": 3.420788979336256e-05, "loss": 1.0903, "step": 6819, "task_loss": 1.1898716688156128 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1163665056228638, "epoch": 5.76, "learning_rate": 3.420475892298059e-05, "loss": 0.973, "step": 6820, "task_loss": 1.2270997762680054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6710965633392334, "epoch": 5.77, "learning_rate": 3.420162805259862e-05, "loss": 0.83, "step": 6821, "task_loss": 0.4136931598186493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4684794843196869, "epoch": 5.77, "learning_rate": 3.419849718221666e-05, "loss": 0.9133, "step": 6822, "task_loss": 0.35520118474960327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9962427616119385, "epoch": 5.77, "learning_rate": 3.419536631183469e-05, "loss": 1.263, "step": 6823, "task_loss": 1.0314804315567017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7658458948135376, "epoch": 5.77, "learning_rate": 3.4192235441452725e-05, "loss": 0.8929, "step": 6824, "task_loss": 0.2688702940940857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.021777868270874, "epoch": 5.77, "learning_rate": 3.418910457107076e-05, "loss": 0.9915, "step": 6825, "task_loss": 1.1615300178527832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.245121717453003, "epoch": 5.77, "learning_rate": 3.4185973700688795e-05, "loss": 1.1103, "step": 6826, "task_loss": 1.2703949213027954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4270362854003906, "epoch": 5.77, "learning_rate": 3.418284283030683e-05, "loss": 0.9767, "step": 6827, "task_loss": 1.418039083480835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2760478258132935, "epoch": 5.77, "learning_rate": 3.417971195992486e-05, "loss": 0.9907, "step": 6828, "task_loss": 0.6571986675262451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7135458588600159, "epoch": 5.77, "learning_rate": 3.41765810895429e-05, "loss": 1.0037, "step": 6829, "task_loss": 0.7373326420783997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.615248203277588, "epoch": 5.77, "learning_rate": 3.417345021916093e-05, "loss": 1.1691, "step": 6830, "task_loss": 0.8501812815666199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7608495950698853, "epoch": 5.77, "learning_rate": 3.417031934877896e-05, "loss": 0.8967, "step": 6831, "task_loss": 0.6247343420982361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.04063081741333, "epoch": 5.77, "learning_rate": 3.416718847839699e-05, "loss": 0.7798, "step": 6832, "task_loss": 1.2698545455932617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5338000059127808, "epoch": 5.78, "learning_rate": 3.416405760801503e-05, "loss": 1.0478, "step": 6833, "task_loss": 1.2023999691009521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2492692470550537, "epoch": 5.78, "learning_rate": 3.416092673763306e-05, "loss": 1.1526, "step": 6834, "task_loss": 1.365518569946289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5214076042175293, "epoch": 5.78, "learning_rate": 3.4157795867251095e-05, "loss": 0.9933, "step": 6835, "task_loss": 1.349713921546936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6935690641403198, "epoch": 5.78, "learning_rate": 3.415466499686913e-05, "loss": 1.2689, "step": 6836, "task_loss": 0.7943674921989441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.009320855140686, "epoch": 5.78, "learning_rate": 3.4151534126487165e-05, "loss": 0.7462, "step": 6837, "task_loss": 1.1938101053237915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6338495016098022, "epoch": 5.78, "learning_rate": 3.41484032561052e-05, "loss": 0.853, "step": 6838, "task_loss": 0.5158208012580872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.971611499786377, "epoch": 5.78, "learning_rate": 3.414527238572323e-05, "loss": 0.8237, "step": 6839, "task_loss": 1.6068977117538452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.359584093093872, "epoch": 5.78, "learning_rate": 3.414214151534126e-05, "loss": 1.2777, "step": 6840, "task_loss": 1.1482033729553223 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0501737594604492, "epoch": 5.78, "learning_rate": 3.41390106449593e-05, "loss": 0.8824, "step": 6841, "task_loss": 0.9465667605400085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3951951861381531, "epoch": 5.78, "learning_rate": 3.413587977457733e-05, "loss": 0.7065, "step": 6842, "task_loss": 1.1950072050094604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.594423770904541, "epoch": 5.78, "learning_rate": 3.413274890419537e-05, "loss": 1.2345, "step": 6843, "task_loss": 0.9702833294868469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1328197717666626, "epoch": 5.78, "learning_rate": 3.41296180338134e-05, "loss": 0.9268, "step": 6844, "task_loss": 1.1243171691894531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5340468287467957, "epoch": 5.79, "learning_rate": 3.412648716343143e-05, "loss": 0.8624, "step": 6845, "task_loss": 0.4273616373538971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9580963850021362, "epoch": 5.79, "learning_rate": 3.412335629304947e-05, "loss": 1.0753, "step": 6846, "task_loss": 0.26788759231567383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4972524046897888, "epoch": 5.79, "learning_rate": 3.4120225422667503e-05, "loss": 0.9109, "step": 6847, "task_loss": 0.9120998382568359 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5880687832832336, "epoch": 5.79, "learning_rate": 3.411709455228554e-05, "loss": 0.8267, "step": 6848, "task_loss": 0.9142454862594604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6654239296913147, "epoch": 5.79, "learning_rate": 3.4113963681903574e-05, "loss": 0.7429, "step": 6849, "task_loss": 1.176025152206421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0166188478469849, "epoch": 5.79, "learning_rate": 3.4110832811521606e-05, "loss": 0.9591, "step": 6850, "task_loss": 0.6287904977798462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9425041079521179, "epoch": 5.79, "learning_rate": 3.410770194113964e-05, "loss": 1.0516, "step": 6851, "task_loss": 1.3071733713150024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3912030458450317, "epoch": 5.79, "learning_rate": 3.4104571070757676e-05, "loss": 1.0111, "step": 6852, "task_loss": 0.8126278519630432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2673314809799194, "epoch": 5.79, "learning_rate": 3.410144020037571e-05, "loss": 1.1916, "step": 6853, "task_loss": 1.9539610147476196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.038001537322998, "epoch": 5.79, "learning_rate": 3.409830932999374e-05, "loss": 0.962, "step": 6854, "task_loss": 0.9244308471679688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6748457551002502, "epoch": 5.79, "learning_rate": 3.409517845961178e-05, "loss": 0.7105, "step": 6855, "task_loss": 0.7693513035774231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.236301302909851, "epoch": 5.79, "learning_rate": 3.409204758922981e-05, "loss": 0.8634, "step": 6856, "task_loss": 0.8071554899215698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5743858218193054, "epoch": 5.8, "learning_rate": 3.408891671884784e-05, "loss": 0.8862, "step": 6857, "task_loss": 0.8154314160346985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.070645809173584, "epoch": 5.8, "learning_rate": 3.408578584846587e-05, "loss": 1.0289, "step": 6858, "task_loss": 1.2735791206359863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3428841829299927, "epoch": 5.8, "learning_rate": 3.408265497808391e-05, "loss": 0.5345, "step": 6859, "task_loss": 0.6374341249465942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.141222357749939, "epoch": 5.8, "learning_rate": 3.4079524107701944e-05, "loss": 0.8658, "step": 6860, "task_loss": 1.8804285526275635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0763192176818848, "epoch": 5.8, "learning_rate": 3.4076393237319975e-05, "loss": 0.7824, "step": 6861, "task_loss": 0.6950896978378296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7557249665260315, "epoch": 5.8, "learning_rate": 3.407326236693801e-05, "loss": 0.7998, "step": 6862, "task_loss": 1.2648613452911377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7542039155960083, "epoch": 5.8, "learning_rate": 3.4070131496556046e-05, "loss": 0.7622, "step": 6863, "task_loss": 0.9475991129875183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1826690435409546, "epoch": 5.8, "learning_rate": 3.406700062617408e-05, "loss": 0.8203, "step": 6864, "task_loss": 1.0393774509429932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7192468643188477, "epoch": 5.8, "learning_rate": 3.406386975579211e-05, "loss": 0.9574, "step": 6865, "task_loss": 0.8472422361373901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9191077947616577, "epoch": 5.8, "learning_rate": 3.406073888541015e-05, "loss": 0.7947, "step": 6866, "task_loss": 1.0084340572357178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6269532442092896, "epoch": 5.8, "learning_rate": 3.405760801502818e-05, "loss": 0.8619, "step": 6867, "task_loss": 0.46473291516304016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5782675743103027, "epoch": 5.81, "learning_rate": 3.405447714464621e-05, "loss": 0.8092, "step": 6868, "task_loss": 0.5977188348770142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7057332992553711, "epoch": 5.81, "learning_rate": 3.405134627426424e-05, "loss": 0.839, "step": 6869, "task_loss": 1.0370389223098755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3104698657989502, "epoch": 5.81, "learning_rate": 3.404821540388228e-05, "loss": 0.7795, "step": 6870, "task_loss": 0.9600040912628174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7795807123184204, "epoch": 5.81, "learning_rate": 3.4045084533500314e-05, "loss": 0.7837, "step": 6871, "task_loss": 1.6263827085494995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0544183254241943, "epoch": 5.81, "learning_rate": 3.4041953663118345e-05, "loss": 0.7588, "step": 6872, "task_loss": 1.3256231546401978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4946787357330322, "epoch": 5.81, "learning_rate": 3.403882279273638e-05, "loss": 0.9356, "step": 6873, "task_loss": 0.6464933156967163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0191198587417603, "epoch": 5.81, "learning_rate": 3.4035691922354416e-05, "loss": 1.2304, "step": 6874, "task_loss": 0.2674356997013092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4595634341239929, "epoch": 5.81, "learning_rate": 3.403256105197245e-05, "loss": 0.8882, "step": 6875, "task_loss": 0.5411006808280945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2538901567459106, "epoch": 5.81, "learning_rate": 3.402943018159048e-05, "loss": 1.0324, "step": 6876, "task_loss": 0.6482666730880737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0995458364486694, "epoch": 5.81, "learning_rate": 3.402629931120852e-05, "loss": 0.9915, "step": 6877, "task_loss": 1.4579235315322876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8104075193405151, "epoch": 5.81, "learning_rate": 3.402316844082655e-05, "loss": 1.0222, "step": 6878, "task_loss": 0.8962030410766602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.515081763267517, "epoch": 5.81, "learning_rate": 3.402003757044459e-05, "loss": 1.5485, "step": 6879, "task_loss": 1.108875036239624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8721487522125244, "epoch": 5.82, "learning_rate": 3.401690670006262e-05, "loss": 1.3957, "step": 6880, "task_loss": 0.46737322211265564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0425803661346436, "epoch": 5.82, "learning_rate": 3.401377582968066e-05, "loss": 0.911, "step": 6881, "task_loss": 1.4012101888656616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4466670751571655, "epoch": 5.82, "learning_rate": 3.401064495929869e-05, "loss": 1.2219, "step": 6882, "task_loss": 1.5687122344970703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.968562126159668, "epoch": 5.82, "learning_rate": 3.400751408891672e-05, "loss": 0.776, "step": 6883, "task_loss": 0.6972676515579224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7438424825668335, "epoch": 5.82, "learning_rate": 3.4004383218534754e-05, "loss": 0.6576, "step": 6884, "task_loss": 1.001470685005188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.816953182220459, "epoch": 5.82, "learning_rate": 3.400125234815279e-05, "loss": 0.7847, "step": 6885, "task_loss": 0.25818002223968506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7609403133392334, "epoch": 5.82, "learning_rate": 3.3998121477770824e-05, "loss": 0.8793, "step": 6886, "task_loss": 0.4908868074417114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3069195747375488, "epoch": 5.82, "learning_rate": 3.3994990607388856e-05, "loss": 0.9857, "step": 6887, "task_loss": 0.9700629711151123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6917468309402466, "epoch": 5.82, "learning_rate": 3.399185973700689e-05, "loss": 0.7312, "step": 6888, "task_loss": 0.5384666919708252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8386602997779846, "epoch": 5.82, "learning_rate": 3.3988728866624926e-05, "loss": 0.8016, "step": 6889, "task_loss": 0.4348175823688507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2580536603927612, "epoch": 5.82, "learning_rate": 3.398559799624296e-05, "loss": 0.8496, "step": 6890, "task_loss": 1.4188299179077148 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5993764400482178, "epoch": 5.82, "learning_rate": 3.398246712586099e-05, "loss": 0.8375, "step": 6891, "task_loss": 0.33297666907310486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7948870062828064, "epoch": 5.83, "learning_rate": 3.397933625547903e-05, "loss": 0.7961, "step": 6892, "task_loss": 0.2505059838294983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2725639343261719, "epoch": 5.83, "learning_rate": 3.397620538509706e-05, "loss": 1.0554, "step": 6893, "task_loss": 0.9319187998771667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7564842104911804, "epoch": 5.83, "learning_rate": 3.397307451471509e-05, "loss": 0.9573, "step": 6894, "task_loss": 1.0442701578140259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9171180725097656, "epoch": 5.83, "learning_rate": 3.3969943644333124e-05, "loss": 0.8526, "step": 6895, "task_loss": 0.6920954585075378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7373217344284058, "epoch": 5.83, "learning_rate": 3.396681277395116e-05, "loss": 1.0694, "step": 6896, "task_loss": 1.4309650659561157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8659797310829163, "epoch": 5.83, "learning_rate": 3.3963681903569194e-05, "loss": 0.8737, "step": 6897, "task_loss": 1.4518369436264038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5154917240142822, "epoch": 5.83, "learning_rate": 3.3960551033187226e-05, "loss": 1.2808, "step": 6898, "task_loss": 0.6414453387260437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5040260553359985, "epoch": 5.83, "learning_rate": 3.395742016280526e-05, "loss": 0.7678, "step": 6899, "task_loss": 0.5746818780899048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.144068956375122, "epoch": 5.83, "learning_rate": 3.3954289292423296e-05, "loss": 0.951, "step": 6900, "task_loss": 1.360191822052002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6666527986526489, "epoch": 5.83, "learning_rate": 3.395115842204133e-05, "loss": 0.8821, "step": 6901, "task_loss": 1.1487798690795898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9893760681152344, "epoch": 5.83, "learning_rate": 3.394802755165936e-05, "loss": 0.9523, "step": 6902, "task_loss": 0.48197755217552185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7012921571731567, "epoch": 5.83, "learning_rate": 3.39448966812774e-05, "loss": 0.9816, "step": 6903, "task_loss": 0.9121934175491333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4635307192802429, "epoch": 5.84, "learning_rate": 3.394176581089543e-05, "loss": 0.8953, "step": 6904, "task_loss": 1.349608063697815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5223485827445984, "epoch": 5.84, "learning_rate": 3.393863494051346e-05, "loss": 0.8265, "step": 6905, "task_loss": 1.064967155456543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4845011234283447, "epoch": 5.84, "learning_rate": 3.3935504070131494e-05, "loss": 0.9164, "step": 6906, "task_loss": 1.1695574522018433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2955421209335327, "epoch": 5.84, "learning_rate": 3.393237319974953e-05, "loss": 1.0058, "step": 6907, "task_loss": 0.6141376495361328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6880213022232056, "epoch": 5.84, "learning_rate": 3.3929242329367564e-05, "loss": 0.8055, "step": 6908, "task_loss": 1.4867212772369385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6852675676345825, "epoch": 5.84, "learning_rate": 3.3926111458985596e-05, "loss": 0.9541, "step": 6909, "task_loss": 0.7258515954017639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.726584792137146, "epoch": 5.84, "learning_rate": 3.3922980588603634e-05, "loss": 1.0001, "step": 6910, "task_loss": 1.1998764276504517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0975732803344727, "epoch": 5.84, "learning_rate": 3.3919849718221666e-05, "loss": 0.8728, "step": 6911, "task_loss": 1.0396242141723633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0266728401184082, "epoch": 5.84, "learning_rate": 3.39167188478397e-05, "loss": 0.8062, "step": 6912, "task_loss": 0.7804949879646301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1047775745391846, "epoch": 5.84, "learning_rate": 3.3913587977457737e-05, "loss": 0.8363, "step": 6913, "task_loss": 1.1864643096923828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8188768029212952, "epoch": 5.84, "learning_rate": 3.391045710707577e-05, "loss": 0.7211, "step": 6914, "task_loss": 1.1286306381225586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8852542042732239, "epoch": 5.84, "learning_rate": 3.390732623669381e-05, "loss": 0.801, "step": 6915, "task_loss": 1.0500050783157349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7649552822113037, "epoch": 5.85, "learning_rate": 3.390419536631184e-05, "loss": 0.8658, "step": 6916, "task_loss": 0.5435249209403992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6489030718803406, "epoch": 5.85, "learning_rate": 3.390106449592987e-05, "loss": 0.7202, "step": 6917, "task_loss": 0.6309554576873779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0053887367248535, "epoch": 5.85, "learning_rate": 3.389793362554791e-05, "loss": 1.0222, "step": 6918, "task_loss": 1.739039421081543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5563018321990967, "epoch": 5.85, "learning_rate": 3.389480275516594e-05, "loss": 0.9521, "step": 6919, "task_loss": 0.9977294206619263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3279426097869873, "epoch": 5.85, "learning_rate": 3.389167188478397e-05, "loss": 0.9352, "step": 6920, "task_loss": 1.7805975675582886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3097933530807495, "epoch": 5.85, "learning_rate": 3.3888541014402004e-05, "loss": 0.8942, "step": 6921, "task_loss": 1.011176586151123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4297526478767395, "epoch": 5.85, "learning_rate": 3.388541014402004e-05, "loss": 0.7953, "step": 6922, "task_loss": 0.5049508213996887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8888627290725708, "epoch": 5.85, "learning_rate": 3.3882279273638075e-05, "loss": 0.7726, "step": 6923, "task_loss": 0.32750824093818665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5377997159957886, "epoch": 5.85, "learning_rate": 3.3879148403256106e-05, "loss": 0.8999, "step": 6924, "task_loss": 0.4693158268928528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8020017743110657, "epoch": 5.85, "learning_rate": 3.387601753287414e-05, "loss": 0.8197, "step": 6925, "task_loss": 0.2112639844417572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6258788704872131, "epoch": 5.85, "learning_rate": 3.387288666249218e-05, "loss": 1.1205, "step": 6926, "task_loss": 0.6032505631446838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.065102219581604, "epoch": 5.85, "learning_rate": 3.386975579211021e-05, "loss": 0.9574, "step": 6927, "task_loss": 0.8468900322914124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0820438861846924, "epoch": 5.86, "learning_rate": 3.386662492172824e-05, "loss": 1.0097, "step": 6928, "task_loss": 0.8839014172554016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.892052412033081, "epoch": 5.86, "learning_rate": 3.386349405134628e-05, "loss": 0.8887, "step": 6929, "task_loss": 1.2531615495681763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6819164752960205, "epoch": 5.86, "learning_rate": 3.386036318096431e-05, "loss": 0.9909, "step": 6930, "task_loss": 1.3475241661071777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7486989498138428, "epoch": 5.86, "learning_rate": 3.385723231058234e-05, "loss": 0.8757, "step": 6931, "task_loss": 1.4608715772628784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4066157341003418, "epoch": 5.86, "learning_rate": 3.3854101440200374e-05, "loss": 1.1617, "step": 6932, "task_loss": 1.7208361625671387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6046029329299927, "epoch": 5.86, "learning_rate": 3.385097056981841e-05, "loss": 0.6236, "step": 6933, "task_loss": 0.7152430415153503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6018414497375488, "epoch": 5.86, "learning_rate": 3.3847839699436445e-05, "loss": 0.8526, "step": 6934, "task_loss": 0.6856856942176819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.168152093887329, "epoch": 5.86, "learning_rate": 3.3844708829054476e-05, "loss": 0.7882, "step": 6935, "task_loss": 2.0066795349121094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.482743263244629, "epoch": 5.86, "learning_rate": 3.384157795867251e-05, "loss": 1.1237, "step": 6936, "task_loss": 0.7621709704399109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.797637939453125, "epoch": 5.86, "learning_rate": 3.383844708829055e-05, "loss": 0.923, "step": 6937, "task_loss": 0.8066670894622803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.979820728302002, "epoch": 5.86, "learning_rate": 3.383531621790858e-05, "loss": 0.9065, "step": 6938, "task_loss": 0.2894704043865204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0127232074737549, "epoch": 5.87, "learning_rate": 3.383218534752661e-05, "loss": 0.8139, "step": 6939, "task_loss": 0.8275284171104431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.647672712802887, "epoch": 5.87, "learning_rate": 3.382905447714465e-05, "loss": 0.9558, "step": 6940, "task_loss": 0.8882201910018921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9000583291053772, "epoch": 5.87, "learning_rate": 3.382592360676268e-05, "loss": 0.8457, "step": 6941, "task_loss": 0.6951413750648499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7572883367538452, "epoch": 5.87, "learning_rate": 3.382279273638071e-05, "loss": 0.7004, "step": 6942, "task_loss": 0.37907737493515015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5620328783988953, "epoch": 5.87, "learning_rate": 3.3819661865998744e-05, "loss": 1.0055, "step": 6943, "task_loss": 1.1281148195266724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9190694093704224, "epoch": 5.87, "learning_rate": 3.381653099561678e-05, "loss": 0.922, "step": 6944, "task_loss": 0.4500877559185028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6308400630950928, "epoch": 5.87, "learning_rate": 3.3813400125234815e-05, "loss": 0.8812, "step": 6945, "task_loss": 1.0601125955581665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7859493494033813, "epoch": 5.87, "learning_rate": 3.381026925485285e-05, "loss": 0.8406, "step": 6946, "task_loss": 0.327406108379364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6470530033111572, "epoch": 5.87, "learning_rate": 3.3807138384470885e-05, "loss": 0.7739, "step": 6947, "task_loss": 0.6219868659973145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9917117357254028, "epoch": 5.87, "learning_rate": 3.3804007514088923e-05, "loss": 0.7731, "step": 6948, "task_loss": 1.9097692966461182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6216516494750977, "epoch": 5.87, "learning_rate": 3.3800876643706955e-05, "loss": 0.705, "step": 6949, "task_loss": 0.751078724861145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7703490853309631, "epoch": 5.87, "learning_rate": 3.379774577332499e-05, "loss": 1.0291, "step": 6950, "task_loss": 0.9537156224250793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0145093202590942, "epoch": 5.88, "learning_rate": 3.379461490294302e-05, "loss": 1.1315, "step": 6951, "task_loss": 1.2060378789901733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7957545518875122, "epoch": 5.88, "learning_rate": 3.379148403256106e-05, "loss": 0.9805, "step": 6952, "task_loss": 0.7363992929458618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.097123384475708, "epoch": 5.88, "learning_rate": 3.378835316217909e-05, "loss": 0.7294, "step": 6953, "task_loss": 1.7358918190002441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8580490350723267, "epoch": 5.88, "learning_rate": 3.378522229179712e-05, "loss": 0.9497, "step": 6954, "task_loss": 0.32238245010375977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8695194721221924, "epoch": 5.88, "learning_rate": 3.378209142141516e-05, "loss": 0.9058, "step": 6955, "task_loss": 0.5766901969909668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4134443998336792, "epoch": 5.88, "learning_rate": 3.377896055103319e-05, "loss": 1.2652, "step": 6956, "task_loss": 1.4127092361450195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1398154497146606, "epoch": 5.88, "learning_rate": 3.377582968065122e-05, "loss": 1.0596, "step": 6957, "task_loss": 1.115541696548462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5135980248451233, "epoch": 5.88, "learning_rate": 3.3772698810269255e-05, "loss": 0.8607, "step": 6958, "task_loss": 0.5383583903312683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.403607726097107, "epoch": 5.88, "learning_rate": 3.376956793988729e-05, "loss": 1.0361, "step": 6959, "task_loss": 1.283463954925537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.684861958026886, "epoch": 5.88, "learning_rate": 3.3766437069505325e-05, "loss": 0.7722, "step": 6960, "task_loss": 0.6889323592185974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0562732219696045, "epoch": 5.88, "learning_rate": 3.376330619912336e-05, "loss": 0.9859, "step": 6961, "task_loss": 1.106767177581787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6342775821685791, "epoch": 5.88, "learning_rate": 3.376017532874139e-05, "loss": 0.7703, "step": 6962, "task_loss": 0.323528528213501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7526289224624634, "epoch": 5.89, "learning_rate": 3.375704445835943e-05, "loss": 0.8064, "step": 6963, "task_loss": 1.1200135946273804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8584944009780884, "epoch": 5.89, "learning_rate": 3.375391358797746e-05, "loss": 0.7038, "step": 6964, "task_loss": 0.8151698708534241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5969104766845703, "epoch": 5.89, "learning_rate": 3.375078271759549e-05, "loss": 0.94, "step": 6965, "task_loss": 0.5064789652824402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35814568400382996, "epoch": 5.89, "learning_rate": 3.374765184721353e-05, "loss": 0.9584, "step": 6966, "task_loss": 0.532548189163208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9311122298240662, "epoch": 5.89, "learning_rate": 3.374452097683156e-05, "loss": 1.1112, "step": 6967, "task_loss": 1.0992355346679688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45583125948905945, "epoch": 5.89, "learning_rate": 3.374139010644959e-05, "loss": 0.8817, "step": 6968, "task_loss": 0.690330445766449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1459267139434814, "epoch": 5.89, "learning_rate": 3.3738259236067625e-05, "loss": 1.1938, "step": 6969, "task_loss": 1.6711328029632568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49957916140556335, "epoch": 5.89, "learning_rate": 3.373512836568566e-05, "loss": 0.8122, "step": 6970, "task_loss": 0.3798987567424774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1093542575836182, "epoch": 5.89, "learning_rate": 3.3731997495303695e-05, "loss": 1.048, "step": 6971, "task_loss": 0.592887818813324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9689028263092041, "epoch": 5.89, "learning_rate": 3.372886662492173e-05, "loss": 0.7901, "step": 6972, "task_loss": 0.9549815654754639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8956030607223511, "epoch": 5.89, "learning_rate": 3.372573575453976e-05, "loss": 0.7514, "step": 6973, "task_loss": 1.5284082889556885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2454735040664673, "epoch": 5.89, "learning_rate": 3.37226048841578e-05, "loss": 1.0115, "step": 6974, "task_loss": 1.6454213857650757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5085554122924805, "epoch": 5.9, "learning_rate": 3.371947401377583e-05, "loss": 1.0599, "step": 6975, "task_loss": 1.7485417127609253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8349713087081909, "epoch": 5.9, "learning_rate": 3.371634314339386e-05, "loss": 0.9457, "step": 6976, "task_loss": 0.7349706888198853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6365257501602173, "epoch": 5.9, "learning_rate": 3.37132122730119e-05, "loss": 0.916, "step": 6977, "task_loss": 0.6570352911949158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6128274202346802, "epoch": 5.9, "learning_rate": 3.371008140262993e-05, "loss": 0.9637, "step": 6978, "task_loss": 1.1949692964553833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8825644254684448, "epoch": 5.9, "learning_rate": 3.370695053224796e-05, "loss": 1.0096, "step": 6979, "task_loss": 0.7607545852661133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9415459632873535, "epoch": 5.9, "learning_rate": 3.3703819661866e-05, "loss": 1.0202, "step": 6980, "task_loss": 1.0818666219711304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6278276443481445, "epoch": 5.9, "learning_rate": 3.370068879148403e-05, "loss": 0.7911, "step": 6981, "task_loss": 1.9155120849609375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6858258247375488, "epoch": 5.9, "learning_rate": 3.369755792110207e-05, "loss": 0.7891, "step": 6982, "task_loss": 0.4791381359100342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7624328136444092, "epoch": 5.9, "learning_rate": 3.3694427050720104e-05, "loss": 1.089, "step": 6983, "task_loss": 1.1145966053009033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7897566556930542, "epoch": 5.9, "learning_rate": 3.3691296180338135e-05, "loss": 0.9009, "step": 6984, "task_loss": 1.169705867767334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.917468249797821, "epoch": 5.9, "learning_rate": 3.3688165309956174e-05, "loss": 0.6691, "step": 6985, "task_loss": 0.6427200436592102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7485520839691162, "epoch": 5.9, "learning_rate": 3.3685034439574206e-05, "loss": 0.9433, "step": 6986, "task_loss": 1.2266640663146973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6695336699485779, "epoch": 5.91, "learning_rate": 3.368190356919224e-05, "loss": 0.8212, "step": 6987, "task_loss": 0.9373969435691833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0041837692260742, "epoch": 5.91, "learning_rate": 3.3678772698810276e-05, "loss": 0.8565, "step": 6988, "task_loss": 1.3345056772232056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9342938661575317, "epoch": 5.91, "learning_rate": 3.367564182842831e-05, "loss": 0.7835, "step": 6989, "task_loss": 0.5277103781700134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9468790888786316, "epoch": 5.91, "learning_rate": 3.367251095804634e-05, "loss": 0.9617, "step": 6990, "task_loss": 2.1691794395446777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.470137596130371, "epoch": 5.91, "learning_rate": 3.366938008766437e-05, "loss": 0.9962, "step": 6991, "task_loss": 1.9285677671432495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1503443717956543, "epoch": 5.91, "learning_rate": 3.366624921728241e-05, "loss": 1.1732, "step": 6992, "task_loss": 1.8206219673156738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8682222962379456, "epoch": 5.91, "learning_rate": 3.366311834690044e-05, "loss": 0.9261, "step": 6993, "task_loss": 0.9568256735801697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7670912742614746, "epoch": 5.91, "learning_rate": 3.3659987476518473e-05, "loss": 1.2775, "step": 6994, "task_loss": 1.870788335800171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.974036693572998, "epoch": 5.91, "learning_rate": 3.3656856606136505e-05, "loss": 0.9542, "step": 6995, "task_loss": 0.48656201362609863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5446070432662964, "epoch": 5.91, "learning_rate": 3.3653725735754544e-05, "loss": 0.9941, "step": 6996, "task_loss": 0.8594799041748047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.669047474861145, "epoch": 5.91, "learning_rate": 3.3650594865372576e-05, "loss": 0.9497, "step": 6997, "task_loss": 0.35954445600509644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.209742546081543, "epoch": 5.91, "learning_rate": 3.364746399499061e-05, "loss": 1.0117, "step": 6998, "task_loss": 1.690645694732666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0742379426956177, "epoch": 5.92, "learning_rate": 3.364433312460864e-05, "loss": 0.8424, "step": 6999, "task_loss": 1.1364433765411377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5611460208892822, "epoch": 5.92, "learning_rate": 3.364120225422668e-05, "loss": 0.7199, "step": 7000, "task_loss": 0.4880939722061157 }, { "epoch": 5.92, "eval_accuracy": 0.8769108910891089, "eval_loss": 0.5619150400161743, "eval_runtime": 208.304, "eval_samples_per_second": 121.217, "eval_steps_per_second": 0.951, "step": 7000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8906183242797852, "epoch": 5.92, "learning_rate": 3.363807138384471e-05, "loss": 0.7704, "step": 7001, "task_loss": 1.2520197629928589 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4863299131393433, "epoch": 5.92, "learning_rate": 3.363494051346274e-05, "loss": 0.9069, "step": 7002, "task_loss": 1.3162511587142944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7060438394546509, "epoch": 5.92, "learning_rate": 3.363180964308078e-05, "loss": 0.8722, "step": 7003, "task_loss": 0.5009607672691345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0671329498291016, "epoch": 5.92, "learning_rate": 3.362867877269881e-05, "loss": 0.9496, "step": 7004, "task_loss": 1.6627230644226074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0057008266448975, "epoch": 5.92, "learning_rate": 3.362554790231684e-05, "loss": 1.012, "step": 7005, "task_loss": 0.4154372811317444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5751818418502808, "epoch": 5.92, "learning_rate": 3.3622417031934875e-05, "loss": 0.7194, "step": 7006, "task_loss": 0.13957597315311432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8671864867210388, "epoch": 5.92, "learning_rate": 3.3619286161552914e-05, "loss": 0.8025, "step": 7007, "task_loss": 0.9912954568862915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.068942904472351, "epoch": 5.92, "learning_rate": 3.3616155291170945e-05, "loss": 0.8545, "step": 7008, "task_loss": 1.4549864530563354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8456752300262451, "epoch": 5.92, "learning_rate": 3.361302442078898e-05, "loss": 0.8988, "step": 7009, "task_loss": 1.2727543115615845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9132617115974426, "epoch": 5.93, "learning_rate": 3.360989355040701e-05, "loss": 0.788, "step": 7010, "task_loss": 1.027687668800354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0513267517089844, "epoch": 5.93, "learning_rate": 3.360676268002505e-05, "loss": 0.7134, "step": 7011, "task_loss": 1.333055853843689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6172952651977539, "epoch": 5.93, "learning_rate": 3.360363180964308e-05, "loss": 0.7945, "step": 7012, "task_loss": 0.5341715812683105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9521008729934692, "epoch": 5.93, "learning_rate": 3.360050093926112e-05, "loss": 0.7429, "step": 7013, "task_loss": 0.696173906326294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0868479013442993, "epoch": 5.93, "learning_rate": 3.359737006887915e-05, "loss": 1.0702, "step": 7014, "task_loss": 0.9238091707229614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0579335689544678, "epoch": 5.93, "learning_rate": 3.359423919849719e-05, "loss": 1.2167, "step": 7015, "task_loss": 1.3333156108856201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7053456902503967, "epoch": 5.93, "learning_rate": 3.359110832811522e-05, "loss": 0.9526, "step": 7016, "task_loss": 1.3505308628082275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5673398971557617, "epoch": 5.93, "learning_rate": 3.358797745773325e-05, "loss": 0.7713, "step": 7017, "task_loss": 0.6700510382652283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.035088300704956, "epoch": 5.93, "learning_rate": 3.358484658735129e-05, "loss": 0.9676, "step": 7018, "task_loss": 0.4453567564487457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4665793776512146, "epoch": 5.93, "learning_rate": 3.358171571696932e-05, "loss": 0.7533, "step": 7019, "task_loss": 0.3984585702419281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.33893620967865, "epoch": 5.93, "learning_rate": 3.3578584846587354e-05, "loss": 0.9096, "step": 7020, "task_loss": 1.0983123779296875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3163965940475464, "epoch": 5.93, "learning_rate": 3.3575453976205386e-05, "loss": 1.0225, "step": 7021, "task_loss": 1.781326413154602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1348402500152588, "epoch": 5.94, "learning_rate": 3.3572323105823424e-05, "loss": 1.0579, "step": 7022, "task_loss": 1.2585880756378174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6900599002838135, "epoch": 5.94, "learning_rate": 3.3569192235441456e-05, "loss": 0.7206, "step": 7023, "task_loss": 1.0540709495544434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.431294322013855, "epoch": 5.94, "learning_rate": 3.356606136505949e-05, "loss": 0.7273, "step": 7024, "task_loss": 0.6996491551399231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9741184115409851, "epoch": 5.94, "learning_rate": 3.3562930494677526e-05, "loss": 0.9428, "step": 7025, "task_loss": 0.8213583827018738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5386438369750977, "epoch": 5.94, "learning_rate": 3.355979962429556e-05, "loss": 1.0465, "step": 7026, "task_loss": 0.7210488319396973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8500163555145264, "epoch": 5.94, "learning_rate": 3.355666875391359e-05, "loss": 0.8687, "step": 7027, "task_loss": 1.0407205820083618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5496389269828796, "epoch": 5.94, "learning_rate": 3.355353788353162e-05, "loss": 0.7468, "step": 7028, "task_loss": 0.5562518835067749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8758108615875244, "epoch": 5.94, "learning_rate": 3.355040701314966e-05, "loss": 0.7124, "step": 7029, "task_loss": 1.5659304857254028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39644455909729004, "epoch": 5.94, "learning_rate": 3.354727614276769e-05, "loss": 0.575, "step": 7030, "task_loss": 0.3709390163421631 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7413907051086426, "epoch": 5.94, "learning_rate": 3.3544145272385724e-05, "loss": 0.863, "step": 7031, "task_loss": 1.6617825031280518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1032668352127075, "epoch": 5.94, "learning_rate": 3.3541014402003756e-05, "loss": 1.0913, "step": 7032, "task_loss": 1.2655178308486938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3417713642120361, "epoch": 5.94, "learning_rate": 3.3537883531621794e-05, "loss": 1.0807, "step": 7033, "task_loss": 1.5017151832580566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2964028120040894, "epoch": 5.95, "learning_rate": 3.3534752661239826e-05, "loss": 0.7395, "step": 7034, "task_loss": 1.2132079601287842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5822587609291077, "epoch": 5.95, "learning_rate": 3.353162179085786e-05, "loss": 0.8753, "step": 7035, "task_loss": 0.6125070452690125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5543727278709412, "epoch": 5.95, "learning_rate": 3.352849092047589e-05, "loss": 0.6868, "step": 7036, "task_loss": 0.6453002095222473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3028368949890137, "epoch": 5.95, "learning_rate": 3.352536005009393e-05, "loss": 1.2008, "step": 7037, "task_loss": 1.3286495208740234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7343938946723938, "epoch": 5.95, "learning_rate": 3.352222917971196e-05, "loss": 0.7103, "step": 7038, "task_loss": 1.2269842624664307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0263383388519287, "epoch": 5.95, "learning_rate": 3.351909830932999e-05, "loss": 0.8294, "step": 7039, "task_loss": 2.6456289291381836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8525549173355103, "epoch": 5.95, "learning_rate": 3.351596743894803e-05, "loss": 0.8707, "step": 7040, "task_loss": 0.41463392972946167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7448689937591553, "epoch": 5.95, "learning_rate": 3.351283656856606e-05, "loss": 0.7213, "step": 7041, "task_loss": 0.3309512734413147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8351993560791016, "epoch": 5.95, "learning_rate": 3.3509705698184094e-05, "loss": 0.8594, "step": 7042, "task_loss": 0.40957143902778625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5271467566490173, "epoch": 5.95, "learning_rate": 3.3506574827802126e-05, "loss": 0.8178, "step": 7043, "task_loss": 1.0805801153182983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0209192037582397, "epoch": 5.95, "learning_rate": 3.3503443957420164e-05, "loss": 0.8927, "step": 7044, "task_loss": 1.3757762908935547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6806366443634033, "epoch": 5.95, "learning_rate": 3.3500313087038196e-05, "loss": 0.6832, "step": 7045, "task_loss": 1.0204750299453735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6285085082054138, "epoch": 5.96, "learning_rate": 3.349718221665623e-05, "loss": 0.7491, "step": 7046, "task_loss": 0.7429376244544983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.617598295211792, "epoch": 5.96, "learning_rate": 3.3494051346274266e-05, "loss": 0.7873, "step": 7047, "task_loss": 0.26115840673446655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4596919119358063, "epoch": 5.96, "learning_rate": 3.34909204758923e-05, "loss": 0.8253, "step": 7048, "task_loss": 0.880122184753418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7382920980453491, "epoch": 5.96, "learning_rate": 3.3487789605510337e-05, "loss": 0.8321, "step": 7049, "task_loss": 0.6393198370933533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2226290702819824, "epoch": 5.96, "learning_rate": 3.348465873512837e-05, "loss": 0.9264, "step": 7050, "task_loss": 2.237197160720825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4245729446411133, "epoch": 5.96, "learning_rate": 3.348152786474641e-05, "loss": 0.8228, "step": 7051, "task_loss": 0.3740636110305786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45692163705825806, "epoch": 5.96, "learning_rate": 3.347839699436444e-05, "loss": 0.8283, "step": 7052, "task_loss": 0.42779892683029175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8256547451019287, "epoch": 5.96, "learning_rate": 3.347526612398247e-05, "loss": 0.82, "step": 7053, "task_loss": 0.6906659603118896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.395383358001709, "epoch": 5.96, "learning_rate": 3.34721352536005e-05, "loss": 1.2477, "step": 7054, "task_loss": 1.416027545928955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6091088056564331, "epoch": 5.96, "learning_rate": 3.346900438321854e-05, "loss": 1.015, "step": 7055, "task_loss": 0.4617844820022583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6917176246643066, "epoch": 5.96, "learning_rate": 3.346587351283657e-05, "loss": 1.3806, "step": 7056, "task_loss": 0.29632890224456787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3845575451850891, "epoch": 5.96, "learning_rate": 3.3462742642454604e-05, "loss": 0.9016, "step": 7057, "task_loss": 0.5488486289978027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.739377498626709, "epoch": 5.97, "learning_rate": 3.3459611772072636e-05, "loss": 0.9424, "step": 7058, "task_loss": 1.1865136623382568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.236553430557251, "epoch": 5.97, "learning_rate": 3.3456480901690675e-05, "loss": 0.9718, "step": 7059, "task_loss": 0.9009526371955872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8758556246757507, "epoch": 5.97, "learning_rate": 3.3453350031308707e-05, "loss": 0.7943, "step": 7060, "task_loss": 1.385522484779358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.9400608539581299, "epoch": 5.97, "learning_rate": 3.345021916092674e-05, "loss": 1.1545, "step": 7061, "task_loss": 1.414088249206543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9217976331710815, "epoch": 5.97, "learning_rate": 3.344708829054478e-05, "loss": 1.0379, "step": 7062, "task_loss": 0.6501072645187378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48230910301208496, "epoch": 5.97, "learning_rate": 3.344395742016281e-05, "loss": 0.7876, "step": 7063, "task_loss": 0.5304593443870544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9775518178939819, "epoch": 5.97, "learning_rate": 3.344082654978084e-05, "loss": 0.7657, "step": 7064, "task_loss": 0.9181693196296692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6108933091163635, "epoch": 5.97, "learning_rate": 3.343769567939887e-05, "loss": 0.8019, "step": 7065, "task_loss": 0.9651261568069458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8108817934989929, "epoch": 5.97, "learning_rate": 3.343456480901691e-05, "loss": 0.8267, "step": 7066, "task_loss": 0.5517512559890747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2069931030273438, "epoch": 5.97, "learning_rate": 3.343143393863494e-05, "loss": 0.8986, "step": 7067, "task_loss": 1.2029309272766113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5490375757217407, "epoch": 5.97, "learning_rate": 3.3428303068252974e-05, "loss": 0.9319, "step": 7068, "task_loss": 1.4454314708709717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5749841928482056, "epoch": 5.97, "learning_rate": 3.3425172197871006e-05, "loss": 0.9857, "step": 7069, "task_loss": 1.0127569437026978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4807184040546417, "epoch": 5.98, "learning_rate": 3.3422041327489045e-05, "loss": 0.7692, "step": 7070, "task_loss": 0.5612149834632874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1538658142089844, "epoch": 5.98, "learning_rate": 3.3418910457107076e-05, "loss": 1.1191, "step": 7071, "task_loss": 1.9996652603149414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5448070764541626, "epoch": 5.98, "learning_rate": 3.341577958672511e-05, "loss": 0.8016, "step": 7072, "task_loss": 1.1052345037460327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9874866008758545, "epoch": 5.98, "learning_rate": 3.341264871634314e-05, "loss": 0.9853, "step": 7073, "task_loss": 0.9079083204269409 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5054412484169006, "epoch": 5.98, "learning_rate": 3.340951784596118e-05, "loss": 0.9837, "step": 7074, "task_loss": 0.4206732511520386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7025010585784912, "epoch": 5.98, "learning_rate": 3.340638697557921e-05, "loss": 0.899, "step": 7075, "task_loss": 0.41056159138679504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.775099515914917, "epoch": 5.98, "learning_rate": 3.340325610519724e-05, "loss": 0.7523, "step": 7076, "task_loss": 0.6254719495773315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9208609461784363, "epoch": 5.98, "learning_rate": 3.340012523481528e-05, "loss": 0.88, "step": 7077, "task_loss": 0.35189077258110046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39377376437187195, "epoch": 5.98, "learning_rate": 3.339699436443331e-05, "loss": 0.6568, "step": 7078, "task_loss": 0.14919373393058777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1851025819778442, "epoch": 5.98, "learning_rate": 3.3393863494051344e-05, "loss": 0.8418, "step": 7079, "task_loss": 2.042318820953369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5039405822753906, "epoch": 5.98, "learning_rate": 3.339073262366938e-05, "loss": 0.8211, "step": 7080, "task_loss": 0.7742667198181152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5731079578399658, "epoch": 5.99, "learning_rate": 3.3387601753287415e-05, "loss": 1.1985, "step": 7081, "task_loss": 0.8183073401451111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5645502209663391, "epoch": 5.99, "learning_rate": 3.338447088290545e-05, "loss": 0.7224, "step": 7082, "task_loss": 0.9291947484016418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.986526608467102, "epoch": 5.99, "learning_rate": 3.3381340012523485e-05, "loss": 1.1815, "step": 7083, "task_loss": 1.024829626083374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2367472648620605, "epoch": 5.99, "learning_rate": 3.337820914214152e-05, "loss": 1.1401, "step": 7084, "task_loss": 1.4710835218429565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5398554801940918, "epoch": 5.99, "learning_rate": 3.3375078271759555e-05, "loss": 0.7025, "step": 7085, "task_loss": 1.035278081893921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8989774584770203, "epoch": 5.99, "learning_rate": 3.337194740137759e-05, "loss": 0.8415, "step": 7086, "task_loss": 0.23946191370487213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7800000905990601, "epoch": 5.99, "learning_rate": 3.336881653099562e-05, "loss": 0.7552, "step": 7087, "task_loss": 0.8897056579589844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7641382217407227, "epoch": 5.99, "learning_rate": 3.336568566061366e-05, "loss": 0.6771, "step": 7088, "task_loss": 0.7020187377929688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.482863187789917, "epoch": 5.99, "learning_rate": 3.336255479023169e-05, "loss": 0.8798, "step": 7089, "task_loss": 2.1682565212249756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.010953664779663, "epoch": 5.99, "learning_rate": 3.335942391984972e-05, "loss": 0.6976, "step": 7090, "task_loss": 0.5193395614624023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9101821184158325, "epoch": 5.99, "learning_rate": 3.335629304946775e-05, "loss": 0.9579, "step": 7091, "task_loss": 1.1549477577209473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7038790583610535, "epoch": 5.99, "learning_rate": 3.335316217908579e-05, "loss": 0.7601, "step": 7092, "task_loss": 1.8974837064743042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6320763826370239, "epoch": 6.0, "learning_rate": 3.335003130870382e-05, "loss": 0.8569, "step": 7093, "task_loss": 0.3574194014072418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1798663139343262, "epoch": 6.0, "learning_rate": 3.3346900438321855e-05, "loss": 0.8561, "step": 7094, "task_loss": 1.789441704750061 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7975437641143799, "epoch": 6.0, "learning_rate": 3.334376956793989e-05, "loss": 1.0391, "step": 7095, "task_loss": 0.9586820006370544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7537767291069031, "epoch": 6.0, "learning_rate": 3.3340638697557925e-05, "loss": 0.8774, "step": 7096, "task_loss": 1.2718219757080078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7234456539154053, "epoch": 6.0, "learning_rate": 3.333750782717596e-05, "loss": 0.9765, "step": 7097, "task_loss": 0.559874415397644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2088359594345093, "epoch": 6.0, "learning_rate": 3.333437695679399e-05, "loss": 1.0003, "step": 7098, "task_loss": 2.1032533645629883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.679758608341217, "epoch": 6.0, "learning_rate": 3.333124608641203e-05, "loss": 1.4004, "step": 7099, "task_loss": 0.47318151593208313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1021302938461304, "epoch": 6.0, "learning_rate": 3.332811521603006e-05, "loss": 0.8617, "step": 7100, "task_loss": 0.9095768928527832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3967905044555664, "epoch": 6.0, "learning_rate": 3.332498434564809e-05, "loss": 0.8241, "step": 7101, "task_loss": 1.4700758457183838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.654239296913147, "epoch": 6.0, "learning_rate": 3.332185347526612e-05, "loss": 0.7002, "step": 7102, "task_loss": 0.4470694065093994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5954717397689819, "epoch": 6.0, "learning_rate": 3.331872260488416e-05, "loss": 0.7775, "step": 7103, "task_loss": 0.7401841282844543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.126165747642517, "epoch": 6.01, "learning_rate": 3.331559173450219e-05, "loss": 1.076, "step": 7104, "task_loss": 1.0465149879455566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6140373349189758, "epoch": 6.01, "learning_rate": 3.3312460864120225e-05, "loss": 0.8377, "step": 7105, "task_loss": 0.2588038444519043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7610269784927368, "epoch": 6.01, "learning_rate": 3.3309329993738257e-05, "loss": 0.8052, "step": 7106, "task_loss": 0.9429497718811035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1553045511245728, "epoch": 6.01, "learning_rate": 3.3306199123356295e-05, "loss": 0.9032, "step": 7107, "task_loss": 1.0622531175613403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0097765922546387, "epoch": 6.01, "learning_rate": 3.330306825297433e-05, "loss": 0.7894, "step": 7108, "task_loss": 1.038678526878357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.681251049041748, "epoch": 6.01, "learning_rate": 3.329993738259236e-05, "loss": 0.6623, "step": 7109, "task_loss": 1.0179443359375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7760499715805054, "epoch": 6.01, "learning_rate": 3.329680651221039e-05, "loss": 0.7031, "step": 7110, "task_loss": 0.19899749755859375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6570858955383301, "epoch": 6.01, "learning_rate": 3.329367564182843e-05, "loss": 1.0014, "step": 7111, "task_loss": 0.6964564323425293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.674100399017334, "epoch": 6.01, "learning_rate": 3.329054477144646e-05, "loss": 0.9504, "step": 7112, "task_loss": 0.9719404578208923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.102778673171997, "epoch": 6.01, "learning_rate": 3.328741390106449e-05, "loss": 0.819, "step": 7113, "task_loss": 1.629730463027954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4875057637691498, "epoch": 6.01, "learning_rate": 3.328428303068253e-05, "loss": 0.5475, "step": 7114, "task_loss": 0.26268017292022705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8423915505409241, "epoch": 6.01, "learning_rate": 3.328115216030056e-05, "loss": 0.7591, "step": 7115, "task_loss": 0.85528165102005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.543428361415863, "epoch": 6.02, "learning_rate": 3.32780212899186e-05, "loss": 0.8037, "step": 7116, "task_loss": 0.27379533648490906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8035752773284912, "epoch": 6.02, "learning_rate": 3.327489041953663e-05, "loss": 0.6155, "step": 7117, "task_loss": 0.7575610280036926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5280598402023315, "epoch": 6.02, "learning_rate": 3.327175954915467e-05, "loss": 0.9449, "step": 7118, "task_loss": 0.4941561818122864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1316498517990112, "epoch": 6.02, "learning_rate": 3.3268628678772704e-05, "loss": 0.7994, "step": 7119, "task_loss": 1.7000086307525635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7515990138053894, "epoch": 6.02, "learning_rate": 3.3265497808390735e-05, "loss": 0.8322, "step": 7120, "task_loss": 1.2276591062545776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0779197216033936, "epoch": 6.02, "learning_rate": 3.326236693800877e-05, "loss": 1.0153, "step": 7121, "task_loss": 0.8753231763839722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5904220938682556, "epoch": 6.02, "learning_rate": 3.3259236067626806e-05, "loss": 0.7407, "step": 7122, "task_loss": 0.4869740605354309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9736806154251099, "epoch": 6.02, "learning_rate": 3.325610519724484e-05, "loss": 0.9283, "step": 7123, "task_loss": 0.5570206046104431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8651782274246216, "epoch": 6.02, "learning_rate": 3.325297432686287e-05, "loss": 0.7891, "step": 7124, "task_loss": 0.4614850878715515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.969072163105011, "epoch": 6.02, "learning_rate": 3.324984345648091e-05, "loss": 0.9104, "step": 7125, "task_loss": 1.2380605936050415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.883641242980957, "epoch": 6.02, "learning_rate": 3.324671258609894e-05, "loss": 0.8143, "step": 7126, "task_loss": 1.1405855417251587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0118870735168457, "epoch": 6.02, "learning_rate": 3.324358171571697e-05, "loss": 0.8951, "step": 7127, "task_loss": 0.4603268504142761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9792618751525879, "epoch": 6.03, "learning_rate": 3.3240450845335e-05, "loss": 0.9961, "step": 7128, "task_loss": 1.0825004577636719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8720901012420654, "epoch": 6.03, "learning_rate": 3.323731997495304e-05, "loss": 0.8776, "step": 7129, "task_loss": 0.8366198539733887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4043658971786499, "epoch": 6.03, "learning_rate": 3.3234189104571074e-05, "loss": 0.8267, "step": 7130, "task_loss": 0.554754912853241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0256009101867676, "epoch": 6.03, "learning_rate": 3.3231058234189105e-05, "loss": 1.1102, "step": 7131, "task_loss": 0.740744411945343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0878318548202515, "epoch": 6.03, "learning_rate": 3.322792736380714e-05, "loss": 0.8517, "step": 7132, "task_loss": 0.5452767610549927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8231265544891357, "epoch": 6.03, "learning_rate": 3.3224796493425176e-05, "loss": 0.8808, "step": 7133, "task_loss": 0.6466094851493835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0914666652679443, "epoch": 6.03, "learning_rate": 3.322166562304321e-05, "loss": 0.7846, "step": 7134, "task_loss": 0.7787254452705383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1151008605957031, "epoch": 6.03, "learning_rate": 3.321853475266124e-05, "loss": 0.8814, "step": 7135, "task_loss": 1.4832638502120972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.813103437423706, "epoch": 6.03, "learning_rate": 3.321540388227928e-05, "loss": 0.7374, "step": 7136, "task_loss": 0.3174043595790863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.910874605178833, "epoch": 6.03, "learning_rate": 3.321227301189731e-05, "loss": 0.8363, "step": 7137, "task_loss": 0.7693797945976257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9142129421234131, "epoch": 6.03, "learning_rate": 3.320914214151534e-05, "loss": 0.9324, "step": 7138, "task_loss": 0.8982295393943787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37546807527542114, "epoch": 6.03, "learning_rate": 3.320601127113337e-05, "loss": 0.5764, "step": 7139, "task_loss": 0.8095757365226746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0992403030395508, "epoch": 6.04, "learning_rate": 3.320288040075141e-05, "loss": 0.9189, "step": 7140, "task_loss": 0.46530717611312866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7777402400970459, "epoch": 6.04, "learning_rate": 3.3199749530369443e-05, "loss": 1.0851, "step": 7141, "task_loss": 0.7369075417518616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6534335613250732, "epoch": 6.04, "learning_rate": 3.3196618659987475e-05, "loss": 0.776, "step": 7142, "task_loss": 0.9643679857254028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9301661252975464, "epoch": 6.04, "learning_rate": 3.319348778960551e-05, "loss": 1.0388, "step": 7143, "task_loss": 0.3028635084629059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6001161932945251, "epoch": 6.04, "learning_rate": 3.3190356919223546e-05, "loss": 0.8049, "step": 7144, "task_loss": 0.6823374032974243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.778194785118103, "epoch": 6.04, "learning_rate": 3.318722604884158e-05, "loss": 0.8403, "step": 7145, "task_loss": 1.8501709699630737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.528243899345398, "epoch": 6.04, "learning_rate": 3.318409517845961e-05, "loss": 0.8909, "step": 7146, "task_loss": 0.4290428161621094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4963325262069702, "epoch": 6.04, "learning_rate": 3.318096430807765e-05, "loss": 0.953, "step": 7147, "task_loss": 0.9002098441123962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8333567380905151, "epoch": 6.04, "learning_rate": 3.317783343769568e-05, "loss": 0.8475, "step": 7148, "task_loss": 0.6226149201393127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2251341342926025, "epoch": 6.04, "learning_rate": 3.317470256731372e-05, "loss": 0.8703, "step": 7149, "task_loss": 0.7961824536323547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0000324249267578, "epoch": 6.04, "learning_rate": 3.317157169693175e-05, "loss": 0.9418, "step": 7150, "task_loss": 0.6477830410003662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4890737533569336, "epoch": 6.04, "learning_rate": 3.316844082654978e-05, "loss": 0.635, "step": 7151, "task_loss": 0.5696635246276855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6887727379798889, "epoch": 6.05, "learning_rate": 3.316530995616782e-05, "loss": 0.844, "step": 7152, "task_loss": 0.54001784324646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35125404596328735, "epoch": 6.05, "learning_rate": 3.316217908578585e-05, "loss": 0.7269, "step": 7153, "task_loss": 0.21781465411186218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9590969085693359, "epoch": 6.05, "learning_rate": 3.3159048215403884e-05, "loss": 0.8396, "step": 7154, "task_loss": 0.7027856707572937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9902321100234985, "epoch": 6.05, "learning_rate": 3.315591734502192e-05, "loss": 0.9205, "step": 7155, "task_loss": 0.9727063179016113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8938536643981934, "epoch": 6.05, "learning_rate": 3.3152786474639954e-05, "loss": 0.9698, "step": 7156, "task_loss": 1.0942778587341309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4156970977783203, "epoch": 6.05, "learning_rate": 3.3149655604257986e-05, "loss": 0.7891, "step": 7157, "task_loss": 0.19956891238689423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7018998861312866, "epoch": 6.05, "learning_rate": 3.314652473387602e-05, "loss": 0.8201, "step": 7158, "task_loss": 0.41200289130210876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5255618095397949, "epoch": 6.05, "learning_rate": 3.3143393863494056e-05, "loss": 0.5821, "step": 7159, "task_loss": 0.2490938901901245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9157636165618896, "epoch": 6.05, "learning_rate": 3.314026299311209e-05, "loss": 0.7466, "step": 7160, "task_loss": 1.1193571090698242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8180738687515259, "epoch": 6.05, "learning_rate": 3.313713212273012e-05, "loss": 0.8354, "step": 7161, "task_loss": 1.6205967664718628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0271759033203125, "epoch": 6.05, "learning_rate": 3.313400125234816e-05, "loss": 0.9372, "step": 7162, "task_loss": 0.9879305958747864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8472605347633362, "epoch": 6.05, "learning_rate": 3.313087038196619e-05, "loss": 0.5789, "step": 7163, "task_loss": 0.3783654272556305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5172886252403259, "epoch": 6.06, "learning_rate": 3.312773951158422e-05, "loss": 0.6104, "step": 7164, "task_loss": 0.21511127054691315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.059041976928711, "epoch": 6.06, "learning_rate": 3.3124608641202254e-05, "loss": 0.8279, "step": 7165, "task_loss": 0.42630329728126526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48553749918937683, "epoch": 6.06, "learning_rate": 3.312147777082029e-05, "loss": 0.5652, "step": 7166, "task_loss": 0.41387590765953064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5201667547225952, "epoch": 6.06, "learning_rate": 3.3118346900438324e-05, "loss": 0.659, "step": 7167, "task_loss": 0.3395191431045532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6622797846794128, "epoch": 6.06, "learning_rate": 3.3115216030056356e-05, "loss": 0.7926, "step": 7168, "task_loss": 0.08317013829946518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8807845115661621, "epoch": 6.06, "learning_rate": 3.311208515967439e-05, "loss": 0.8638, "step": 7169, "task_loss": 0.8471167087554932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8999598622322083, "epoch": 6.06, "learning_rate": 3.3108954289292426e-05, "loss": 0.7561, "step": 7170, "task_loss": 0.897377610206604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.116554617881775, "epoch": 6.06, "learning_rate": 3.310582341891046e-05, "loss": 1.0756, "step": 7171, "task_loss": 1.4311957359313965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1784402132034302, "epoch": 6.06, "learning_rate": 3.310269254852849e-05, "loss": 0.913, "step": 7172, "task_loss": 0.6084955334663391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0737801790237427, "epoch": 6.06, "learning_rate": 3.309956167814653e-05, "loss": 1.0027, "step": 7173, "task_loss": 1.303377389907837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9058513641357422, "epoch": 6.06, "learning_rate": 3.309643080776456e-05, "loss": 0.9985, "step": 7174, "task_loss": 1.1446502208709717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1382520198822021, "epoch": 6.07, "learning_rate": 3.309329993738259e-05, "loss": 0.8164, "step": 7175, "task_loss": 0.8581617474555969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7085356116294861, "epoch": 6.07, "learning_rate": 3.3090169067000624e-05, "loss": 0.6511, "step": 7176, "task_loss": 0.15360970795154572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6311012506484985, "epoch": 6.07, "learning_rate": 3.308703819661866e-05, "loss": 0.9414, "step": 7177, "task_loss": 0.20489773154258728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.375742644071579, "epoch": 6.07, "learning_rate": 3.3083907326236694e-05, "loss": 0.9145, "step": 7178, "task_loss": 0.7220238447189331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5543814301490784, "epoch": 6.07, "learning_rate": 3.3080776455854726e-05, "loss": 0.9033, "step": 7179, "task_loss": 0.7823293209075928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1370460987091064, "epoch": 6.07, "learning_rate": 3.307764558547276e-05, "loss": 0.8611, "step": 7180, "task_loss": 1.2608683109283447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.785584568977356, "epoch": 6.07, "learning_rate": 3.3074514715090796e-05, "loss": 0.8759, "step": 7181, "task_loss": 1.4929392337799072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3566522598266602, "epoch": 6.07, "learning_rate": 3.307138384470883e-05, "loss": 0.9266, "step": 7182, "task_loss": 1.114668369293213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5956746935844421, "epoch": 6.07, "learning_rate": 3.3068252974326866e-05, "loss": 0.7237, "step": 7183, "task_loss": 0.2632259726524353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5571725964546204, "epoch": 6.07, "learning_rate": 3.30651221039449e-05, "loss": 0.9992, "step": 7184, "task_loss": 0.08946076780557632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3214747905731201, "epoch": 6.07, "learning_rate": 3.306199123356294e-05, "loss": 0.9412, "step": 7185, "task_loss": 0.6194925308227539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5688329935073853, "epoch": 6.07, "learning_rate": 3.305886036318097e-05, "loss": 0.9268, "step": 7186, "task_loss": 1.6872705221176147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5641540884971619, "epoch": 6.08, "learning_rate": 3.3055729492799e-05, "loss": 0.6981, "step": 7187, "task_loss": 0.7155500054359436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1988914012908936, "epoch": 6.08, "learning_rate": 3.305259862241704e-05, "loss": 1.0173, "step": 7188, "task_loss": 1.7789032459259033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7102258205413818, "epoch": 6.08, "learning_rate": 3.304946775203507e-05, "loss": 0.7672, "step": 7189, "task_loss": 1.1162161827087402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0625098943710327, "epoch": 6.08, "learning_rate": 3.30463368816531e-05, "loss": 0.8041, "step": 7190, "task_loss": 0.943272054195404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6266254186630249, "epoch": 6.08, "learning_rate": 3.3043206011271134e-05, "loss": 0.6724, "step": 7191, "task_loss": 0.8303369283676147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3988325595855713, "epoch": 6.08, "learning_rate": 3.304007514088917e-05, "loss": 1.2212, "step": 7192, "task_loss": 1.351244330406189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8924081325531006, "epoch": 6.08, "learning_rate": 3.3036944270507204e-05, "loss": 0.8101, "step": 7193, "task_loss": 0.4890406131744385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6934230327606201, "epoch": 6.08, "learning_rate": 3.3033813400125236e-05, "loss": 0.8124, "step": 7194, "task_loss": 0.35416528582572937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.008449673652649, "epoch": 6.08, "learning_rate": 3.303068252974327e-05, "loss": 0.7062, "step": 7195, "task_loss": 0.6381125450134277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8176251649856567, "epoch": 6.08, "learning_rate": 3.3027551659361307e-05, "loss": 0.9339, "step": 7196, "task_loss": 0.7797865867614746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7067531943321228, "epoch": 6.08, "learning_rate": 3.302442078897934e-05, "loss": 0.8761, "step": 7197, "task_loss": 0.3746439814567566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49363973736763, "epoch": 6.08, "learning_rate": 3.302128991859737e-05, "loss": 0.7128, "step": 7198, "task_loss": 0.803598940372467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42610877752304077, "epoch": 6.09, "learning_rate": 3.301815904821541e-05, "loss": 0.5311, "step": 7199, "task_loss": 0.7414965629577637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5406156778335571, "epoch": 6.09, "learning_rate": 3.301502817783344e-05, "loss": 0.6334, "step": 7200, "task_loss": 0.49655041098594666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5218297839164734, "epoch": 6.09, "learning_rate": 3.301189730745147e-05, "loss": 0.9439, "step": 7201, "task_loss": 0.5105909109115601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7390440702438354, "epoch": 6.09, "learning_rate": 3.3008766437069504e-05, "loss": 0.7001, "step": 7202, "task_loss": 0.7765141725540161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5943429470062256, "epoch": 6.09, "learning_rate": 3.300563556668754e-05, "loss": 0.7821, "step": 7203, "task_loss": 0.809565544128418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8465864062309265, "epoch": 6.09, "learning_rate": 3.3002504696305574e-05, "loss": 1.0961, "step": 7204, "task_loss": 1.623387098312378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4604342579841614, "epoch": 6.09, "learning_rate": 3.2999373825923606e-05, "loss": 0.8162, "step": 7205, "task_loss": 0.3889962434768677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9671773314476013, "epoch": 6.09, "learning_rate": 3.299624295554164e-05, "loss": 0.9616, "step": 7206, "task_loss": 1.0913949012756348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7427002191543579, "epoch": 6.09, "learning_rate": 3.2993112085159677e-05, "loss": 0.74, "step": 7207, "task_loss": 0.33837154507637024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8232676982879639, "epoch": 6.09, "learning_rate": 3.298998121477771e-05, "loss": 0.8125, "step": 7208, "task_loss": 1.0828593969345093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7201483249664307, "epoch": 6.09, "learning_rate": 3.298685034439574e-05, "loss": 0.7961, "step": 7209, "task_loss": 0.9159234166145325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.746338427066803, "epoch": 6.09, "learning_rate": 3.298371947401378e-05, "loss": 0.7384, "step": 7210, "task_loss": 0.8733844757080078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4114792346954346, "epoch": 6.1, "learning_rate": 3.298058860363181e-05, "loss": 0.993, "step": 7211, "task_loss": 1.81383216381073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7213934659957886, "epoch": 6.1, "learning_rate": 3.297745773324984e-05, "loss": 0.7316, "step": 7212, "task_loss": 0.9568859934806824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.39216947555542, "epoch": 6.1, "learning_rate": 3.2974326862867874e-05, "loss": 0.8099, "step": 7213, "task_loss": 1.096618413925171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.071499228477478, "epoch": 6.1, "learning_rate": 3.297119599248591e-05, "loss": 0.9951, "step": 7214, "task_loss": 1.2135751247406006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7367380261421204, "epoch": 6.1, "learning_rate": 3.2968065122103944e-05, "loss": 0.6778, "step": 7215, "task_loss": 1.3366519212722778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8281238079071045, "epoch": 6.1, "learning_rate": 3.2964934251721976e-05, "loss": 0.8731, "step": 7216, "task_loss": 1.3886027336120605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4121251106262207, "epoch": 6.1, "learning_rate": 3.2961803381340015e-05, "loss": 1.1652, "step": 7217, "task_loss": 1.4266517162322998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7329757809638977, "epoch": 6.1, "learning_rate": 3.2958672510958046e-05, "loss": 0.6124, "step": 7218, "task_loss": 1.5782407522201538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1108438968658447, "epoch": 6.1, "learning_rate": 3.2955541640576085e-05, "loss": 0.8535, "step": 7219, "task_loss": 0.6909410953521729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6377221941947937, "epoch": 6.1, "learning_rate": 3.295241077019412e-05, "loss": 0.8193, "step": 7220, "task_loss": 0.0900893583893776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 2.324317216873169, "epoch": 6.1, "learning_rate": 3.2949279899812155e-05, "loss": 1.0227, "step": 7221, "task_loss": 1.6604886054992676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9877737164497375, "epoch": 6.1, "learning_rate": 3.294614902943019e-05, "loss": 0.8798, "step": 7222, "task_loss": 1.2588233947753906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8070646524429321, "epoch": 6.11, "learning_rate": 3.294301815904822e-05, "loss": 0.9446, "step": 7223, "task_loss": 0.5455974340438843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2271943092346191, "epoch": 6.11, "learning_rate": 3.293988728866625e-05, "loss": 0.9868, "step": 7224, "task_loss": 1.5881534814834595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7457537651062012, "epoch": 6.11, "learning_rate": 3.293675641828429e-05, "loss": 0.7422, "step": 7225, "task_loss": 1.2155075073242188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7189030647277832, "epoch": 6.11, "learning_rate": 3.293362554790232e-05, "loss": 0.9107, "step": 7226, "task_loss": 0.7629039287567139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6626107096672058, "epoch": 6.11, "learning_rate": 3.293049467752035e-05, "loss": 0.8595, "step": 7227, "task_loss": 0.8797781467437744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2543718814849854, "epoch": 6.11, "learning_rate": 3.2927363807138385e-05, "loss": 0.916, "step": 7228, "task_loss": 1.115072250366211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3938689231872559, "epoch": 6.11, "learning_rate": 3.292423293675642e-05, "loss": 0.8787, "step": 7229, "task_loss": 1.6555532217025757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5407358407974243, "epoch": 6.11, "learning_rate": 3.2921102066374455e-05, "loss": 0.9254, "step": 7230, "task_loss": 0.678122878074646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2137285470962524, "epoch": 6.11, "learning_rate": 3.291797119599249e-05, "loss": 0.9025, "step": 7231, "task_loss": 0.8060026168823242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6895594596862793, "epoch": 6.11, "learning_rate": 3.291484032561052e-05, "loss": 0.7984, "step": 7232, "task_loss": 0.3280414938926697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6324382424354553, "epoch": 6.11, "learning_rate": 3.291170945522856e-05, "loss": 0.7469, "step": 7233, "task_loss": 0.8111462593078613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4343222379684448, "epoch": 6.11, "learning_rate": 3.290857858484659e-05, "loss": 0.8268, "step": 7234, "task_loss": 0.487801730632782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5764298439025879, "epoch": 6.12, "learning_rate": 3.290544771446462e-05, "loss": 0.9203, "step": 7235, "task_loss": 0.6802922487258911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5608959197998047, "epoch": 6.12, "learning_rate": 3.290231684408266e-05, "loss": 0.7417, "step": 7236, "task_loss": 0.4217790365219116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9254236817359924, "epoch": 6.12, "learning_rate": 3.289918597370069e-05, "loss": 0.957, "step": 7237, "task_loss": 1.7511011362075806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.016242265701294, "epoch": 6.12, "learning_rate": 3.289605510331872e-05, "loss": 0.7137, "step": 7238, "task_loss": 0.5952972173690796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1129541397094727, "epoch": 6.12, "learning_rate": 3.2892924232936754e-05, "loss": 0.8471, "step": 7239, "task_loss": 1.572275996208191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0704169273376465, "epoch": 6.12, "learning_rate": 3.288979336255479e-05, "loss": 0.8061, "step": 7240, "task_loss": 0.4230737090110779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7752761244773865, "epoch": 6.12, "learning_rate": 3.2886662492172825e-05, "loss": 0.9037, "step": 7241, "task_loss": 0.5870619416236877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0267882347106934, "epoch": 6.12, "learning_rate": 3.288353162179086e-05, "loss": 0.8525, "step": 7242, "task_loss": 1.0652021169662476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6259417533874512, "epoch": 6.12, "learning_rate": 3.288040075140889e-05, "loss": 0.7123, "step": 7243, "task_loss": 0.1983826607465744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7136424779891968, "epoch": 6.12, "learning_rate": 3.287726988102693e-05, "loss": 0.758, "step": 7244, "task_loss": 1.190396785736084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9669165015220642, "epoch": 6.12, "learning_rate": 3.287413901064496e-05, "loss": 0.8418, "step": 7245, "task_loss": 0.8177785873413086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7350203990936279, "epoch": 6.13, "learning_rate": 3.287100814026299e-05, "loss": 0.8488, "step": 7246, "task_loss": 1.458043098449707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7546119689941406, "epoch": 6.13, "learning_rate": 3.286787726988103e-05, "loss": 0.6627, "step": 7247, "task_loss": 1.4475014209747314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8292967081069946, "epoch": 6.13, "learning_rate": 3.286474639949906e-05, "loss": 0.8145, "step": 7248, "task_loss": 1.1629430055618286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7749509811401367, "epoch": 6.13, "learning_rate": 3.286161552911709e-05, "loss": 0.7604, "step": 7249, "task_loss": 0.4459178149700165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.120512843132019, "epoch": 6.13, "learning_rate": 3.285848465873513e-05, "loss": 0.9219, "step": 7250, "task_loss": 0.7899014949798584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8496295809745789, "epoch": 6.13, "learning_rate": 3.285535378835316e-05, "loss": 0.9876, "step": 7251, "task_loss": 1.2626323699951172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9511977434158325, "epoch": 6.13, "learning_rate": 3.28522229179712e-05, "loss": 0.9788, "step": 7252, "task_loss": 1.418232798576355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7151068449020386, "epoch": 6.13, "learning_rate": 3.284909204758923e-05, "loss": 0.8037, "step": 7253, "task_loss": 0.8156168460845947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8169862031936646, "epoch": 6.13, "learning_rate": 3.2845961177207265e-05, "loss": 0.7924, "step": 7254, "task_loss": 0.4374980628490448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43245869874954224, "epoch": 6.13, "learning_rate": 3.2842830306825304e-05, "loss": 0.6304, "step": 7255, "task_loss": 0.8856765031814575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6001585721969604, "epoch": 6.13, "learning_rate": 3.2839699436443335e-05, "loss": 0.817, "step": 7256, "task_loss": 0.3264996409416199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6130332946777344, "epoch": 6.13, "learning_rate": 3.283656856606137e-05, "loss": 0.763, "step": 7257, "task_loss": 0.8805605173110962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9700088500976562, "epoch": 6.14, "learning_rate": 3.2833437695679406e-05, "loss": 1.0116, "step": 7258, "task_loss": 1.4806195497512817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.769963264465332, "epoch": 6.14, "learning_rate": 3.283030682529744e-05, "loss": 0.9732, "step": 7259, "task_loss": 0.9595872163772583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1159865856170654, "epoch": 6.14, "learning_rate": 3.282717595491547e-05, "loss": 0.9749, "step": 7260, "task_loss": 1.403681993484497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1071548461914062, "epoch": 6.14, "learning_rate": 3.28240450845335e-05, "loss": 0.8232, "step": 7261, "task_loss": 0.5911343097686768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4633708596229553, "epoch": 6.14, "learning_rate": 3.282091421415154e-05, "loss": 0.6738, "step": 7262, "task_loss": 0.10251853615045547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7144120931625366, "epoch": 6.14, "learning_rate": 3.281778334376957e-05, "loss": 0.8064, "step": 7263, "task_loss": 1.0622191429138184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.552836000919342, "epoch": 6.14, "learning_rate": 3.28146524733876e-05, "loss": 0.7656, "step": 7264, "task_loss": 1.1450711488723755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2204803228378296, "epoch": 6.14, "learning_rate": 3.2811521603005635e-05, "loss": 0.8694, "step": 7265, "task_loss": 1.472355604171753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9937061071395874, "epoch": 6.14, "learning_rate": 3.2808390732623674e-05, "loss": 1.0217, "step": 7266, "task_loss": 0.8035581707954407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5292882919311523, "epoch": 6.14, "learning_rate": 3.2805259862241705e-05, "loss": 0.6245, "step": 7267, "task_loss": 0.9698188304901123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4934617280960083, "epoch": 6.14, "learning_rate": 3.280212899185974e-05, "loss": 0.5657, "step": 7268, "task_loss": 0.8939364552497864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7787710428237915, "epoch": 6.14, "learning_rate": 3.279899812147777e-05, "loss": 0.6803, "step": 7269, "task_loss": 0.4162706434726715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3583146333694458, "epoch": 6.15, "learning_rate": 3.279586725109581e-05, "loss": 1.0086, "step": 7270, "task_loss": 0.12387027591466904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3930201530456543, "epoch": 6.15, "learning_rate": 3.279273638071384e-05, "loss": 1.0179, "step": 7271, "task_loss": 1.0771373510360718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7735474109649658, "epoch": 6.15, "learning_rate": 3.278960551033187e-05, "loss": 0.6251, "step": 7272, "task_loss": 1.0378391742706299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6677408218383789, "epoch": 6.15, "learning_rate": 3.278647463994991e-05, "loss": 0.6246, "step": 7273, "task_loss": 0.41958147287368774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9064851403236389, "epoch": 6.15, "learning_rate": 3.278334376956794e-05, "loss": 0.9806, "step": 7274, "task_loss": 0.7262861728668213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.353480339050293, "epoch": 6.15, "learning_rate": 3.278021289918597e-05, "loss": 0.9739, "step": 7275, "task_loss": 2.595172166824341 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7869868278503418, "epoch": 6.15, "learning_rate": 3.2777082028804005e-05, "loss": 0.645, "step": 7276, "task_loss": 1.1938377618789673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.713625967502594, "epoch": 6.15, "learning_rate": 3.2773951158422044e-05, "loss": 0.73, "step": 7277, "task_loss": 0.48615118861198425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8915488719940186, "epoch": 6.15, "learning_rate": 3.2770820288040075e-05, "loss": 0.6819, "step": 7278, "task_loss": 0.31078222393989563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9282675981521606, "epoch": 6.15, "learning_rate": 3.276768941765811e-05, "loss": 0.8373, "step": 7279, "task_loss": 0.70148104429245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2089672088623047, "epoch": 6.15, "learning_rate": 3.276455854727614e-05, "loss": 0.7231, "step": 7280, "task_loss": 1.127596378326416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9949270486831665, "epoch": 6.15, "learning_rate": 3.276142767689418e-05, "loss": 0.8589, "step": 7281, "task_loss": 0.402688592672348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.613531231880188, "epoch": 6.16, "learning_rate": 3.275829680651221e-05, "loss": 0.6807, "step": 7282, "task_loss": 0.4912908375263214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8819475769996643, "epoch": 6.16, "learning_rate": 3.275516593613024e-05, "loss": 1.0005, "step": 7283, "task_loss": 0.26097986102104187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5484189987182617, "epoch": 6.16, "learning_rate": 3.275203506574828e-05, "loss": 0.6486, "step": 7284, "task_loss": 0.293957382440567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42914727330207825, "epoch": 6.16, "learning_rate": 3.274890419536631e-05, "loss": 0.5924, "step": 7285, "task_loss": 0.611358642578125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.015190839767456, "epoch": 6.16, "learning_rate": 3.274577332498435e-05, "loss": 0.8787, "step": 7286, "task_loss": 0.7021806836128235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.083798885345459, "epoch": 6.16, "learning_rate": 3.274264245460238e-05, "loss": 0.8532, "step": 7287, "task_loss": 0.396785706281662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7775477766990662, "epoch": 6.16, "learning_rate": 3.273951158422042e-05, "loss": 0.7774, "step": 7288, "task_loss": 0.7240651845932007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5509438514709473, "epoch": 6.16, "learning_rate": 3.273638071383845e-05, "loss": 1.3026, "step": 7289, "task_loss": 1.3912397623062134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.109861135482788, "epoch": 6.16, "learning_rate": 3.2733249843456484e-05, "loss": 0.7012, "step": 7290, "task_loss": 0.3849550187587738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8794814944267273, "epoch": 6.16, "learning_rate": 3.2730118973074516e-05, "loss": 0.9183, "step": 7291, "task_loss": 0.5939740538597107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47601574659347534, "epoch": 6.16, "learning_rate": 3.2726988102692554e-05, "loss": 0.9039, "step": 7292, "task_loss": 0.3551356792449951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8585556149482727, "epoch": 6.16, "learning_rate": 3.2723857232310586e-05, "loss": 0.8394, "step": 7293, "task_loss": 0.871902346611023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7592427134513855, "epoch": 6.17, "learning_rate": 3.272072636192862e-05, "loss": 0.6137, "step": 7294, "task_loss": 0.5438873171806335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7113496661186218, "epoch": 6.17, "learning_rate": 3.2717595491546656e-05, "loss": 0.756, "step": 7295, "task_loss": 0.4256446361541748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7635424137115479, "epoch": 6.17, "learning_rate": 3.271446462116469e-05, "loss": 0.6516, "step": 7296, "task_loss": 0.974885106086731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3699941635131836, "epoch": 6.17, "learning_rate": 3.271133375078272e-05, "loss": 0.7476, "step": 7297, "task_loss": 0.9819663166999817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8456522822380066, "epoch": 6.17, "learning_rate": 3.270820288040075e-05, "loss": 0.7659, "step": 7298, "task_loss": 0.9681025147438049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0990904569625854, "epoch": 6.17, "learning_rate": 3.270507201001879e-05, "loss": 0.6878, "step": 7299, "task_loss": 0.9404675364494324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6646413803100586, "epoch": 6.17, "learning_rate": 3.270194113963682e-05, "loss": 0.7453, "step": 7300, "task_loss": 0.34146803617477417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1818881034851074, "epoch": 6.17, "learning_rate": 3.2698810269254854e-05, "loss": 0.886, "step": 7301, "task_loss": 1.5100347995758057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4099397659301758, "epoch": 6.17, "learning_rate": 3.2695679398872885e-05, "loss": 0.7068, "step": 7302, "task_loss": 0.39185234904289246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7659066915512085, "epoch": 6.17, "learning_rate": 3.2692548528490924e-05, "loss": 0.8172, "step": 7303, "task_loss": 0.5856067538261414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.554448127746582, "epoch": 6.17, "learning_rate": 3.2689417658108956e-05, "loss": 0.6753, "step": 7304, "task_loss": 0.14331193268299103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8544219732284546, "epoch": 6.17, "learning_rate": 3.268628678772699e-05, "loss": 0.9254, "step": 7305, "task_loss": 1.6448262929916382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8244155645370483, "epoch": 6.18, "learning_rate": 3.268315591734502e-05, "loss": 0.689, "step": 7306, "task_loss": 0.28305783867836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8546792268753052, "epoch": 6.18, "learning_rate": 3.268002504696306e-05, "loss": 0.8742, "step": 7307, "task_loss": 0.8718169927597046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8263027667999268, "epoch": 6.18, "learning_rate": 3.267689417658109e-05, "loss": 0.9463, "step": 7308, "task_loss": 1.6825847625732422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7141494750976562, "epoch": 6.18, "learning_rate": 3.267376330619912e-05, "loss": 0.8149, "step": 7309, "task_loss": 0.7658374905586243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4280381500720978, "epoch": 6.18, "learning_rate": 3.267063243581716e-05, "loss": 0.7367, "step": 7310, "task_loss": 0.08514410257339478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.142547369003296, "epoch": 6.18, "learning_rate": 3.266750156543519e-05, "loss": 0.852, "step": 7311, "task_loss": 1.0472568273544312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5313748121261597, "epoch": 6.18, "learning_rate": 3.2664370695053224e-05, "loss": 0.845, "step": 7312, "task_loss": 0.6642627120018005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.735202968120575, "epoch": 6.18, "learning_rate": 3.2661239824671255e-05, "loss": 0.8748, "step": 7313, "task_loss": 0.8031681776046753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45836949348449707, "epoch": 6.18, "learning_rate": 3.2658108954289294e-05, "loss": 0.8338, "step": 7314, "task_loss": 0.5234813690185547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.266512155532837, "epoch": 6.18, "learning_rate": 3.2654978083907326e-05, "loss": 0.8349, "step": 7315, "task_loss": 1.4111123085021973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9350287914276123, "epoch": 6.18, "learning_rate": 3.265184721352536e-05, "loss": 0.9556, "step": 7316, "task_loss": 1.0906126499176025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6406686902046204, "epoch": 6.19, "learning_rate": 3.2648716343143396e-05, "loss": 0.7669, "step": 7317, "task_loss": 1.0115569829940796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6262810826301575, "epoch": 6.19, "learning_rate": 3.264558547276143e-05, "loss": 0.8443, "step": 7318, "task_loss": 1.051641821861267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0566011667251587, "epoch": 6.19, "learning_rate": 3.2642454602379466e-05, "loss": 0.765, "step": 7319, "task_loss": 0.6180446743965149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8334544897079468, "epoch": 6.19, "learning_rate": 3.26393237319975e-05, "loss": 0.6594, "step": 7320, "task_loss": 1.4069479703903198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0540422201156616, "epoch": 6.19, "learning_rate": 3.263619286161553e-05, "loss": 0.9425, "step": 7321, "task_loss": 1.8022420406341553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1184041500091553, "epoch": 6.19, "learning_rate": 3.263306199123357e-05, "loss": 0.8045, "step": 7322, "task_loss": 1.5007835626602173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5582259297370911, "epoch": 6.19, "learning_rate": 3.26299311208516e-05, "loss": 0.7688, "step": 7323, "task_loss": 0.6438292264938354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8635067939758301, "epoch": 6.19, "learning_rate": 3.262680025046963e-05, "loss": 0.7471, "step": 7324, "task_loss": 0.2745419144630432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7822767496109009, "epoch": 6.19, "learning_rate": 3.262366938008767e-05, "loss": 0.9873, "step": 7325, "task_loss": 1.1534905433654785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6158425807952881, "epoch": 6.19, "learning_rate": 3.26205385097057e-05, "loss": 0.6633, "step": 7326, "task_loss": 0.18320229649543762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5610923767089844, "epoch": 6.19, "learning_rate": 3.2617407639323734e-05, "loss": 0.7544, "step": 7327, "task_loss": 0.3549681007862091 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5028581619262695, "epoch": 6.19, "learning_rate": 3.2614276768941766e-05, "loss": 0.629, "step": 7328, "task_loss": 0.8936004638671875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7661334276199341, "epoch": 6.2, "learning_rate": 3.2611145898559805e-05, "loss": 0.933, "step": 7329, "task_loss": 0.5842020511627197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9984524250030518, "epoch": 6.2, "learning_rate": 3.2608015028177836e-05, "loss": 0.8314, "step": 7330, "task_loss": 0.44862762093544006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5166067481040955, "epoch": 6.2, "learning_rate": 3.260488415779587e-05, "loss": 0.871, "step": 7331, "task_loss": 1.3327159881591797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8523465394973755, "epoch": 6.2, "learning_rate": 3.260175328741391e-05, "loss": 0.7789, "step": 7332, "task_loss": 0.430145263671875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0481736660003662, "epoch": 6.2, "learning_rate": 3.259862241703194e-05, "loss": 0.852, "step": 7333, "task_loss": 1.8336578607559204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.567872166633606, "epoch": 6.2, "learning_rate": 3.259549154664997e-05, "loss": 0.6544, "step": 7334, "task_loss": 0.8347977995872498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7673084139823914, "epoch": 6.2, "learning_rate": 3.2592360676268e-05, "loss": 0.9003, "step": 7335, "task_loss": 0.5178557634353638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5604450702667236, "epoch": 6.2, "learning_rate": 3.258922980588604e-05, "loss": 0.6481, "step": 7336, "task_loss": 0.3341405391693115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0049335956573486, "epoch": 6.2, "learning_rate": 3.258609893550407e-05, "loss": 0.8434, "step": 7337, "task_loss": 1.1404205560684204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9021672010421753, "epoch": 6.2, "learning_rate": 3.2582968065122104e-05, "loss": 0.7789, "step": 7338, "task_loss": 1.407596468925476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6901568174362183, "epoch": 6.2, "learning_rate": 3.2579837194740136e-05, "loss": 0.6133, "step": 7339, "task_loss": 0.4225136637687683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.131493091583252, "epoch": 6.2, "learning_rate": 3.2576706324358174e-05, "loss": 0.9917, "step": 7340, "task_loss": 0.8385602235794067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0074397325515747, "epoch": 6.21, "learning_rate": 3.2573575453976206e-05, "loss": 0.8648, "step": 7341, "task_loss": 1.052875280380249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6577971577644348, "epoch": 6.21, "learning_rate": 3.257044458359424e-05, "loss": 0.837, "step": 7342, "task_loss": 1.072435975074768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8135558366775513, "epoch": 6.21, "learning_rate": 3.256731371321227e-05, "loss": 0.8538, "step": 7343, "task_loss": 0.7651817798614502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6326394081115723, "epoch": 6.21, "learning_rate": 3.256418284283031e-05, "loss": 0.7389, "step": 7344, "task_loss": 0.6769306659698486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8081209659576416, "epoch": 6.21, "learning_rate": 3.256105197244834e-05, "loss": 0.8023, "step": 7345, "task_loss": 0.43164172768592834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.631638765335083, "epoch": 6.21, "learning_rate": 3.255792110206637e-05, "loss": 0.9805, "step": 7346, "task_loss": 1.2753331661224365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8709975481033325, "epoch": 6.21, "learning_rate": 3.255479023168441e-05, "loss": 0.7583, "step": 7347, "task_loss": 1.063697099685669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6055271625518799, "epoch": 6.21, "learning_rate": 3.255165936130244e-05, "loss": 0.7847, "step": 7348, "task_loss": 0.5219417810440063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3159587383270264, "epoch": 6.21, "learning_rate": 3.2548528490920474e-05, "loss": 0.9735, "step": 7349, "task_loss": 2.6025660037994385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.608206033706665, "epoch": 6.21, "learning_rate": 3.2545397620538506e-05, "loss": 0.9723, "step": 7350, "task_loss": 0.5630318522453308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0121443271636963, "epoch": 6.21, "learning_rate": 3.2542266750156544e-05, "loss": 0.8835, "step": 7351, "task_loss": 1.1738219261169434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0399011373519897, "epoch": 6.21, "learning_rate": 3.2539135879774576e-05, "loss": 1.0145, "step": 7352, "task_loss": 0.840496301651001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4685901999473572, "epoch": 6.22, "learning_rate": 3.2536005009392615e-05, "loss": 0.8034, "step": 7353, "task_loss": 1.4474095106124878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9569340944290161, "epoch": 6.22, "learning_rate": 3.2532874139010647e-05, "loss": 0.9772, "step": 7354, "task_loss": 1.0089342594146729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7145942449569702, "epoch": 6.22, "learning_rate": 3.2529743268628685e-05, "loss": 0.8766, "step": 7355, "task_loss": 0.5248070955276489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.729185163974762, "epoch": 6.22, "learning_rate": 3.252661239824672e-05, "loss": 0.8852, "step": 7356, "task_loss": 1.2106950283050537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9990220069885254, "epoch": 6.22, "learning_rate": 3.252348152786475e-05, "loss": 0.8269, "step": 7357, "task_loss": 1.160241723060608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.118987798690796, "epoch": 6.22, "learning_rate": 3.252035065748279e-05, "loss": 0.9173, "step": 7358, "task_loss": 0.6607280373573303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8286850452423096, "epoch": 6.22, "learning_rate": 3.251721978710082e-05, "loss": 0.6231, "step": 7359, "task_loss": 0.8342722654342651 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6244360208511353, "epoch": 6.22, "learning_rate": 3.251408891671885e-05, "loss": 0.6295, "step": 7360, "task_loss": 0.7555914521217346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0811771154403687, "epoch": 6.22, "learning_rate": 3.251095804633688e-05, "loss": 0.9123, "step": 7361, "task_loss": 1.3200687170028687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6783034801483154, "epoch": 6.22, "learning_rate": 3.250782717595492e-05, "loss": 0.7624, "step": 7362, "task_loss": 0.5870183110237122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7520180940628052, "epoch": 6.22, "learning_rate": 3.250469630557295e-05, "loss": 0.8536, "step": 7363, "task_loss": 0.7762349843978882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9261247515678406, "epoch": 6.22, "learning_rate": 3.2501565435190985e-05, "loss": 0.8138, "step": 7364, "task_loss": 0.851408064365387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5763198137283325, "epoch": 6.23, "learning_rate": 3.2498434564809016e-05, "loss": 0.7453, "step": 7365, "task_loss": 0.8733024597167969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5538883209228516, "epoch": 6.23, "learning_rate": 3.2495303694427055e-05, "loss": 0.8088, "step": 7366, "task_loss": 0.15584638714790344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7479410171508789, "epoch": 6.23, "learning_rate": 3.249217282404509e-05, "loss": 0.6796, "step": 7367, "task_loss": 1.2751922607421875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7282272577285767, "epoch": 6.23, "learning_rate": 3.248904195366312e-05, "loss": 1.0999, "step": 7368, "task_loss": 0.685931921005249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7193988561630249, "epoch": 6.23, "learning_rate": 3.248591108328116e-05, "loss": 0.7411, "step": 7369, "task_loss": 0.9118191003799438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1510860919952393, "epoch": 6.23, "learning_rate": 3.248278021289919e-05, "loss": 0.9045, "step": 7370, "task_loss": 2.2200164794921875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5760983824729919, "epoch": 6.23, "learning_rate": 3.247964934251722e-05, "loss": 0.7473, "step": 7371, "task_loss": 0.3485730290412903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7540810108184814, "epoch": 6.23, "learning_rate": 3.247651847213525e-05, "loss": 0.7086, "step": 7372, "task_loss": 0.576192319393158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0534639358520508, "epoch": 6.23, "learning_rate": 3.247338760175329e-05, "loss": 0.8418, "step": 7373, "task_loss": 1.67023503780365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0563576221466064, "epoch": 6.23, "learning_rate": 3.247025673137132e-05, "loss": 0.797, "step": 7374, "task_loss": 1.0620529651641846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9320428371429443, "epoch": 6.23, "learning_rate": 3.2467125860989355e-05, "loss": 0.734, "step": 7375, "task_loss": 0.8374480605125427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5938379764556885, "epoch": 6.23, "learning_rate": 3.2463994990607386e-05, "loss": 0.7542, "step": 7376, "task_loss": 1.2183725833892822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5426383018493652, "epoch": 6.24, "learning_rate": 3.2460864120225425e-05, "loss": 0.8956, "step": 7377, "task_loss": 0.5199726819992065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.868013322353363, "epoch": 6.24, "learning_rate": 3.245773324984346e-05, "loss": 0.7547, "step": 7378, "task_loss": 1.263550877571106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6810237169265747, "epoch": 6.24, "learning_rate": 3.245460237946149e-05, "loss": 0.7896, "step": 7379, "task_loss": 0.814238429069519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5597769618034363, "epoch": 6.24, "learning_rate": 3.245147150907952e-05, "loss": 0.7247, "step": 7380, "task_loss": 0.7658183574676514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7371374368667603, "epoch": 6.24, "learning_rate": 3.244834063869756e-05, "loss": 0.8625, "step": 7381, "task_loss": 1.167942762374878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7553545236587524, "epoch": 6.24, "learning_rate": 3.244520976831559e-05, "loss": 0.9424, "step": 7382, "task_loss": 1.477034568786621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.07838773727417, "epoch": 6.24, "learning_rate": 3.244207889793362e-05, "loss": 0.9534, "step": 7383, "task_loss": 1.3485196828842163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5835322737693787, "epoch": 6.24, "learning_rate": 3.243894802755166e-05, "loss": 0.5998, "step": 7384, "task_loss": 0.5484700202941895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7088229060173035, "epoch": 6.24, "learning_rate": 3.243581715716969e-05, "loss": 0.7638, "step": 7385, "task_loss": 0.5915820002555847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9483219981193542, "epoch": 6.24, "learning_rate": 3.243268628678773e-05, "loss": 0.6439, "step": 7386, "task_loss": 0.3198609948158264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8334802389144897, "epoch": 6.24, "learning_rate": 3.242955541640576e-05, "loss": 0.6791, "step": 7387, "task_loss": 1.191267490386963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0218048095703125, "epoch": 6.24, "learning_rate": 3.2426424546023795e-05, "loss": 1.0107, "step": 7388, "task_loss": 0.357349157333374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7202906608581543, "epoch": 6.25, "learning_rate": 3.2423293675641833e-05, "loss": 0.7359, "step": 7389, "task_loss": 0.7320424318313599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7397348880767822, "epoch": 6.25, "learning_rate": 3.2420162805259865e-05, "loss": 0.766, "step": 7390, "task_loss": 0.32219526171684265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5715610980987549, "epoch": 6.25, "learning_rate": 3.24170319348779e-05, "loss": 0.7207, "step": 7391, "task_loss": 0.6734588146209717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1444363594055176, "epoch": 6.25, "learning_rate": 3.2413901064495936e-05, "loss": 0.8393, "step": 7392, "task_loss": 0.9344111084938049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7653133869171143, "epoch": 6.25, "learning_rate": 3.241077019411397e-05, "loss": 0.6705, "step": 7393, "task_loss": 0.7780746221542358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6980164051055908, "epoch": 6.25, "learning_rate": 3.2407639323732e-05, "loss": 0.6833, "step": 7394, "task_loss": 0.8973962664604187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3346365988254547, "epoch": 6.25, "learning_rate": 3.240450845335004e-05, "loss": 0.6376, "step": 7395, "task_loss": 0.41439101099967957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33774346113204956, "epoch": 6.25, "learning_rate": 3.240137758296807e-05, "loss": 0.5951, "step": 7396, "task_loss": 0.35691532492637634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8311076760292053, "epoch": 6.25, "learning_rate": 3.23982467125861e-05, "loss": 0.7282, "step": 7397, "task_loss": 1.0328325033187866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6381131410598755, "epoch": 6.25, "learning_rate": 3.239511584220413e-05, "loss": 0.9033, "step": 7398, "task_loss": 0.37309229373931885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8622696399688721, "epoch": 6.25, "learning_rate": 3.239198497182217e-05, "loss": 0.8809, "step": 7399, "task_loss": 1.7849137783050537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6526895761489868, "epoch": 6.26, "learning_rate": 3.23888541014402e-05, "loss": 0.7641, "step": 7400, "task_loss": 0.44649896025657654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6780917048454285, "epoch": 6.26, "learning_rate": 3.2385723231058235e-05, "loss": 0.7605, "step": 7401, "task_loss": 0.6705424189567566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8677247166633606, "epoch": 6.26, "learning_rate": 3.238259236067627e-05, "loss": 0.9181, "step": 7402, "task_loss": 0.8932814598083496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6970074772834778, "epoch": 6.26, "learning_rate": 3.2379461490294305e-05, "loss": 0.7122, "step": 7403, "task_loss": 0.40466731786727905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8236110806465149, "epoch": 6.26, "learning_rate": 3.237633061991234e-05, "loss": 0.6832, "step": 7404, "task_loss": 0.4035249650478363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8181665539741516, "epoch": 6.26, "learning_rate": 3.237319974953037e-05, "loss": 0.7612, "step": 7405, "task_loss": 0.6940217614173889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4214785099029541, "epoch": 6.26, "learning_rate": 3.237006887914841e-05, "loss": 0.9708, "step": 7406, "task_loss": 0.52804034948349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8273749947547913, "epoch": 6.26, "learning_rate": 3.236693800876644e-05, "loss": 0.6935, "step": 7407, "task_loss": 1.422276258468628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8654565811157227, "epoch": 6.26, "learning_rate": 3.236380713838447e-05, "loss": 1.0151, "step": 7408, "task_loss": 1.1338298320770264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8421639800071716, "epoch": 6.26, "learning_rate": 3.23606762680025e-05, "loss": 0.8043, "step": 7409, "task_loss": 0.9895845055580139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5976762771606445, "epoch": 6.26, "learning_rate": 3.235754539762054e-05, "loss": 0.9577, "step": 7410, "task_loss": 0.8442704081535339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9261865615844727, "epoch": 6.26, "learning_rate": 3.235441452723857e-05, "loss": 0.8026, "step": 7411, "task_loss": 1.1595200300216675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.574245274066925, "epoch": 6.27, "learning_rate": 3.2351283656856605e-05, "loss": 0.6244, "step": 7412, "task_loss": 0.6329626441001892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9840944409370422, "epoch": 6.27, "learning_rate": 3.234815278647464e-05, "loss": 0.9254, "step": 7413, "task_loss": 1.5791887044906616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6749947667121887, "epoch": 6.27, "learning_rate": 3.2345021916092675e-05, "loss": 0.7222, "step": 7414, "task_loss": 1.1688979864120483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9623695015907288, "epoch": 6.27, "learning_rate": 3.234189104571071e-05, "loss": 0.9083, "step": 7415, "task_loss": 1.1182655096054077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1748355627059937, "epoch": 6.27, "learning_rate": 3.233876017532874e-05, "loss": 1.0005, "step": 7416, "task_loss": 2.1206398010253906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5514520406723022, "epoch": 6.27, "learning_rate": 3.233562930494677e-05, "loss": 0.6926, "step": 7417, "task_loss": 0.5716879963874817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9593885540962219, "epoch": 6.27, "learning_rate": 3.233249843456481e-05, "loss": 0.7248, "step": 7418, "task_loss": 1.0415023565292358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7995764017105103, "epoch": 6.27, "learning_rate": 3.232936756418284e-05, "loss": 0.8309, "step": 7419, "task_loss": 0.3686891198158264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9709974527359009, "epoch": 6.27, "learning_rate": 3.232623669380088e-05, "loss": 0.8339, "step": 7420, "task_loss": 1.1966924667358398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8422772288322449, "epoch": 6.27, "learning_rate": 3.232310582341891e-05, "loss": 0.761, "step": 7421, "task_loss": 0.9485624432563782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8435242176055908, "epoch": 6.27, "learning_rate": 3.231997495303695e-05, "loss": 0.9091, "step": 7422, "task_loss": 0.3267122209072113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6868990659713745, "epoch": 6.27, "learning_rate": 3.231684408265498e-05, "loss": 0.7087, "step": 7423, "task_loss": 0.43273812532424927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6506779789924622, "epoch": 6.28, "learning_rate": 3.2313713212273013e-05, "loss": 0.6833, "step": 7424, "task_loss": 0.3093213438987732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8560687899589539, "epoch": 6.28, "learning_rate": 3.231058234189105e-05, "loss": 0.7457, "step": 7425, "task_loss": 0.5923929214477539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6620567440986633, "epoch": 6.28, "learning_rate": 3.2307451471509084e-05, "loss": 0.959, "step": 7426, "task_loss": 0.9801532030105591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4280385673046112, "epoch": 6.28, "learning_rate": 3.2304320601127116e-05, "loss": 0.8896, "step": 7427, "task_loss": 0.4752841889858246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5911158323287964, "epoch": 6.28, "learning_rate": 3.230118973074515e-05, "loss": 0.645, "step": 7428, "task_loss": 0.4199666380882263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6800719499588013, "epoch": 6.28, "learning_rate": 3.2298058860363186e-05, "loss": 0.6578, "step": 7429, "task_loss": 0.41622182726860046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9445850849151611, "epoch": 6.28, "learning_rate": 3.229492798998122e-05, "loss": 0.838, "step": 7430, "task_loss": 0.5747875571250916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0229312181472778, "epoch": 6.28, "learning_rate": 3.229179711959925e-05, "loss": 0.9249, "step": 7431, "task_loss": 0.7882843613624573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6005703210830688, "epoch": 6.28, "learning_rate": 3.228866624921729e-05, "loss": 0.624, "step": 7432, "task_loss": 0.8276329636573792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6262754201889038, "epoch": 6.28, "learning_rate": 3.228553537883532e-05, "loss": 0.7589, "step": 7433, "task_loss": 0.6553522348403931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.02410089969635, "epoch": 6.28, "learning_rate": 3.228240450845335e-05, "loss": 0.9908, "step": 7434, "task_loss": 0.5441640615463257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7210803627967834, "epoch": 6.28, "learning_rate": 3.2279273638071383e-05, "loss": 0.8997, "step": 7435, "task_loss": 0.3290754556655884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7145229578018188, "epoch": 6.29, "learning_rate": 3.227614276768942e-05, "loss": 0.7993, "step": 7436, "task_loss": 0.3828011155128479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8112426996231079, "epoch": 6.29, "learning_rate": 3.2273011897307454e-05, "loss": 0.7007, "step": 7437, "task_loss": 0.27596530318260193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1923717260360718, "epoch": 6.29, "learning_rate": 3.2269881026925486e-05, "loss": 1.2506, "step": 7438, "task_loss": 0.9145166277885437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5470309853553772, "epoch": 6.29, "learning_rate": 3.226675015654352e-05, "loss": 0.6496, "step": 7439, "task_loss": 0.8498038649559021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9760289192199707, "epoch": 6.29, "learning_rate": 3.2263619286161556e-05, "loss": 0.7266, "step": 7440, "task_loss": 2.129641532897949 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.576645016670227, "epoch": 6.29, "learning_rate": 3.226048841577959e-05, "loss": 0.8721, "step": 7441, "task_loss": 0.757213830947876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5402920246124268, "epoch": 6.29, "learning_rate": 3.225735754539762e-05, "loss": 0.8471, "step": 7442, "task_loss": 0.9763416647911072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.889595091342926, "epoch": 6.29, "learning_rate": 3.225422667501566e-05, "loss": 0.7082, "step": 7443, "task_loss": 0.672132134437561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4399985373020172, "epoch": 6.29, "learning_rate": 3.225109580463369e-05, "loss": 0.581, "step": 7444, "task_loss": 0.24576522409915924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8105830550193787, "epoch": 6.29, "learning_rate": 3.224796493425172e-05, "loss": 0.8828, "step": 7445, "task_loss": 0.2975460886955261 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9648386240005493, "epoch": 6.29, "learning_rate": 3.224483406386975e-05, "loss": 0.7892, "step": 7446, "task_loss": 0.9940316081047058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7414273023605347, "epoch": 6.29, "learning_rate": 3.224170319348779e-05, "loss": 0.7462, "step": 7447, "task_loss": 0.5254045724868774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4915833473205566, "epoch": 6.3, "learning_rate": 3.2238572323105824e-05, "loss": 1.055, "step": 7448, "task_loss": 1.0540252923965454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6795961856842041, "epoch": 6.3, "learning_rate": 3.2235441452723855e-05, "loss": 0.6569, "step": 7449, "task_loss": 0.9914053678512573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8248546123504639, "epoch": 6.3, "learning_rate": 3.223231058234189e-05, "loss": 1.0832, "step": 7450, "task_loss": 1.3167691230773926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7716928720474243, "epoch": 6.3, "learning_rate": 3.2229179711959926e-05, "loss": 0.9186, "step": 7451, "task_loss": 0.7149481773376465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0360779762268066, "epoch": 6.3, "learning_rate": 3.222604884157796e-05, "loss": 0.7852, "step": 7452, "task_loss": 0.34496423602104187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1228570938110352, "epoch": 6.3, "learning_rate": 3.2222917971195996e-05, "loss": 0.8835, "step": 7453, "task_loss": 1.1450875997543335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8143109083175659, "epoch": 6.3, "learning_rate": 3.221978710081403e-05, "loss": 0.6481, "step": 7454, "task_loss": 0.7261930108070374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4085594415664673, "epoch": 6.3, "learning_rate": 3.221665623043206e-05, "loss": 1.0962, "step": 7455, "task_loss": 2.4656362533569336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7279154062271118, "epoch": 6.3, "learning_rate": 3.22135253600501e-05, "loss": 0.7576, "step": 7456, "task_loss": 0.6204026341438293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8057430386543274, "epoch": 6.3, "learning_rate": 3.221039448966813e-05, "loss": 0.7579, "step": 7457, "task_loss": 1.5672756433486938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7166203260421753, "epoch": 6.3, "learning_rate": 3.220726361928617e-05, "loss": 0.6089, "step": 7458, "task_loss": 0.7504241466522217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6333633065223694, "epoch": 6.3, "learning_rate": 3.22041327489042e-05, "loss": 0.7652, "step": 7459, "task_loss": 1.1526726484298706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8660963773727417, "epoch": 6.31, "learning_rate": 3.220100187852223e-05, "loss": 0.8031, "step": 7460, "task_loss": 0.7135440111160278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.702748715877533, "epoch": 6.31, "learning_rate": 3.2197871008140264e-05, "loss": 0.5781, "step": 7461, "task_loss": 0.3461496829986572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.216275930404663, "epoch": 6.31, "learning_rate": 3.21947401377583e-05, "loss": 0.7225, "step": 7462, "task_loss": 1.0143492221832275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5138231515884399, "epoch": 6.31, "learning_rate": 3.2191609267376334e-05, "loss": 0.8309, "step": 7463, "task_loss": 0.30901190638542175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7602196931838989, "epoch": 6.31, "learning_rate": 3.2188478396994366e-05, "loss": 0.7336, "step": 7464, "task_loss": 0.7954379320144653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7193101644515991, "epoch": 6.31, "learning_rate": 3.21853475266124e-05, "loss": 0.9184, "step": 7465, "task_loss": 0.28763407468795776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9027704000473022, "epoch": 6.31, "learning_rate": 3.2182216656230436e-05, "loss": 0.7375, "step": 7466, "task_loss": 0.5247845649719238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.926295816898346, "epoch": 6.31, "learning_rate": 3.217908578584847e-05, "loss": 0.9356, "step": 7467, "task_loss": 0.7811524868011475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.101093053817749, "epoch": 6.31, "learning_rate": 3.21759549154665e-05, "loss": 0.8218, "step": 7468, "task_loss": 1.0289026498794556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6726568937301636, "epoch": 6.31, "learning_rate": 3.217282404508454e-05, "loss": 0.9791, "step": 7469, "task_loss": 0.9150856733322144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8906072378158569, "epoch": 6.31, "learning_rate": 3.216969317470257e-05, "loss": 0.9721, "step": 7470, "task_loss": 0.7563080191612244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6941787004470825, "epoch": 6.32, "learning_rate": 3.21665623043206e-05, "loss": 0.8581, "step": 7471, "task_loss": 1.071071743965149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.000005841255188, "epoch": 6.32, "learning_rate": 3.2163431433938634e-05, "loss": 0.9226, "step": 7472, "task_loss": 0.9647865891456604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6309512257575989, "epoch": 6.32, "learning_rate": 3.216030056355667e-05, "loss": 0.9129, "step": 7473, "task_loss": 1.3836534023284912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9664393663406372, "epoch": 6.32, "learning_rate": 3.2157169693174704e-05, "loss": 0.8609, "step": 7474, "task_loss": 0.9827758073806763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.003150463104248, "epoch": 6.32, "learning_rate": 3.2154038822792736e-05, "loss": 1.1213, "step": 7475, "task_loss": 1.0840613842010498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8131544589996338, "epoch": 6.32, "learning_rate": 3.215090795241077e-05, "loss": 0.8567, "step": 7476, "task_loss": 1.222765326499939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35337790846824646, "epoch": 6.32, "learning_rate": 3.2147777082028806e-05, "loss": 0.6279, "step": 7477, "task_loss": 0.6814793944358826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8773435354232788, "epoch": 6.32, "learning_rate": 3.214464621164684e-05, "loss": 0.8573, "step": 7478, "task_loss": 0.8420472145080566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2544610500335693, "epoch": 6.32, "learning_rate": 3.214151534126487e-05, "loss": 0.7761, "step": 7479, "task_loss": 1.3027994632720947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.685028076171875, "epoch": 6.32, "learning_rate": 3.213838447088291e-05, "loss": 0.8589, "step": 7480, "task_loss": 1.3783290386199951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7589131593704224, "epoch": 6.32, "learning_rate": 3.213525360050094e-05, "loss": 0.7379, "step": 7481, "task_loss": 0.14405246078968048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9581469297409058, "epoch": 6.32, "learning_rate": 3.213212273011897e-05, "loss": 0.9525, "step": 7482, "task_loss": 0.634421169757843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6621205806732178, "epoch": 6.33, "learning_rate": 3.2128991859737004e-05, "loss": 0.5888, "step": 7483, "task_loss": 0.744815468788147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9470300078392029, "epoch": 6.33, "learning_rate": 3.212586098935504e-05, "loss": 0.9186, "step": 7484, "task_loss": 1.0826746225357056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0506389141082764, "epoch": 6.33, "learning_rate": 3.2122730118973074e-05, "loss": 0.6811, "step": 7485, "task_loss": 0.7253723740577698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4846612811088562, "epoch": 6.33, "learning_rate": 3.2119599248591106e-05, "loss": 0.5779, "step": 7486, "task_loss": 0.7904396057128906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7930373549461365, "epoch": 6.33, "learning_rate": 3.2116468378209144e-05, "loss": 0.7512, "step": 7487, "task_loss": 1.115469217300415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7100651860237122, "epoch": 6.33, "learning_rate": 3.2113337507827176e-05, "loss": 0.7237, "step": 7488, "task_loss": 0.32203036546707153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23832671344280243, "epoch": 6.33, "learning_rate": 3.2110206637445215e-05, "loss": 0.5028, "step": 7489, "task_loss": 0.026655249297618866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0970988273620605, "epoch": 6.33, "learning_rate": 3.2107075767063247e-05, "loss": 0.8541, "step": 7490, "task_loss": 0.5515878796577454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.074319839477539, "epoch": 6.33, "learning_rate": 3.210394489668128e-05, "loss": 0.8627, "step": 7491, "task_loss": 0.565703809261322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5564824342727661, "epoch": 6.33, "learning_rate": 3.210081402629932e-05, "loss": 0.5551, "step": 7492, "task_loss": 0.14472678303718567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.899785578250885, "epoch": 6.33, "learning_rate": 3.209768315591735e-05, "loss": 0.9029, "step": 7493, "task_loss": 1.0291674137115479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.450152188539505, "epoch": 6.33, "learning_rate": 3.209455228553538e-05, "loss": 0.6244, "step": 7494, "task_loss": 0.9464607238769531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.703309178352356, "epoch": 6.34, "learning_rate": 3.209142141515342e-05, "loss": 0.8082, "step": 7495, "task_loss": 0.6757782101631165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4824652373790741, "epoch": 6.34, "learning_rate": 3.208829054477145e-05, "loss": 0.705, "step": 7496, "task_loss": 0.32549816370010376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9337853193283081, "epoch": 6.34, "learning_rate": 3.208515967438948e-05, "loss": 0.8718, "step": 7497, "task_loss": 0.8378446102142334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9630289077758789, "epoch": 6.34, "learning_rate": 3.2082028804007514e-05, "loss": 0.96, "step": 7498, "task_loss": 0.5180336833000183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.039438247680664, "epoch": 6.34, "learning_rate": 3.207889793362555e-05, "loss": 0.7804, "step": 7499, "task_loss": 0.708196222782135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3987133502960205, "epoch": 6.34, "learning_rate": 3.2075767063243585e-05, "loss": 0.8874, "step": 7500, "task_loss": 1.122597575187683 }, { "epoch": 6.34, "eval_accuracy": 0.8859009900990099, "eval_loss": 0.5150541663169861, "eval_runtime": 207.5341, "eval_samples_per_second": 121.667, "eval_steps_per_second": 0.954, "step": 7500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6812906265258789, "epoch": 6.34, "learning_rate": 3.2072636192861617e-05, "loss": 0.6515, "step": 7501, "task_loss": 1.3381730318069458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8760244250297546, "epoch": 6.34, "learning_rate": 3.206950532247965e-05, "loss": 0.9656, "step": 7502, "task_loss": 0.3292783200740814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6205652356147766, "epoch": 6.34, "learning_rate": 3.206637445209769e-05, "loss": 0.6707, "step": 7503, "task_loss": 0.5079911947250366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9885909557342529, "epoch": 6.34, "learning_rate": 3.206324358171572e-05, "loss": 0.8933, "step": 7504, "task_loss": 0.49076223373413086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6667767763137817, "epoch": 6.34, "learning_rate": 3.206011271133375e-05, "loss": 0.7226, "step": 7505, "task_loss": 0.54465252161026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9210776686668396, "epoch": 6.34, "learning_rate": 3.205698184095179e-05, "loss": 0.7849, "step": 7506, "task_loss": 1.1767936944961548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6544913053512573, "epoch": 6.35, "learning_rate": 3.205385097056982e-05, "loss": 0.6687, "step": 7507, "task_loss": 0.3415253758430481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.76054447889328, "epoch": 6.35, "learning_rate": 3.205072010018785e-05, "loss": 0.7657, "step": 7508, "task_loss": 0.12800724804401398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6236101984977722, "epoch": 6.35, "learning_rate": 3.2047589229805884e-05, "loss": 0.7418, "step": 7509, "task_loss": 1.2994208335876465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0708763599395752, "epoch": 6.35, "learning_rate": 3.204445835942392e-05, "loss": 0.7895, "step": 7510, "task_loss": 0.8831343054771423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47562164068222046, "epoch": 6.35, "learning_rate": 3.2041327489041955e-05, "loss": 0.6625, "step": 7511, "task_loss": 0.6505880355834961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7993870377540588, "epoch": 6.35, "learning_rate": 3.2038196618659986e-05, "loss": 0.7345, "step": 7512, "task_loss": 0.9117836952209473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5385242104530334, "epoch": 6.35, "learning_rate": 3.203506574827802e-05, "loss": 0.8261, "step": 7513, "task_loss": 0.5952461957931519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6708654761314392, "epoch": 6.35, "learning_rate": 3.203193487789606e-05, "loss": 0.7911, "step": 7514, "task_loss": 0.5502325296401978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4554188847541809, "epoch": 6.35, "learning_rate": 3.202880400751409e-05, "loss": 0.5195, "step": 7515, "task_loss": 0.698704719543457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6352314352989197, "epoch": 6.35, "learning_rate": 3.202567313713212e-05, "loss": 0.6974, "step": 7516, "task_loss": 0.7300940752029419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6675108075141907, "epoch": 6.35, "learning_rate": 3.202254226675016e-05, "loss": 0.6792, "step": 7517, "task_loss": 0.14790290594100952 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6253761053085327, "epoch": 6.35, "learning_rate": 3.201941139636819e-05, "loss": 0.6829, "step": 7518, "task_loss": 0.8725154995918274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.809104859828949, "epoch": 6.36, "learning_rate": 3.201628052598622e-05, "loss": 0.8115, "step": 7519, "task_loss": 0.3916974663734436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6327226161956787, "epoch": 6.36, "learning_rate": 3.2013149655604254e-05, "loss": 0.9309, "step": 7520, "task_loss": 0.5607863068580627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7538047432899475, "epoch": 6.36, "learning_rate": 3.201001878522229e-05, "loss": 0.8589, "step": 7521, "task_loss": 0.9558371901512146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9267904758453369, "epoch": 6.36, "learning_rate": 3.2006887914840325e-05, "loss": 0.8529, "step": 7522, "task_loss": 2.22011137008667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7496516704559326, "epoch": 6.36, "learning_rate": 3.200375704445836e-05, "loss": 0.7505, "step": 7523, "task_loss": 0.4661013185977936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9078434705734253, "epoch": 6.36, "learning_rate": 3.2000626174076395e-05, "loss": 0.965, "step": 7524, "task_loss": 1.5518735647201538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8179677128791809, "epoch": 6.36, "learning_rate": 3.1997495303694433e-05, "loss": 0.9207, "step": 7525, "task_loss": 0.7300041913986206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.718685507774353, "epoch": 6.36, "learning_rate": 3.1994364433312465e-05, "loss": 0.6681, "step": 7526, "task_loss": 1.075690746307373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5531846284866333, "epoch": 6.36, "learning_rate": 3.19912335629305e-05, "loss": 0.6441, "step": 7527, "task_loss": 0.3312350809574127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.664029061794281, "epoch": 6.36, "learning_rate": 3.1988102692548536e-05, "loss": 0.7701, "step": 7528, "task_loss": 0.5269252061843872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1983354091644287, "epoch": 6.36, "learning_rate": 3.198497182216657e-05, "loss": 0.823, "step": 7529, "task_loss": 1.5140810012817383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.331059455871582, "epoch": 6.36, "learning_rate": 3.19818409517846e-05, "loss": 0.8758, "step": 7530, "task_loss": 0.5454875826835632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4719383120536804, "epoch": 6.37, "learning_rate": 3.197871008140263e-05, "loss": 0.4857, "step": 7531, "task_loss": 0.15604467689990997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40381938219070435, "epoch": 6.37, "learning_rate": 3.197557921102067e-05, "loss": 0.8323, "step": 7532, "task_loss": 0.5693575739860535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9504426717758179, "epoch": 6.37, "learning_rate": 3.19724483406387e-05, "loss": 0.8622, "step": 7533, "task_loss": 0.8538603782653809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9107181429862976, "epoch": 6.37, "learning_rate": 3.196931747025673e-05, "loss": 0.6421, "step": 7534, "task_loss": 0.21300949156284332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5069432258605957, "epoch": 6.37, "learning_rate": 3.1966186599874765e-05, "loss": 0.7239, "step": 7535, "task_loss": 0.9375752210617065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8242259621620178, "epoch": 6.37, "learning_rate": 3.19630557294928e-05, "loss": 0.9482, "step": 7536, "task_loss": 0.4123106896877289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6034528017044067, "epoch": 6.37, "learning_rate": 3.1959924859110835e-05, "loss": 0.7175, "step": 7537, "task_loss": 0.1559513360261917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2128949165344238, "epoch": 6.37, "learning_rate": 3.195679398872887e-05, "loss": 0.7635, "step": 7538, "task_loss": 1.4280129671096802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0972540378570557, "epoch": 6.37, "learning_rate": 3.19536631183469e-05, "loss": 0.9381, "step": 7539, "task_loss": 1.247320294380188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7718982100486755, "epoch": 6.37, "learning_rate": 3.195053224796494e-05, "loss": 0.7915, "step": 7540, "task_loss": 0.5747050046920776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4294615089893341, "epoch": 6.37, "learning_rate": 3.194740137758297e-05, "loss": 0.5128, "step": 7541, "task_loss": 0.866732120513916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.603005051612854, "epoch": 6.38, "learning_rate": 3.1944270507201e-05, "loss": 0.685, "step": 7542, "task_loss": 0.16332091391086578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6379014849662781, "epoch": 6.38, "learning_rate": 3.194113963681904e-05, "loss": 0.7306, "step": 7543, "task_loss": 0.636384904384613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.671053946018219, "epoch": 6.38, "learning_rate": 3.193800876643707e-05, "loss": 0.6512, "step": 7544, "task_loss": 0.5821993947029114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.662284791469574, "epoch": 6.38, "learning_rate": 3.19348778960551e-05, "loss": 0.8411, "step": 7545, "task_loss": 0.8194592595100403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9783964157104492, "epoch": 6.38, "learning_rate": 3.1931747025673135e-05, "loss": 0.8269, "step": 7546, "task_loss": 0.4949536621570587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6581668853759766, "epoch": 6.38, "learning_rate": 3.192861615529117e-05, "loss": 0.6584, "step": 7547, "task_loss": 1.1025019884109497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8151091933250427, "epoch": 6.38, "learning_rate": 3.1925485284909205e-05, "loss": 0.6556, "step": 7548, "task_loss": 1.0013774633407593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.014056921005249, "epoch": 6.38, "learning_rate": 3.192235441452724e-05, "loss": 0.906, "step": 7549, "task_loss": 0.8731791973114014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9717416763305664, "epoch": 6.38, "learning_rate": 3.191922354414527e-05, "loss": 0.5802, "step": 7550, "task_loss": 0.5659821629524231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0729018449783325, "epoch": 6.38, "learning_rate": 3.191609267376331e-05, "loss": 0.8608, "step": 7551, "task_loss": 1.0094596147537231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5544007420539856, "epoch": 6.38, "learning_rate": 3.191296180338134e-05, "loss": 0.7297, "step": 7552, "task_loss": 0.8668425679206848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7458093166351318, "epoch": 6.38, "learning_rate": 3.190983093299937e-05, "loss": 0.6445, "step": 7553, "task_loss": 0.6147090196609497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5068514347076416, "epoch": 6.39, "learning_rate": 3.190670006261741e-05, "loss": 0.737, "step": 7554, "task_loss": 0.9087813496589661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4570161998271942, "epoch": 6.39, "learning_rate": 3.190356919223544e-05, "loss": 0.7218, "step": 7555, "task_loss": 0.6515787839889526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7966903448104858, "epoch": 6.39, "learning_rate": 3.190043832185348e-05, "loss": 0.8077, "step": 7556, "task_loss": 0.8880345821380615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6884174346923828, "epoch": 6.39, "learning_rate": 3.189730745147151e-05, "loss": 0.8633, "step": 7557, "task_loss": 0.9999603033065796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7360155582427979, "epoch": 6.39, "learning_rate": 3.189417658108954e-05, "loss": 0.8232, "step": 7558, "task_loss": 1.0479848384857178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5156943798065186, "epoch": 6.39, "learning_rate": 3.189104571070758e-05, "loss": 0.8268, "step": 7559, "task_loss": 0.4795456826686859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.024117350578308, "epoch": 6.39, "learning_rate": 3.1887914840325614e-05, "loss": 0.7708, "step": 7560, "task_loss": 1.1907360553741455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9101576805114746, "epoch": 6.39, "learning_rate": 3.1884783969943645e-05, "loss": 0.907, "step": 7561, "task_loss": 1.5590813159942627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8851885795593262, "epoch": 6.39, "learning_rate": 3.1881653099561684e-05, "loss": 0.6374, "step": 7562, "task_loss": 1.0053704977035522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8995898962020874, "epoch": 6.39, "learning_rate": 3.1878522229179716e-05, "loss": 0.9784, "step": 7563, "task_loss": 0.4449736177921295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8887133598327637, "epoch": 6.39, "learning_rate": 3.187539135879775e-05, "loss": 0.8521, "step": 7564, "task_loss": 1.1701650619506836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9051982164382935, "epoch": 6.39, "learning_rate": 3.1872260488415786e-05, "loss": 0.8651, "step": 7565, "task_loss": 1.275010108947754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2912986278533936, "epoch": 6.4, "learning_rate": 3.186912961803382e-05, "loss": 0.9843, "step": 7566, "task_loss": 1.215599775314331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0953575372695923, "epoch": 6.4, "learning_rate": 3.186599874765185e-05, "loss": 0.8963, "step": 7567, "task_loss": 0.5515112280845642 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1782463788986206, "epoch": 6.4, "learning_rate": 3.186286787726988e-05, "loss": 0.8748, "step": 7568, "task_loss": 0.8249827027320862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1353119611740112, "epoch": 6.4, "learning_rate": 3.185973700688792e-05, "loss": 1.3275, "step": 7569, "task_loss": 0.7329065799713135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5609924793243408, "epoch": 6.4, "learning_rate": 3.185660613650595e-05, "loss": 0.7357, "step": 7570, "task_loss": 0.6496579647064209 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.036138892173767, "epoch": 6.4, "learning_rate": 3.1853475266123983e-05, "loss": 0.9188, "step": 7571, "task_loss": 0.9693177342414856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8939152956008911, "epoch": 6.4, "learning_rate": 3.1850344395742015e-05, "loss": 0.8527, "step": 7572, "task_loss": 0.8829478025436401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5028553009033203, "epoch": 6.4, "learning_rate": 3.1847213525360054e-05, "loss": 0.8382, "step": 7573, "task_loss": 0.6834889650344849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0788118839263916, "epoch": 6.4, "learning_rate": 3.1844082654978086e-05, "loss": 0.8525, "step": 7574, "task_loss": 1.0270379781723022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0837304592132568, "epoch": 6.4, "learning_rate": 3.184095178459612e-05, "loss": 0.966, "step": 7575, "task_loss": 1.3593568801879883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7939642667770386, "epoch": 6.4, "learning_rate": 3.183782091421415e-05, "loss": 0.7942, "step": 7576, "task_loss": 0.5198790431022644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5254908800125122, "epoch": 6.4, "learning_rate": 3.183469004383219e-05, "loss": 0.7755, "step": 7577, "task_loss": 0.6223665475845337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4680120646953583, "epoch": 6.41, "learning_rate": 3.183155917345022e-05, "loss": 0.6707, "step": 7578, "task_loss": 0.7587383389472961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6698700189590454, "epoch": 6.41, "learning_rate": 3.182842830306825e-05, "loss": 0.6625, "step": 7579, "task_loss": 0.6195158958435059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.351672887802124, "epoch": 6.41, "learning_rate": 3.182529743268629e-05, "loss": 0.9536, "step": 7580, "task_loss": 1.8347722291946411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6328564286231995, "epoch": 6.41, "learning_rate": 3.182216656230432e-05, "loss": 0.6824, "step": 7581, "task_loss": 0.5620028376579285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4479998052120209, "epoch": 6.41, "learning_rate": 3.1819035691922353e-05, "loss": 0.583, "step": 7582, "task_loss": 0.7044345736503601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3250396251678467, "epoch": 6.41, "learning_rate": 3.1815904821540385e-05, "loss": 0.9398, "step": 7583, "task_loss": 1.016648292541504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0218809843063354, "epoch": 6.41, "learning_rate": 3.1812773951158424e-05, "loss": 0.7267, "step": 7584, "task_loss": 0.6241471171379089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8318659067153931, "epoch": 6.41, "learning_rate": 3.1809643080776456e-05, "loss": 0.7367, "step": 7585, "task_loss": 1.2073498964309692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6789474487304688, "epoch": 6.41, "learning_rate": 3.180651221039449e-05, "loss": 0.825, "step": 7586, "task_loss": 0.9279670715332031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8696539402008057, "epoch": 6.41, "learning_rate": 3.180338134001252e-05, "loss": 0.8243, "step": 7587, "task_loss": 1.3673352003097534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8100711107254028, "epoch": 6.41, "learning_rate": 3.180025046963056e-05, "loss": 0.6907, "step": 7588, "task_loss": 0.9283299446105957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8278863430023193, "epoch": 6.41, "learning_rate": 3.179711959924859e-05, "loss": 0.8731, "step": 7589, "task_loss": 1.0366407632827759 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8958315849304199, "epoch": 6.42, "learning_rate": 3.179398872886663e-05, "loss": 0.7144, "step": 7590, "task_loss": 1.1317579746246338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4662649929523468, "epoch": 6.42, "learning_rate": 3.179085785848466e-05, "loss": 0.6585, "step": 7591, "task_loss": 0.9184021353721619 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5793616771697998, "epoch": 6.42, "learning_rate": 3.17877269881027e-05, "loss": 0.8722, "step": 7592, "task_loss": 0.34505495429039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.768463134765625, "epoch": 6.42, "learning_rate": 3.178459611772073e-05, "loss": 0.7462, "step": 7593, "task_loss": 0.5643362998962402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8710842132568359, "epoch": 6.42, "learning_rate": 3.178146524733876e-05, "loss": 0.9484, "step": 7594, "task_loss": 1.6717013120651245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6361879110336304, "epoch": 6.42, "learning_rate": 3.17783343769568e-05, "loss": 0.6518, "step": 7595, "task_loss": 1.3739758729934692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6814865469932556, "epoch": 6.42, "learning_rate": 3.177520350657483e-05, "loss": 0.9102, "step": 7596, "task_loss": 0.4365799129009247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2005712985992432, "epoch": 6.42, "learning_rate": 3.1772072636192864e-05, "loss": 1.0648, "step": 7597, "task_loss": 1.5192835330963135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44751280546188354, "epoch": 6.42, "learning_rate": 3.1768941765810896e-05, "loss": 0.7189, "step": 7598, "task_loss": 0.419102281332016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.708838939666748, "epoch": 6.42, "learning_rate": 3.1765810895428934e-05, "loss": 0.6969, "step": 7599, "task_loss": 1.6274259090423584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0263341665267944, "epoch": 6.42, "learning_rate": 3.1762680025046966e-05, "loss": 0.9435, "step": 7600, "task_loss": 1.2311675548553467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9407000541687012, "epoch": 6.42, "learning_rate": 3.1759549154665e-05, "loss": 0.538, "step": 7601, "task_loss": 2.0963380336761475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5310513377189636, "epoch": 6.43, "learning_rate": 3.1756418284283036e-05, "loss": 0.5617, "step": 7602, "task_loss": 0.5020774006843567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8507021069526672, "epoch": 6.43, "learning_rate": 3.175328741390107e-05, "loss": 0.979, "step": 7603, "task_loss": 0.4415615499019623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7347816824913025, "epoch": 6.43, "learning_rate": 3.17501565435191e-05, "loss": 0.777, "step": 7604, "task_loss": 0.9635474681854248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.008908987045288, "epoch": 6.43, "learning_rate": 3.174702567313713e-05, "loss": 0.9819, "step": 7605, "task_loss": 1.0376302003860474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6438332200050354, "epoch": 6.43, "learning_rate": 3.174389480275517e-05, "loss": 0.6052, "step": 7606, "task_loss": 0.6926084160804749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9981083869934082, "epoch": 6.43, "learning_rate": 3.17407639323732e-05, "loss": 0.7184, "step": 7607, "task_loss": 1.5815067291259766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8608183860778809, "epoch": 6.43, "learning_rate": 3.1737633061991234e-05, "loss": 0.6853, "step": 7608, "task_loss": 0.8194525241851807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9944292306900024, "epoch": 6.43, "learning_rate": 3.1734502191609266e-05, "loss": 0.7575, "step": 7609, "task_loss": 1.4851810932159424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5872108936309814, "epoch": 6.43, "learning_rate": 3.1731371321227304e-05, "loss": 0.8011, "step": 7610, "task_loss": 1.1443538665771484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.566986083984375, "epoch": 6.43, "learning_rate": 3.1728240450845336e-05, "loss": 0.6196, "step": 7611, "task_loss": 1.6337507963180542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5238946080207825, "epoch": 6.43, "learning_rate": 3.172510958046337e-05, "loss": 0.7675, "step": 7612, "task_loss": 0.8925831317901611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9375196695327759, "epoch": 6.44, "learning_rate": 3.17219787100814e-05, "loss": 0.7234, "step": 7613, "task_loss": 0.8456355929374695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9920738339424133, "epoch": 6.44, "learning_rate": 3.171884783969944e-05, "loss": 0.9597, "step": 7614, "task_loss": 1.1556593179702759 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9688533544540405, "epoch": 6.44, "learning_rate": 3.171571696931747e-05, "loss": 1.0826, "step": 7615, "task_loss": 0.993739902973175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7793323993682861, "epoch": 6.44, "learning_rate": 3.17125860989355e-05, "loss": 0.7977, "step": 7616, "task_loss": 1.2173138856887817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6903035044670105, "epoch": 6.44, "learning_rate": 3.170945522855354e-05, "loss": 0.7685, "step": 7617, "task_loss": 0.6263490915298462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.941961944103241, "epoch": 6.44, "learning_rate": 3.170632435817157e-05, "loss": 0.8258, "step": 7618, "task_loss": 1.0879722833633423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5605208873748779, "epoch": 6.44, "learning_rate": 3.1703193487789604e-05, "loss": 0.7409, "step": 7619, "task_loss": 0.9493563771247864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.666745662689209, "epoch": 6.44, "learning_rate": 3.1700062617407636e-05, "loss": 0.7948, "step": 7620, "task_loss": 1.4253982305526733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8872570395469666, "epoch": 6.44, "learning_rate": 3.1696931747025674e-05, "loss": 0.6138, "step": 7621, "task_loss": 0.4962390959262848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6745941638946533, "epoch": 6.44, "learning_rate": 3.1693800876643706e-05, "loss": 0.7934, "step": 7622, "task_loss": 1.2599090337753296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5424239635467529, "epoch": 6.44, "learning_rate": 3.1690670006261745e-05, "loss": 0.8059, "step": 7623, "task_loss": 1.1501002311706543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9772013425827026, "epoch": 6.44, "learning_rate": 3.1687539135879776e-05, "loss": 0.6765, "step": 7624, "task_loss": 0.5244808793067932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5627895593643188, "epoch": 6.45, "learning_rate": 3.168440826549781e-05, "loss": 0.9064, "step": 7625, "task_loss": 0.6678546071052551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6590093374252319, "epoch": 6.45, "learning_rate": 3.168127739511585e-05, "loss": 0.617, "step": 7626, "task_loss": 0.8872880339622498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8041719198226929, "epoch": 6.45, "learning_rate": 3.167814652473388e-05, "loss": 0.8297, "step": 7627, "task_loss": 1.1659132242202759 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9974463582038879, "epoch": 6.45, "learning_rate": 3.167501565435192e-05, "loss": 0.8277, "step": 7628, "task_loss": 1.4033927917480469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6049997806549072, "epoch": 6.45, "learning_rate": 3.167188478396995e-05, "loss": 0.6341, "step": 7629, "task_loss": 0.5362334847450256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7042418122291565, "epoch": 6.45, "learning_rate": 3.166875391358798e-05, "loss": 0.7639, "step": 7630, "task_loss": 0.37747564911842346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5127112865447998, "epoch": 6.45, "learning_rate": 3.166562304320601e-05, "loss": 0.6196, "step": 7631, "task_loss": 0.3362566828727722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6184456944465637, "epoch": 6.45, "learning_rate": 3.166249217282405e-05, "loss": 0.8701, "step": 7632, "task_loss": 0.8504531979560852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8691563606262207, "epoch": 6.45, "learning_rate": 3.165936130244208e-05, "loss": 0.616, "step": 7633, "task_loss": 0.26721906661987305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7801335453987122, "epoch": 6.45, "learning_rate": 3.1656230432060114e-05, "loss": 0.725, "step": 7634, "task_loss": 0.9962859749794006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7143505215644836, "epoch": 6.45, "learning_rate": 3.1653099561678146e-05, "loss": 0.7898, "step": 7635, "task_loss": 0.6744329929351807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9348138570785522, "epoch": 6.45, "learning_rate": 3.1649968691296185e-05, "loss": 0.7609, "step": 7636, "task_loss": 0.6552948355674744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4329158365726471, "epoch": 6.46, "learning_rate": 3.1646837820914217e-05, "loss": 0.712, "step": 7637, "task_loss": 0.029787909239530563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.827167809009552, "epoch": 6.46, "learning_rate": 3.164370695053225e-05, "loss": 0.7342, "step": 7638, "task_loss": 1.9862949848175049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9012694358825684, "epoch": 6.46, "learning_rate": 3.164057608015029e-05, "loss": 1.1292, "step": 7639, "task_loss": 0.3677302300930023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4041471481323242, "epoch": 6.46, "learning_rate": 3.163744520976832e-05, "loss": 0.6802, "step": 7640, "task_loss": 0.6890875697135925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9831558465957642, "epoch": 6.46, "learning_rate": 3.163431433938635e-05, "loss": 0.7098, "step": 7641, "task_loss": 1.2324531078338623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1411819458007812, "epoch": 6.46, "learning_rate": 3.163118346900438e-05, "loss": 0.7693, "step": 7642, "task_loss": 0.5828860402107239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1715357303619385, "epoch": 6.46, "learning_rate": 3.162805259862242e-05, "loss": 0.988, "step": 7643, "task_loss": 0.8240180015563965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6314629316329956, "epoch": 6.46, "learning_rate": 3.162492172824045e-05, "loss": 0.7242, "step": 7644, "task_loss": 0.5874351263046265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.830039381980896, "epoch": 6.46, "learning_rate": 3.1621790857858484e-05, "loss": 0.7398, "step": 7645, "task_loss": 0.841687798500061 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0663511753082275, "epoch": 6.46, "learning_rate": 3.1618659987476516e-05, "loss": 0.9648, "step": 7646, "task_loss": 0.630853533744812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9221290349960327, "epoch": 6.46, "learning_rate": 3.1615529117094555e-05, "loss": 0.7259, "step": 7647, "task_loss": 0.5431566834449768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8596539497375488, "epoch": 6.46, "learning_rate": 3.1612398246712586e-05, "loss": 0.595, "step": 7648, "task_loss": 0.20980525016784668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1265596151351929, "epoch": 6.47, "learning_rate": 3.160926737633062e-05, "loss": 0.758, "step": 7649, "task_loss": 0.5243829488754272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5225988626480103, "epoch": 6.47, "learning_rate": 3.160613650594865e-05, "loss": 0.7085, "step": 7650, "task_loss": 0.44991764426231384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6893805861473083, "epoch": 6.47, "learning_rate": 3.160300563556669e-05, "loss": 0.7527, "step": 7651, "task_loss": 0.5876893997192383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7877744436264038, "epoch": 6.47, "learning_rate": 3.159987476518472e-05, "loss": 0.7839, "step": 7652, "task_loss": 0.9410611391067505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4869353473186493, "epoch": 6.47, "learning_rate": 3.159674389480275e-05, "loss": 0.704, "step": 7653, "task_loss": 0.4039267897605896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9674853682518005, "epoch": 6.47, "learning_rate": 3.159361302442079e-05, "loss": 0.8921, "step": 7654, "task_loss": 0.44126829504966736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0483808517456055, "epoch": 6.47, "learning_rate": 3.159048215403882e-05, "loss": 0.918, "step": 7655, "task_loss": 1.101696252822876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9213892221450806, "epoch": 6.47, "learning_rate": 3.1587351283656854e-05, "loss": 0.6684, "step": 7656, "task_loss": 2.034076452255249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0082134008407593, "epoch": 6.47, "learning_rate": 3.158422041327489e-05, "loss": 1.0081, "step": 7657, "task_loss": 1.1376502513885498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0821309089660645, "epoch": 6.47, "learning_rate": 3.1581089542892925e-05, "loss": 0.9256, "step": 7658, "task_loss": 1.708998441696167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5553170442581177, "epoch": 6.47, "learning_rate": 3.157795867251096e-05, "loss": 0.5726, "step": 7659, "task_loss": 0.7550135850906372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5976896286010742, "epoch": 6.47, "learning_rate": 3.1574827802128995e-05, "loss": 0.7492, "step": 7660, "task_loss": 0.4778181314468384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46948477625846863, "epoch": 6.48, "learning_rate": 3.157169693174703e-05, "loss": 0.5647, "step": 7661, "task_loss": 1.0257878303527832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49179941415786743, "epoch": 6.48, "learning_rate": 3.1568566061365065e-05, "loss": 0.5881, "step": 7662, "task_loss": 0.4419666826725006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46895888447761536, "epoch": 6.48, "learning_rate": 3.15654351909831e-05, "loss": 0.5469, "step": 7663, "task_loss": 0.2538011372089386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8366434574127197, "epoch": 6.48, "learning_rate": 3.156230432060113e-05, "loss": 0.9224, "step": 7664, "task_loss": 1.8677151203155518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0663557052612305, "epoch": 6.48, "learning_rate": 3.155917345021917e-05, "loss": 0.8315, "step": 7665, "task_loss": 1.2807142734527588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.749636173248291, "epoch": 6.48, "learning_rate": 3.15560425798372e-05, "loss": 0.7416, "step": 7666, "task_loss": 0.789298415184021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6270514726638794, "epoch": 6.48, "learning_rate": 3.155291170945523e-05, "loss": 0.5969, "step": 7667, "task_loss": 0.40874233841896057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9139413237571716, "epoch": 6.48, "learning_rate": 3.154978083907326e-05, "loss": 0.6524, "step": 7668, "task_loss": 0.9401371479034424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5629290342330933, "epoch": 6.48, "learning_rate": 3.15466499686913e-05, "loss": 0.6412, "step": 7669, "task_loss": 0.5256693959236145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7415757179260254, "epoch": 6.48, "learning_rate": 3.154351909830933e-05, "loss": 0.6117, "step": 7670, "task_loss": 0.9132529497146606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7528166770935059, "epoch": 6.48, "learning_rate": 3.1540388227927365e-05, "loss": 1.0208, "step": 7671, "task_loss": 0.6377136707305908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8992076516151428, "epoch": 6.48, "learning_rate": 3.15372573575454e-05, "loss": 0.8354, "step": 7672, "task_loss": 1.1065391302108765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4487900733947754, "epoch": 6.49, "learning_rate": 3.1534126487163435e-05, "loss": 0.6187, "step": 7673, "task_loss": 0.244768425822258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0532032251358032, "epoch": 6.49, "learning_rate": 3.153099561678147e-05, "loss": 0.9628, "step": 7674, "task_loss": 0.5893939137458801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6598324775695801, "epoch": 6.49, "learning_rate": 3.15278647463995e-05, "loss": 0.8505, "step": 7675, "task_loss": 0.6305468678474426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6039859652519226, "epoch": 6.49, "learning_rate": 3.152473387601754e-05, "loss": 0.7191, "step": 7676, "task_loss": 0.5010570883750916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9106802344322205, "epoch": 6.49, "learning_rate": 3.152160300563557e-05, "loss": 0.8253, "step": 7677, "task_loss": 0.5663142800331116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7352885007858276, "epoch": 6.49, "learning_rate": 3.15184721352536e-05, "loss": 0.8824, "step": 7678, "task_loss": 0.5396860241889954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2876198291778564, "epoch": 6.49, "learning_rate": 3.151534126487163e-05, "loss": 0.9093, "step": 7679, "task_loss": 1.0494147539138794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6980348825454712, "epoch": 6.49, "learning_rate": 3.151221039448967e-05, "loss": 0.7361, "step": 7680, "task_loss": 0.9117630124092102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46353551745414734, "epoch": 6.49, "learning_rate": 3.15090795241077e-05, "loss": 0.6334, "step": 7681, "task_loss": 0.36286160349845886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7133407592773438, "epoch": 6.49, "learning_rate": 3.1505948653725735e-05, "loss": 0.8921, "step": 7682, "task_loss": 0.8625688552856445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.513655424118042, "epoch": 6.49, "learning_rate": 3.1502817783343767e-05, "loss": 0.7353, "step": 7683, "task_loss": 0.10599762201309204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7369003295898438, "epoch": 6.5, "learning_rate": 3.1499686912961805e-05, "loss": 0.776, "step": 7684, "task_loss": 1.2413831949234009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8954565525054932, "epoch": 6.5, "learning_rate": 3.149655604257984e-05, "loss": 0.7492, "step": 7685, "task_loss": 1.06400728225708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6053361296653748, "epoch": 6.5, "learning_rate": 3.149342517219787e-05, "loss": 0.6993, "step": 7686, "task_loss": 0.5562010407447815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9251982569694519, "epoch": 6.5, "learning_rate": 3.14902943018159e-05, "loss": 0.8112, "step": 7687, "task_loss": 0.9100286960601807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7710068225860596, "epoch": 6.5, "learning_rate": 3.148716343143394e-05, "loss": 0.7348, "step": 7688, "task_loss": 0.6958635449409485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33464211225509644, "epoch": 6.5, "learning_rate": 3.148403256105197e-05, "loss": 0.6644, "step": 7689, "task_loss": 0.9942583441734314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8195614814758301, "epoch": 6.5, "learning_rate": 3.148090169067001e-05, "loss": 0.7457, "step": 7690, "task_loss": 1.2663259506225586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2742037773132324, "epoch": 6.5, "learning_rate": 3.147777082028804e-05, "loss": 0.7492, "step": 7691, "task_loss": 1.9614146947860718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4903832674026489, "epoch": 6.5, "learning_rate": 3.147463994990607e-05, "loss": 0.7707, "step": 7692, "task_loss": 0.1800331324338913 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9774162769317627, "epoch": 6.5, "learning_rate": 3.147150907952411e-05, "loss": 0.7508, "step": 7693, "task_loss": 0.8215064406394958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4994962811470032, "epoch": 6.5, "learning_rate": 3.146837820914214e-05, "loss": 0.5546, "step": 7694, "task_loss": 0.4344640374183655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5589737892150879, "epoch": 6.5, "learning_rate": 3.146524733876018e-05, "loss": 0.7766, "step": 7695, "task_loss": 0.491283118724823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48162007331848145, "epoch": 6.51, "learning_rate": 3.1462116468378214e-05, "loss": 0.8429, "step": 7696, "task_loss": 1.0376139879226685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.830075740814209, "epoch": 6.51, "learning_rate": 3.1458985597996245e-05, "loss": 0.8876, "step": 7697, "task_loss": 1.3519269227981567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5865181684494019, "epoch": 6.51, "learning_rate": 3.145585472761428e-05, "loss": 0.7435, "step": 7698, "task_loss": 1.0197758674621582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5128817558288574, "epoch": 6.51, "learning_rate": 3.1452723857232316e-05, "loss": 0.7003, "step": 7699, "task_loss": 0.3518710732460022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5181410312652588, "epoch": 6.51, "learning_rate": 3.144959298685035e-05, "loss": 0.5594, "step": 7700, "task_loss": 0.4581563174724579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0079748630523682, "epoch": 6.51, "learning_rate": 3.144646211646838e-05, "loss": 0.9808, "step": 7701, "task_loss": 0.9497845768928528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7430627346038818, "epoch": 6.51, "learning_rate": 3.144333124608642e-05, "loss": 0.6471, "step": 7702, "task_loss": 1.047404170036316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5883327722549438, "epoch": 6.51, "learning_rate": 3.144020037570445e-05, "loss": 0.781, "step": 7703, "task_loss": 0.6532547473907471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.750178873538971, "epoch": 6.51, "learning_rate": 3.143706950532248e-05, "loss": 0.9124, "step": 7704, "task_loss": 0.9133567810058594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9273213148117065, "epoch": 6.51, "learning_rate": 3.143393863494051e-05, "loss": 0.7678, "step": 7705, "task_loss": 0.7650607824325562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7744037508964539, "epoch": 6.51, "learning_rate": 3.143080776455855e-05, "loss": 0.6425, "step": 7706, "task_loss": 0.3944385051727295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5522055625915527, "epoch": 6.51, "learning_rate": 3.1427676894176584e-05, "loss": 0.6028, "step": 7707, "task_loss": 0.7533243298530579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8796084523200989, "epoch": 6.52, "learning_rate": 3.1424546023794615e-05, "loss": 0.9175, "step": 7708, "task_loss": 1.4236035346984863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0031269788742065, "epoch": 6.52, "learning_rate": 3.142141515341265e-05, "loss": 0.6703, "step": 7709, "task_loss": 0.8942893147468567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8623170852661133, "epoch": 6.52, "learning_rate": 3.1418284283030686e-05, "loss": 0.7519, "step": 7710, "task_loss": 1.4656397104263306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9584095478057861, "epoch": 6.52, "learning_rate": 3.141515341264872e-05, "loss": 0.9479, "step": 7711, "task_loss": 0.31307482719421387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9017452001571655, "epoch": 6.52, "learning_rate": 3.141202254226675e-05, "loss": 0.9284, "step": 7712, "task_loss": 0.9922993183135986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6785460710525513, "epoch": 6.52, "learning_rate": 3.140889167188479e-05, "loss": 0.673, "step": 7713, "task_loss": 0.4498891830444336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8160339593887329, "epoch": 6.52, "learning_rate": 3.140576080150282e-05, "loss": 0.7005, "step": 7714, "task_loss": 0.5157793164253235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6815577745437622, "epoch": 6.52, "learning_rate": 3.140262993112085e-05, "loss": 0.7451, "step": 7715, "task_loss": 0.7760279178619385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8612475395202637, "epoch": 6.52, "learning_rate": 3.139949906073888e-05, "loss": 0.7164, "step": 7716, "task_loss": 0.5599210858345032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32469642162323, "epoch": 6.52, "learning_rate": 3.139636819035692e-05, "loss": 0.71, "step": 7717, "task_loss": 0.15821290016174316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6784681081771851, "epoch": 6.52, "learning_rate": 3.1393237319974953e-05, "loss": 0.8183, "step": 7718, "task_loss": 0.47446200251579285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0839258432388306, "epoch": 6.52, "learning_rate": 3.1390106449592985e-05, "loss": 0.9523, "step": 7719, "task_loss": 1.6752817630767822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6625052094459534, "epoch": 6.53, "learning_rate": 3.138697557921102e-05, "loss": 0.7205, "step": 7720, "task_loss": 1.0914117097854614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9562808275222778, "epoch": 6.53, "learning_rate": 3.1383844708829056e-05, "loss": 0.9126, "step": 7721, "task_loss": 1.081510066986084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8557871580123901, "epoch": 6.53, "learning_rate": 3.138071383844709e-05, "loss": 0.9849, "step": 7722, "task_loss": 1.0480533838272095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3878188133239746, "epoch": 6.53, "learning_rate": 3.137758296806512e-05, "loss": 0.5018, "step": 7723, "task_loss": 0.17134755849838257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7676882147789001, "epoch": 6.53, "learning_rate": 3.137445209768316e-05, "loss": 0.7697, "step": 7724, "task_loss": 0.9138239026069641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0533292293548584, "epoch": 6.53, "learning_rate": 3.137132122730119e-05, "loss": 0.9483, "step": 7725, "task_loss": 1.0278276205062866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8130332231521606, "epoch": 6.53, "learning_rate": 3.136819035691923e-05, "loss": 0.6997, "step": 7726, "task_loss": 0.39819976687431335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0263824462890625, "epoch": 6.53, "learning_rate": 3.136505948653726e-05, "loss": 0.8113, "step": 7727, "task_loss": 0.9160223603248596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8794245719909668, "epoch": 6.53, "learning_rate": 3.136192861615529e-05, "loss": 0.6333, "step": 7728, "task_loss": 0.6194890737533569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4026062488555908, "epoch": 6.53, "learning_rate": 3.135879774577333e-05, "loss": 0.5676, "step": 7729, "task_loss": 1.026450753211975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7928111553192139, "epoch": 6.53, "learning_rate": 3.135566687539136e-05, "loss": 0.7768, "step": 7730, "task_loss": 0.20969247817993164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7634894251823425, "epoch": 6.53, "learning_rate": 3.1352536005009394e-05, "loss": 0.7082, "step": 7731, "task_loss": 0.4123113751411438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5964085459709167, "epoch": 6.54, "learning_rate": 3.134940513462743e-05, "loss": 0.7565, "step": 7732, "task_loss": 0.8608435392379761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6422088742256165, "epoch": 6.54, "learning_rate": 3.1346274264245464e-05, "loss": 0.6775, "step": 7733, "task_loss": 0.4063192903995514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49834144115448, "epoch": 6.54, "learning_rate": 3.1343143393863496e-05, "loss": 0.6291, "step": 7734, "task_loss": 0.2229415476322174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5843074917793274, "epoch": 6.54, "learning_rate": 3.134001252348153e-05, "loss": 0.7315, "step": 7735, "task_loss": 0.4842022955417633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5895148515701294, "epoch": 6.54, "learning_rate": 3.1336881653099566e-05, "loss": 0.5233, "step": 7736, "task_loss": 1.0704126358032227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8733500242233276, "epoch": 6.54, "learning_rate": 3.13337507827176e-05, "loss": 0.7321, "step": 7737, "task_loss": 0.6463457942008972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0556710958480835, "epoch": 6.54, "learning_rate": 3.133061991233563e-05, "loss": 0.6531, "step": 7738, "task_loss": 0.5822784900665283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8000163435935974, "epoch": 6.54, "learning_rate": 3.132748904195367e-05, "loss": 0.6126, "step": 7739, "task_loss": 0.7553963661193848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5292446613311768, "epoch": 6.54, "learning_rate": 3.13243581715717e-05, "loss": 0.8158, "step": 7740, "task_loss": 0.6950765252113342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7849242687225342, "epoch": 6.54, "learning_rate": 3.132122730118973e-05, "loss": 0.7061, "step": 7741, "task_loss": 1.5388906002044678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3946203887462616, "epoch": 6.54, "learning_rate": 3.1318096430807764e-05, "loss": 0.5342, "step": 7742, "task_loss": 0.33375680446624756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4591425955295563, "epoch": 6.54, "learning_rate": 3.13149655604258e-05, "loss": 0.703, "step": 7743, "task_loss": 0.42968741059303284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9310051202774048, "epoch": 6.55, "learning_rate": 3.1311834690043834e-05, "loss": 0.6976, "step": 7744, "task_loss": 0.33928775787353516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7622404098510742, "epoch": 6.55, "learning_rate": 3.1308703819661866e-05, "loss": 0.8202, "step": 7745, "task_loss": 0.5207499265670776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1908997297286987, "epoch": 6.55, "learning_rate": 3.13055729492799e-05, "loss": 1.1874, "step": 7746, "task_loss": 0.6128979921340942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0953774452209473, "epoch": 6.55, "learning_rate": 3.1302442078897936e-05, "loss": 0.917, "step": 7747, "task_loss": 1.0828208923339844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7030486464500427, "epoch": 6.55, "learning_rate": 3.129931120851597e-05, "loss": 0.9153, "step": 7748, "task_loss": 1.5283172130584717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8423489332199097, "epoch": 6.55, "learning_rate": 3.1296180338134e-05, "loss": 0.9133, "step": 7749, "task_loss": 0.6184724569320679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.592064380645752, "epoch": 6.55, "learning_rate": 3.129304946775204e-05, "loss": 0.6677, "step": 7750, "task_loss": 1.16681969165802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6282076835632324, "epoch": 6.55, "learning_rate": 3.128991859737007e-05, "loss": 0.7233, "step": 7751, "task_loss": 0.34013813734054565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1212778091430664, "epoch": 6.55, "learning_rate": 3.12867877269881e-05, "loss": 0.8862, "step": 7752, "task_loss": 0.885981559753418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6421414613723755, "epoch": 6.55, "learning_rate": 3.1283656856606134e-05, "loss": 0.9333, "step": 7753, "task_loss": 0.8426740169525146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.909177839756012, "epoch": 6.55, "learning_rate": 3.128052598622417e-05, "loss": 0.8068, "step": 7754, "task_loss": 0.9784224629402161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6276123523712158, "epoch": 6.56, "learning_rate": 3.1277395115842204e-05, "loss": 0.693, "step": 7755, "task_loss": 0.2930087149143219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6698158383369446, "epoch": 6.56, "learning_rate": 3.1274264245460236e-05, "loss": 0.805, "step": 7756, "task_loss": 1.1276750564575195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5771065950393677, "epoch": 6.56, "learning_rate": 3.1271133375078274e-05, "loss": 0.8536, "step": 7757, "task_loss": 0.8202257752418518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8355543613433838, "epoch": 6.56, "learning_rate": 3.1268002504696306e-05, "loss": 0.7671, "step": 7758, "task_loss": 0.7898392677307129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4973604679107666, "epoch": 6.56, "learning_rate": 3.126487163431434e-05, "loss": 0.5626, "step": 7759, "task_loss": 0.10328813642263412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6982392072677612, "epoch": 6.56, "learning_rate": 3.1261740763932376e-05, "loss": 0.8467, "step": 7760, "task_loss": 0.7997664213180542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6570202708244324, "epoch": 6.56, "learning_rate": 3.125860989355041e-05, "loss": 0.8553, "step": 7761, "task_loss": 1.2264115810394287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6291763782501221, "epoch": 6.56, "learning_rate": 3.125547902316845e-05, "loss": 0.6786, "step": 7762, "task_loss": 0.45492851734161377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7704598903656006, "epoch": 6.56, "learning_rate": 3.125234815278648e-05, "loss": 0.7773, "step": 7763, "task_loss": 1.3052699565887451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5199309587478638, "epoch": 6.56, "learning_rate": 3.124921728240451e-05, "loss": 0.5639, "step": 7764, "task_loss": 0.6199285387992859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.649483859539032, "epoch": 6.56, "learning_rate": 3.124608641202255e-05, "loss": 0.9837, "step": 7765, "task_loss": 1.2204707860946655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6507588028907776, "epoch": 6.56, "learning_rate": 3.124295554164058e-05, "loss": 0.8115, "step": 7766, "task_loss": 0.28513821959495544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37810662388801575, "epoch": 6.57, "learning_rate": 3.123982467125861e-05, "loss": 0.7091, "step": 7767, "task_loss": 0.2752509117126465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6285630464553833, "epoch": 6.57, "learning_rate": 3.1236693800876644e-05, "loss": 0.797, "step": 7768, "task_loss": 0.301032692193985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37722986936569214, "epoch": 6.57, "learning_rate": 3.123356293049468e-05, "loss": 0.5529, "step": 7769, "task_loss": 0.026602042838931084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7910642027854919, "epoch": 6.57, "learning_rate": 3.1230432060112715e-05, "loss": 0.8099, "step": 7770, "task_loss": 0.5064657926559448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0066243410110474, "epoch": 6.57, "learning_rate": 3.1227301189730746e-05, "loss": 0.9155, "step": 7771, "task_loss": 1.0802581310272217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4929963946342468, "epoch": 6.57, "learning_rate": 3.122417031934878e-05, "loss": 0.6306, "step": 7772, "task_loss": 0.8994061350822449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6810818910598755, "epoch": 6.57, "learning_rate": 3.122103944896682e-05, "loss": 1.0018, "step": 7773, "task_loss": 1.4307273626327515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4641556441783905, "epoch": 6.57, "learning_rate": 3.121790857858485e-05, "loss": 0.7404, "step": 7774, "task_loss": 0.6032471060752869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7676602602005005, "epoch": 6.57, "learning_rate": 3.121477770820288e-05, "loss": 0.693, "step": 7775, "task_loss": 1.1913480758666992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6716569662094116, "epoch": 6.57, "learning_rate": 3.121164683782092e-05, "loss": 0.7558, "step": 7776, "task_loss": 0.5662922263145447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6996621489524841, "epoch": 6.57, "learning_rate": 3.120851596743895e-05, "loss": 0.6461, "step": 7777, "task_loss": 0.8868237137794495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6807137727737427, "epoch": 6.57, "learning_rate": 3.120538509705698e-05, "loss": 0.6815, "step": 7778, "task_loss": 2.1811330318450928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5409700870513916, "epoch": 6.58, "learning_rate": 3.1202254226675014e-05, "loss": 0.797, "step": 7779, "task_loss": 0.6231014728546143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9853704571723938, "epoch": 6.58, "learning_rate": 3.119912335629305e-05, "loss": 0.7112, "step": 7780, "task_loss": 0.6288038492202759 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.817145824432373, "epoch": 6.58, "learning_rate": 3.1195992485911084e-05, "loss": 0.8283, "step": 7781, "task_loss": 0.42451825737953186 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7505125999450684, "epoch": 6.58, "learning_rate": 3.1192861615529116e-05, "loss": 0.882, "step": 7782, "task_loss": 0.32465946674346924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0316979885101318, "epoch": 6.58, "learning_rate": 3.118973074514715e-05, "loss": 0.6993, "step": 7783, "task_loss": 1.2979589700698853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8815667629241943, "epoch": 6.58, "learning_rate": 3.1186599874765187e-05, "loss": 1.0177, "step": 7784, "task_loss": 1.1194361448287964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47384554147720337, "epoch": 6.58, "learning_rate": 3.118346900438322e-05, "loss": 0.6327, "step": 7785, "task_loss": 0.5807791948318481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8282086253166199, "epoch": 6.58, "learning_rate": 3.118033813400125e-05, "loss": 0.8522, "step": 7786, "task_loss": 0.4251169264316559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.5093570947647095, "epoch": 6.58, "learning_rate": 3.117720726361929e-05, "loss": 0.9348, "step": 7787, "task_loss": 2.3874728679656982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7717408537864685, "epoch": 6.58, "learning_rate": 3.117407639323732e-05, "loss": 0.7361, "step": 7788, "task_loss": 1.1861200332641602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4925551414489746, "epoch": 6.58, "learning_rate": 3.117094552285535e-05, "loss": 0.542, "step": 7789, "task_loss": 0.5556874871253967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9843547344207764, "epoch": 6.58, "learning_rate": 3.1167814652473384e-05, "loss": 0.8677, "step": 7790, "task_loss": 0.6794679760932922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5809696912765503, "epoch": 6.59, "learning_rate": 3.116468378209142e-05, "loss": 0.6588, "step": 7791, "task_loss": 0.9491599202156067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2462503910064697, "epoch": 6.59, "learning_rate": 3.1161552911709454e-05, "loss": 0.9286, "step": 7792, "task_loss": 1.3562273979187012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0422987937927246, "epoch": 6.59, "learning_rate": 3.115842204132749e-05, "loss": 0.7465, "step": 7793, "task_loss": 0.6310070753097534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.012008786201477, "epoch": 6.59, "learning_rate": 3.1155291170945525e-05, "loss": 0.7571, "step": 7794, "task_loss": 0.5945205092430115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5800241231918335, "epoch": 6.59, "learning_rate": 3.1152160300563556e-05, "loss": 0.847, "step": 7795, "task_loss": 0.6979014873504639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6636587381362915, "epoch": 6.59, "learning_rate": 3.1149029430181595e-05, "loss": 0.6557, "step": 7796, "task_loss": 0.8113368153572083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5897580981254578, "epoch": 6.59, "learning_rate": 3.114589855979963e-05, "loss": 0.6969, "step": 7797, "task_loss": 0.4159582257270813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5598008632659912, "epoch": 6.59, "learning_rate": 3.1142767689417665e-05, "loss": 0.7063, "step": 7798, "task_loss": 0.5725383162498474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.445076048374176, "epoch": 6.59, "learning_rate": 3.11396368190357e-05, "loss": 0.7453, "step": 7799, "task_loss": 0.5173043012619019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8305970430374146, "epoch": 6.59, "learning_rate": 3.113650594865373e-05, "loss": 0.7592, "step": 7800, "task_loss": 0.2392357587814331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7252662181854248, "epoch": 6.59, "learning_rate": 3.113337507827176e-05, "loss": 0.6894, "step": 7801, "task_loss": 0.8649057745933533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5937643051147461, "epoch": 6.59, "learning_rate": 3.11302442078898e-05, "loss": 0.7478, "step": 7802, "task_loss": 1.286636471748352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7974721789360046, "epoch": 6.6, "learning_rate": 3.112711333750783e-05, "loss": 0.8662, "step": 7803, "task_loss": 1.4945555925369263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7067269086837769, "epoch": 6.6, "learning_rate": 3.112398246712586e-05, "loss": 0.7206, "step": 7804, "task_loss": 0.6560543179512024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9145022630691528, "epoch": 6.6, "learning_rate": 3.1120851596743895e-05, "loss": 0.8666, "step": 7805, "task_loss": 1.6974306106567383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1147735118865967, "epoch": 6.6, "learning_rate": 3.111772072636193e-05, "loss": 0.9254, "step": 7806, "task_loss": 1.7146272659301758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.055774211883545, "epoch": 6.6, "learning_rate": 3.1114589855979965e-05, "loss": 0.7625, "step": 7807, "task_loss": 1.614280343055725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8186135292053223, "epoch": 6.6, "learning_rate": 3.1111458985598e-05, "loss": 0.6446, "step": 7808, "task_loss": 0.6049314141273499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6119240522384644, "epoch": 6.6, "learning_rate": 3.110832811521603e-05, "loss": 0.5591, "step": 7809, "task_loss": 1.458935022354126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3257405757904053, "epoch": 6.6, "learning_rate": 3.110519724483407e-05, "loss": 0.8412, "step": 7810, "task_loss": 0.7480639219284058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.009911060333252, "epoch": 6.6, "learning_rate": 3.11020663744521e-05, "loss": 0.8068, "step": 7811, "task_loss": 0.907606840133667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3180341422557831, "epoch": 6.6, "learning_rate": 3.109893550407013e-05, "loss": 0.6078, "step": 7812, "task_loss": 0.45485252141952515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5428030490875244, "epoch": 6.6, "learning_rate": 3.109580463368817e-05, "loss": 0.6704, "step": 7813, "task_loss": 0.22354021668434143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4580204486846924, "epoch": 6.6, "learning_rate": 3.10926737633062e-05, "loss": 0.6615, "step": 7814, "task_loss": 0.0581025555729866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6748038530349731, "epoch": 6.61, "learning_rate": 3.108954289292423e-05, "loss": 0.6704, "step": 7815, "task_loss": 1.4242591857910156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4912005066871643, "epoch": 6.61, "learning_rate": 3.1086412022542265e-05, "loss": 0.6126, "step": 7816, "task_loss": 0.15230980515480042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.741539478302002, "epoch": 6.61, "learning_rate": 3.10832811521603e-05, "loss": 0.7237, "step": 7817, "task_loss": 2.1263716220855713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6776713132858276, "epoch": 6.61, "learning_rate": 3.1080150281778335e-05, "loss": 0.7817, "step": 7818, "task_loss": 0.9083625674247742 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7384053468704224, "epoch": 6.61, "learning_rate": 3.107701941139637e-05, "loss": 0.7134, "step": 7819, "task_loss": 0.4966283440589905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7532395124435425, "epoch": 6.61, "learning_rate": 3.10738885410144e-05, "loss": 0.6733, "step": 7820, "task_loss": 1.1035606861114502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9999198317527771, "epoch": 6.61, "learning_rate": 3.107075767063244e-05, "loss": 0.8278, "step": 7821, "task_loss": 0.81703120470047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8103848695755005, "epoch": 6.61, "learning_rate": 3.106762680025047e-05, "loss": 0.8363, "step": 7822, "task_loss": 0.6550906300544739 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6558685302734375, "epoch": 6.61, "learning_rate": 3.10644959298685e-05, "loss": 0.7581, "step": 7823, "task_loss": 0.6165711283683777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6580954790115356, "epoch": 6.61, "learning_rate": 3.106136505948654e-05, "loss": 0.6653, "step": 7824, "task_loss": 0.8322805166244507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38331976532936096, "epoch": 6.61, "learning_rate": 3.105823418910457e-05, "loss": 0.48, "step": 7825, "task_loss": 0.18065746128559113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7441970109939575, "epoch": 6.61, "learning_rate": 3.10551033187226e-05, "loss": 0.8265, "step": 7826, "task_loss": 1.5507702827453613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1339428424835205, "epoch": 6.62, "learning_rate": 3.105197244834064e-05, "loss": 0.9891, "step": 7827, "task_loss": 0.9072702527046204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7895769476890564, "epoch": 6.62, "learning_rate": 3.104884157795867e-05, "loss": 0.8018, "step": 7828, "task_loss": 1.771973729133606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8070446848869324, "epoch": 6.62, "learning_rate": 3.104571070757671e-05, "loss": 0.8209, "step": 7829, "task_loss": 0.6095823049545288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5976200103759766, "epoch": 6.62, "learning_rate": 3.104257983719474e-05, "loss": 0.8164, "step": 7830, "task_loss": 0.4563678503036499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8954372406005859, "epoch": 6.62, "learning_rate": 3.1039448966812775e-05, "loss": 0.6156, "step": 7831, "task_loss": 1.2633955478668213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8336849212646484, "epoch": 6.62, "learning_rate": 3.1036318096430814e-05, "loss": 0.727, "step": 7832, "task_loss": 0.34747079014778137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3035629987716675, "epoch": 6.62, "learning_rate": 3.1033187226048845e-05, "loss": 0.7095, "step": 7833, "task_loss": 0.8001317977905273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6860249042510986, "epoch": 6.62, "learning_rate": 3.103005635566688e-05, "loss": 0.9005, "step": 7834, "task_loss": 0.37612733244895935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8491510152816772, "epoch": 6.62, "learning_rate": 3.1026925485284916e-05, "loss": 0.801, "step": 7835, "task_loss": 1.0654646158218384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.573403537273407, "epoch": 6.62, "learning_rate": 3.102379461490295e-05, "loss": 0.8897, "step": 7836, "task_loss": 0.7413607239723206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1847662925720215, "epoch": 6.62, "learning_rate": 3.102066374452098e-05, "loss": 0.9027, "step": 7837, "task_loss": 0.7340397834777832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.494746208190918, "epoch": 6.63, "learning_rate": 3.101753287413901e-05, "loss": 0.95, "step": 7838, "task_loss": 0.8388433456420898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7944983243942261, "epoch": 6.63, "learning_rate": 3.101440200375705e-05, "loss": 0.8105, "step": 7839, "task_loss": 1.4507156610488892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7479832768440247, "epoch": 6.63, "learning_rate": 3.101127113337508e-05, "loss": 0.6821, "step": 7840, "task_loss": 0.9795768857002258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30935439467430115, "epoch": 6.63, "learning_rate": 3.100814026299311e-05, "loss": 0.6102, "step": 7841, "task_loss": 0.03303993120789528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6686610579490662, "epoch": 6.63, "learning_rate": 3.1005009392611145e-05, "loss": 0.7911, "step": 7842, "task_loss": 0.6055296659469604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7549334764480591, "epoch": 6.63, "learning_rate": 3.1001878522229184e-05, "loss": 0.7868, "step": 7843, "task_loss": 1.056333065032959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0769237279891968, "epoch": 6.63, "learning_rate": 3.0998747651847215e-05, "loss": 0.7515, "step": 7844, "task_loss": 1.344021201133728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1956653594970703, "epoch": 6.63, "learning_rate": 3.099561678146525e-05, "loss": 0.9729, "step": 7845, "task_loss": 1.0827137231826782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1090543270111084, "epoch": 6.63, "learning_rate": 3.099248591108328e-05, "loss": 1.0009, "step": 7846, "task_loss": 2.294351816177368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.164046287536621, "epoch": 6.63, "learning_rate": 3.098935504070132e-05, "loss": 0.77, "step": 7847, "task_loss": 0.5658764839172363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6703422665596008, "epoch": 6.63, "learning_rate": 3.098622417031935e-05, "loss": 0.7102, "step": 7848, "task_loss": 0.8571425080299377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7830971479415894, "epoch": 6.63, "learning_rate": 3.098309329993738e-05, "loss": 0.7118, "step": 7849, "task_loss": 0.9499332308769226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5924041867256165, "epoch": 6.64, "learning_rate": 3.097996242955542e-05, "loss": 0.5809, "step": 7850, "task_loss": 1.1958281993865967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.7045538425445557, "epoch": 6.64, "learning_rate": 3.097683155917345e-05, "loss": 0.8583, "step": 7851, "task_loss": 1.0171397924423218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0536372661590576, "epoch": 6.64, "learning_rate": 3.097370068879148e-05, "loss": 0.8929, "step": 7852, "task_loss": 1.4715747833251953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4510791301727295, "epoch": 6.64, "learning_rate": 3.0970569818409515e-05, "loss": 0.9185, "step": 7853, "task_loss": 1.2247133255004883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6445608735084534, "epoch": 6.64, "learning_rate": 3.0967438948027554e-05, "loss": 0.7364, "step": 7854, "task_loss": 0.15796324610710144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5720125436782837, "epoch": 6.64, "learning_rate": 3.0964308077645585e-05, "loss": 0.6302, "step": 7855, "task_loss": 0.6526627540588379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8593996167182922, "epoch": 6.64, "learning_rate": 3.096117720726362e-05, "loss": 0.8428, "step": 7856, "task_loss": 1.5648798942565918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6933846473693848, "epoch": 6.64, "learning_rate": 3.095804633688165e-05, "loss": 0.6324, "step": 7857, "task_loss": 0.19046542048454285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8903942108154297, "epoch": 6.64, "learning_rate": 3.095491546649969e-05, "loss": 0.9121, "step": 7858, "task_loss": 0.36920106410980225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3940043449401855, "epoch": 6.64, "learning_rate": 3.095178459611772e-05, "loss": 0.9585, "step": 7859, "task_loss": 1.3199647665023804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.665700376033783, "epoch": 6.64, "learning_rate": 3.094865372573576e-05, "loss": 0.7214, "step": 7860, "task_loss": 1.0405535697937012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39294159412384033, "epoch": 6.64, "learning_rate": 3.094552285535379e-05, "loss": 0.5998, "step": 7861, "task_loss": 0.0675031766295433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.01967191696167, "epoch": 6.65, "learning_rate": 3.094239198497182e-05, "loss": 0.9219, "step": 7862, "task_loss": 0.9841682314872742 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5606961250305176, "epoch": 6.65, "learning_rate": 3.093926111458986e-05, "loss": 0.7567, "step": 7863, "task_loss": 0.7634813189506531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0704057216644287, "epoch": 6.65, "learning_rate": 3.093613024420789e-05, "loss": 0.7604, "step": 7864, "task_loss": 1.0676367282867432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0188149213790894, "epoch": 6.65, "learning_rate": 3.093299937382593e-05, "loss": 0.8433, "step": 7865, "task_loss": 1.3246017694473267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5383800268173218, "epoch": 6.65, "learning_rate": 3.092986850344396e-05, "loss": 0.7557, "step": 7866, "task_loss": 0.18932032585144043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6851837635040283, "epoch": 6.65, "learning_rate": 3.0926737633061994e-05, "loss": 0.6274, "step": 7867, "task_loss": 1.1945297718048096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8965051174163818, "epoch": 6.65, "learning_rate": 3.0923606762680026e-05, "loss": 0.7455, "step": 7868, "task_loss": 0.8608350157737732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5254784822463989, "epoch": 6.65, "learning_rate": 3.0920475892298064e-05, "loss": 0.7449, "step": 7869, "task_loss": 0.7571254968643188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42447981238365173, "epoch": 6.65, "learning_rate": 3.0917345021916096e-05, "loss": 0.7027, "step": 7870, "task_loss": 1.397099494934082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0335148572921753, "epoch": 6.65, "learning_rate": 3.091421415153413e-05, "loss": 0.8585, "step": 7871, "task_loss": 1.294325351715088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41244882345199585, "epoch": 6.65, "learning_rate": 3.0911083281152166e-05, "loss": 0.7825, "step": 7872, "task_loss": 1.032364845275879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9099690914154053, "epoch": 6.65, "learning_rate": 3.09079524107702e-05, "loss": 0.7469, "step": 7873, "task_loss": 1.3128854036331177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9677103757858276, "epoch": 6.66, "learning_rate": 3.090482154038823e-05, "loss": 0.8604, "step": 7874, "task_loss": 0.5719103217124939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9494198560714722, "epoch": 6.66, "learning_rate": 3.090169067000626e-05, "loss": 0.8327, "step": 7875, "task_loss": 0.7604184746742249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4780595898628235, "epoch": 6.66, "learning_rate": 3.08985597996243e-05, "loss": 0.5692, "step": 7876, "task_loss": 0.19279123842716217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4833381772041321, "epoch": 6.66, "learning_rate": 3.089542892924233e-05, "loss": 0.5928, "step": 7877, "task_loss": 0.7064904570579529 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9333354234695435, "epoch": 6.66, "learning_rate": 3.0892298058860364e-05, "loss": 0.779, "step": 7878, "task_loss": 1.0014985799789429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.642236590385437, "epoch": 6.66, "learning_rate": 3.0889167188478396e-05, "loss": 0.9117, "step": 7879, "task_loss": 1.2465085983276367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7298465967178345, "epoch": 6.66, "learning_rate": 3.0886036318096434e-05, "loss": 0.6163, "step": 7880, "task_loss": 0.23890121281147003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5414343476295471, "epoch": 6.66, "learning_rate": 3.0882905447714466e-05, "loss": 0.7035, "step": 7881, "task_loss": 0.7312673926353455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7427877187728882, "epoch": 6.66, "learning_rate": 3.08797745773325e-05, "loss": 0.6797, "step": 7882, "task_loss": 0.5666733980178833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5464044213294983, "epoch": 6.66, "learning_rate": 3.087664370695053e-05, "loss": 0.6418, "step": 7883, "task_loss": 0.23757542669773102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9232921600341797, "epoch": 6.66, "learning_rate": 3.087351283656857e-05, "loss": 0.8603, "step": 7884, "task_loss": 0.6532484889030457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5538132190704346, "epoch": 6.66, "learning_rate": 3.08703819661866e-05, "loss": 0.9106, "step": 7885, "task_loss": 1.3001362085342407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8099179863929749, "epoch": 6.67, "learning_rate": 3.086725109580463e-05, "loss": 0.7026, "step": 7886, "task_loss": 1.236095666885376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6212517023086548, "epoch": 6.67, "learning_rate": 3.086412022542267e-05, "loss": 0.7879, "step": 7887, "task_loss": 1.3866441249847412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5920131802558899, "epoch": 6.67, "learning_rate": 3.08609893550407e-05, "loss": 0.7572, "step": 7888, "task_loss": 0.9410519003868103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9802490472793579, "epoch": 6.67, "learning_rate": 3.0857858484658734e-05, "loss": 0.9189, "step": 7889, "task_loss": 1.3675380945205688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8461836576461792, "epoch": 6.67, "learning_rate": 3.0854727614276765e-05, "loss": 0.5765, "step": 7890, "task_loss": 1.3642176389694214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5338706970214844, "epoch": 6.67, "learning_rate": 3.0851596743894804e-05, "loss": 0.5885, "step": 7891, "task_loss": 0.3238854706287384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1104888916015625, "epoch": 6.67, "learning_rate": 3.0848465873512836e-05, "loss": 0.7802, "step": 7892, "task_loss": 0.781108021736145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3980805575847626, "epoch": 6.67, "learning_rate": 3.084533500313087e-05, "loss": 0.7244, "step": 7893, "task_loss": 1.0138074159622192 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5404284000396729, "epoch": 6.67, "learning_rate": 3.0842204132748906e-05, "loss": 0.7778, "step": 7894, "task_loss": 1.1715208292007446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7018006443977356, "epoch": 6.67, "learning_rate": 3.083907326236694e-05, "loss": 0.6435, "step": 7895, "task_loss": 0.9346175193786621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37566906213760376, "epoch": 6.67, "learning_rate": 3.0835942391984976e-05, "loss": 0.678, "step": 7896, "task_loss": 0.5053325295448303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0852903127670288, "epoch": 6.67, "learning_rate": 3.083281152160301e-05, "loss": 0.9121, "step": 7897, "task_loss": 1.0978707075119019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9346021413803101, "epoch": 6.68, "learning_rate": 3.082968065122105e-05, "loss": 0.7328, "step": 7898, "task_loss": 0.9871256947517395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5615735054016113, "epoch": 6.68, "learning_rate": 3.082654978083908e-05, "loss": 0.7322, "step": 7899, "task_loss": 0.788532018661499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6284962296485901, "epoch": 6.68, "learning_rate": 3.082341891045711e-05, "loss": 0.6955, "step": 7900, "task_loss": 0.5042320489883423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7040247321128845, "epoch": 6.68, "learning_rate": 3.082028804007514e-05, "loss": 0.598, "step": 7901, "task_loss": 0.6746094226837158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9123935699462891, "epoch": 6.68, "learning_rate": 3.081715716969318e-05, "loss": 0.7724, "step": 7902, "task_loss": 0.8819447755813599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.367483913898468, "epoch": 6.68, "learning_rate": 3.081402629931121e-05, "loss": 0.6404, "step": 7903, "task_loss": 0.7807163000106812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9524089694023132, "epoch": 6.68, "learning_rate": 3.0810895428929244e-05, "loss": 1.0343, "step": 7904, "task_loss": 2.036872148513794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6188325881958008, "epoch": 6.68, "learning_rate": 3.0807764558547276e-05, "loss": 0.7849, "step": 7905, "task_loss": 1.083208441734314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6731878519058228, "epoch": 6.68, "learning_rate": 3.0804633688165315e-05, "loss": 0.8576, "step": 7906, "task_loss": 1.0257712602615356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6524299383163452, "epoch": 6.68, "learning_rate": 3.0801502817783346e-05, "loss": 0.6192, "step": 7907, "task_loss": 0.39575883746147156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9307807683944702, "epoch": 6.68, "learning_rate": 3.079837194740138e-05, "loss": 0.7677, "step": 7908, "task_loss": 0.38086172938346863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5724180936813354, "epoch": 6.69, "learning_rate": 3.079524107701942e-05, "loss": 0.6702, "step": 7909, "task_loss": 0.7285068035125732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6594997048377991, "epoch": 6.69, "learning_rate": 3.079211020663745e-05, "loss": 0.7587, "step": 7910, "task_loss": 1.1743316650390625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6002024412155151, "epoch": 6.69, "learning_rate": 3.078897933625548e-05, "loss": 0.6248, "step": 7911, "task_loss": 0.37358155846595764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9696728587150574, "epoch": 6.69, "learning_rate": 3.078584846587351e-05, "loss": 0.7782, "step": 7912, "task_loss": 0.9270454049110413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9528603553771973, "epoch": 6.69, "learning_rate": 3.078271759549155e-05, "loss": 0.7149, "step": 7913, "task_loss": 0.5310961604118347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7027051448822021, "epoch": 6.69, "learning_rate": 3.077958672510958e-05, "loss": 0.8873, "step": 7914, "task_loss": 1.121349573135376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4320659339427948, "epoch": 6.69, "learning_rate": 3.0776455854727614e-05, "loss": 0.4677, "step": 7915, "task_loss": 1.2191030979156494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5533809065818787, "epoch": 6.69, "learning_rate": 3.0773324984345646e-05, "loss": 0.6047, "step": 7916, "task_loss": 0.6571025848388672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8710528612136841, "epoch": 6.69, "learning_rate": 3.0770194113963685e-05, "loss": 0.8873, "step": 7917, "task_loss": 0.5058302283287048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5362743139266968, "epoch": 6.69, "learning_rate": 3.0767063243581716e-05, "loss": 0.8926, "step": 7918, "task_loss": 0.10563885420560837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8363993763923645, "epoch": 6.69, "learning_rate": 3.076393237319975e-05, "loss": 0.6404, "step": 7919, "task_loss": 0.6310018301010132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8162299990653992, "epoch": 6.69, "learning_rate": 3.076080150281778e-05, "loss": 0.8566, "step": 7920, "task_loss": 1.2060855627059937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.255800724029541, "epoch": 6.7, "learning_rate": 3.075767063243582e-05, "loss": 0.8815, "step": 7921, "task_loss": 0.9523227214813232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0945072174072266, "epoch": 6.7, "learning_rate": 3.075453976205385e-05, "loss": 0.7355, "step": 7922, "task_loss": 0.4342915713787079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7952897548675537, "epoch": 6.7, "learning_rate": 3.075140889167188e-05, "loss": 0.6896, "step": 7923, "task_loss": 0.9575875997543335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1268901824951172, "epoch": 6.7, "learning_rate": 3.074827802128992e-05, "loss": 0.8557, "step": 7924, "task_loss": 0.2713778018951416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4746420383453369, "epoch": 6.7, "learning_rate": 3.074514715090795e-05, "loss": 0.6247, "step": 7925, "task_loss": 0.6120726466178894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6747310161590576, "epoch": 6.7, "learning_rate": 3.0742016280525984e-05, "loss": 0.6992, "step": 7926, "task_loss": 0.6556189060211182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1576824188232422, "epoch": 6.7, "learning_rate": 3.073888541014402e-05, "loss": 0.5899, "step": 7927, "task_loss": 0.676166296005249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6947252750396729, "epoch": 6.7, "learning_rate": 3.0735754539762054e-05, "loss": 0.7123, "step": 7928, "task_loss": 0.8598650693893433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7326651811599731, "epoch": 6.7, "learning_rate": 3.0732623669380086e-05, "loss": 0.7179, "step": 7929, "task_loss": 0.48558229207992554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6701529026031494, "epoch": 6.7, "learning_rate": 3.0729492798998125e-05, "loss": 0.7749, "step": 7930, "task_loss": 0.9428117275238037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5405824184417725, "epoch": 6.7, "learning_rate": 3.0726361928616157e-05, "loss": 0.7439, "step": 7931, "task_loss": 0.4943445324897766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6988991498947144, "epoch": 6.7, "learning_rate": 3.0723231058234195e-05, "loss": 0.864, "step": 7932, "task_loss": 0.732299268245697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7395046353340149, "epoch": 6.71, "learning_rate": 3.072010018785223e-05, "loss": 0.7329, "step": 7933, "task_loss": 0.9162377715110779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3390333652496338, "epoch": 6.71, "learning_rate": 3.071696931747026e-05, "loss": 0.647, "step": 7934, "task_loss": 0.16739346086978912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5781220197677612, "epoch": 6.71, "learning_rate": 3.07138384470883e-05, "loss": 0.7209, "step": 7935, "task_loss": 1.6372970342636108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9666109085083008, "epoch": 6.71, "learning_rate": 3.071070757670633e-05, "loss": 0.7204, "step": 7936, "task_loss": 1.3068536520004272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.923848032951355, "epoch": 6.71, "learning_rate": 3.070757670632436e-05, "loss": 0.863, "step": 7937, "task_loss": 1.6657745838165283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8401018381118774, "epoch": 6.71, "learning_rate": 3.070444583594239e-05, "loss": 0.6106, "step": 7938, "task_loss": 0.26621219515800476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.084112524986267, "epoch": 6.71, "learning_rate": 3.070131496556043e-05, "loss": 0.7683, "step": 7939, "task_loss": 1.067142128944397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8905581831932068, "epoch": 6.71, "learning_rate": 3.069818409517846e-05, "loss": 0.8928, "step": 7940, "task_loss": 1.202677607536316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8540109395980835, "epoch": 6.71, "learning_rate": 3.0695053224796495e-05, "loss": 0.9002, "step": 7941, "task_loss": 1.053513765335083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.478132426738739, "epoch": 6.71, "learning_rate": 3.0691922354414526e-05, "loss": 0.5133, "step": 7942, "task_loss": 0.9483898878097534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8751832246780396, "epoch": 6.71, "learning_rate": 3.0688791484032565e-05, "loss": 0.5966, "step": 7943, "task_loss": 0.6501158475875854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.699175238609314, "epoch": 6.71, "learning_rate": 3.06856606136506e-05, "loss": 1.0151, "step": 7944, "task_loss": 0.3583155572414398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7464189529418945, "epoch": 6.72, "learning_rate": 3.068252974326863e-05, "loss": 0.7482, "step": 7945, "task_loss": 0.7967494130134583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4984007477760315, "epoch": 6.72, "learning_rate": 3.067939887288667e-05, "loss": 0.7476, "step": 7946, "task_loss": 0.4930814802646637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3452970087528229, "epoch": 6.72, "learning_rate": 3.06762680025047e-05, "loss": 0.5505, "step": 7947, "task_loss": 0.4824996888637543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9079967141151428, "epoch": 6.72, "learning_rate": 3.067313713212273e-05, "loss": 0.756, "step": 7948, "task_loss": 1.1444380283355713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3337257206439972, "epoch": 6.72, "learning_rate": 3.067000626174076e-05, "loss": 0.5181, "step": 7949, "task_loss": 0.3130708634853363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4440925419330597, "epoch": 6.72, "learning_rate": 3.06668753913588e-05, "loss": 0.6898, "step": 7950, "task_loss": 0.48636943101882935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9903525114059448, "epoch": 6.72, "learning_rate": 3.066374452097683e-05, "loss": 0.9494, "step": 7951, "task_loss": 1.4414085149765015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.744814395904541, "epoch": 6.72, "learning_rate": 3.0660613650594865e-05, "loss": 0.7408, "step": 7952, "task_loss": 1.0941481590270996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5669673681259155, "epoch": 6.72, "learning_rate": 3.0657482780212896e-05, "loss": 0.7235, "step": 7953, "task_loss": 0.6772136092185974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5715273022651672, "epoch": 6.72, "learning_rate": 3.0654351909830935e-05, "loss": 0.7911, "step": 7954, "task_loss": 0.5386320948600769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7066366672515869, "epoch": 6.72, "learning_rate": 3.065122103944897e-05, "loss": 0.6468, "step": 7955, "task_loss": 0.6629272699356079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6540637016296387, "epoch": 6.72, "learning_rate": 3.0648090169067e-05, "loss": 0.6546, "step": 7956, "task_loss": 0.7783797383308411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3100660443305969, "epoch": 6.73, "learning_rate": 3.064495929868504e-05, "loss": 0.7583, "step": 7957, "task_loss": 0.5171807408332825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7717970609664917, "epoch": 6.73, "learning_rate": 3.064182842830307e-05, "loss": 0.7179, "step": 7958, "task_loss": 0.9658201932907104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8680041432380676, "epoch": 6.73, "learning_rate": 3.06386975579211e-05, "loss": 0.6671, "step": 7959, "task_loss": 1.0551222562789917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7201470732688904, "epoch": 6.73, "learning_rate": 3.063556668753913e-05, "loss": 0.5952, "step": 7960, "task_loss": 0.685112714767456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0383715629577637, "epoch": 6.73, "learning_rate": 3.063243581715717e-05, "loss": 0.8811, "step": 7961, "task_loss": 0.9338220357894897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6101997494697571, "epoch": 6.73, "learning_rate": 3.06293049467752e-05, "loss": 0.6698, "step": 7962, "task_loss": 0.09364841878414154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9403054714202881, "epoch": 6.73, "learning_rate": 3.062617407639324e-05, "loss": 0.7554, "step": 7963, "task_loss": 0.32351258397102356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9220346808433533, "epoch": 6.73, "learning_rate": 3.062304320601127e-05, "loss": 0.7014, "step": 7964, "task_loss": 0.7576712369918823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7538548111915588, "epoch": 6.73, "learning_rate": 3.0619912335629305e-05, "loss": 0.6788, "step": 7965, "task_loss": 1.0767817497253418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3589867055416107, "epoch": 6.73, "learning_rate": 3.0616781465247343e-05, "loss": 0.7294, "step": 7966, "task_loss": 0.5897250771522522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8686939477920532, "epoch": 6.73, "learning_rate": 3.0613650594865375e-05, "loss": 0.6761, "step": 7967, "task_loss": 0.874840497970581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42193278670310974, "epoch": 6.73, "learning_rate": 3.061051972448341e-05, "loss": 0.7367, "step": 7968, "task_loss": 0.832334578037262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8463811278343201, "epoch": 6.74, "learning_rate": 3.0607388854101446e-05, "loss": 0.6777, "step": 7969, "task_loss": 1.0389213562011719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5044190883636475, "epoch": 6.74, "learning_rate": 3.060425798371948e-05, "loss": 0.6571, "step": 7970, "task_loss": 0.4185419976711273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3555697798728943, "epoch": 6.74, "learning_rate": 3.060112711333751e-05, "loss": 0.5358, "step": 7971, "task_loss": 0.48405492305755615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.002551555633545, "epoch": 6.74, "learning_rate": 3.059799624295555e-05, "loss": 0.8214, "step": 7972, "task_loss": 0.8620414137840271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9709553718566895, "epoch": 6.74, "learning_rate": 3.059486537257358e-05, "loss": 0.7311, "step": 7973, "task_loss": 1.414740800857544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8744901418685913, "epoch": 6.74, "learning_rate": 3.059173450219161e-05, "loss": 0.8868, "step": 7974, "task_loss": 0.630386233329773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5999623537063599, "epoch": 6.74, "learning_rate": 3.058860363180964e-05, "loss": 0.7032, "step": 7975, "task_loss": 0.21744699776172638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0543558597564697, "epoch": 6.74, "learning_rate": 3.058547276142768e-05, "loss": 0.809, "step": 7976, "task_loss": 1.2337418794631958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5550273656845093, "epoch": 6.74, "learning_rate": 3.058234189104571e-05, "loss": 0.6174, "step": 7977, "task_loss": 0.26725032925605774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8222082257270813, "epoch": 6.74, "learning_rate": 3.0579211020663745e-05, "loss": 0.8687, "step": 7978, "task_loss": 1.0747005939483643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1381057500839233, "epoch": 6.74, "learning_rate": 3.057608015028178e-05, "loss": 0.9934, "step": 7979, "task_loss": 0.6794317960739136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7514306306838989, "epoch": 6.75, "learning_rate": 3.0572949279899815e-05, "loss": 0.7798, "step": 7980, "task_loss": 0.9214450120925903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1835412979125977, "epoch": 6.75, "learning_rate": 3.056981840951785e-05, "loss": 0.9492, "step": 7981, "task_loss": 1.094326376914978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5166473984718323, "epoch": 6.75, "learning_rate": 3.056668753913588e-05, "loss": 0.7134, "step": 7982, "task_loss": 0.2569558620452881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6642045974731445, "epoch": 6.75, "learning_rate": 3.056355666875392e-05, "loss": 1.0176, "step": 7983, "task_loss": 0.8207444548606873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20439884066581726, "epoch": 6.75, "learning_rate": 3.056042579837195e-05, "loss": 0.8197, "step": 7984, "task_loss": 0.25879883766174316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6247966289520264, "epoch": 6.75, "learning_rate": 3.055729492798998e-05, "loss": 0.6297, "step": 7985, "task_loss": 0.5506195425987244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4697665572166443, "epoch": 6.75, "learning_rate": 3.055416405760801e-05, "loss": 0.7849, "step": 7986, "task_loss": 1.2391473054885864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8071537017822266, "epoch": 6.75, "learning_rate": 3.055103318722605e-05, "loss": 0.9064, "step": 7987, "task_loss": 1.0094711780548096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7167580127716064, "epoch": 6.75, "learning_rate": 3.054790231684408e-05, "loss": 0.6624, "step": 7988, "task_loss": 1.3834856748580933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9525105357170105, "epoch": 6.75, "learning_rate": 3.0544771446462115e-05, "loss": 0.8786, "step": 7989, "task_loss": 0.6287370920181274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9671080708503723, "epoch": 6.75, "learning_rate": 3.054164057608015e-05, "loss": 0.8511, "step": 7990, "task_loss": 1.27701735496521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7852371335029602, "epoch": 6.75, "learning_rate": 3.0538509705698185e-05, "loss": 0.6646, "step": 7991, "task_loss": 0.41965723037719727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6298502087593079, "epoch": 6.76, "learning_rate": 3.053537883531622e-05, "loss": 0.7731, "step": 7992, "task_loss": 0.717761754989624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.813783586025238, "epoch": 6.76, "learning_rate": 3.053224796493425e-05, "loss": 0.6854, "step": 7993, "task_loss": 0.7985125184059143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41958165168762207, "epoch": 6.76, "learning_rate": 3.052911709455229e-05, "loss": 0.562, "step": 7994, "task_loss": 0.30338987708091736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8144826889038086, "epoch": 6.76, "learning_rate": 3.052598622417032e-05, "loss": 0.7757, "step": 7995, "task_loss": 1.9013867378234863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6854699850082397, "epoch": 6.76, "learning_rate": 3.052285535378835e-05, "loss": 0.7874, "step": 7996, "task_loss": 0.7759087681770325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5973459482192993, "epoch": 6.76, "learning_rate": 3.051972448340639e-05, "loss": 0.7179, "step": 7997, "task_loss": 0.7048553824424744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8944823741912842, "epoch": 6.76, "learning_rate": 3.0516593613024425e-05, "loss": 0.5908, "step": 7998, "task_loss": 1.3196204900741577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0855300426483154, "epoch": 6.76, "learning_rate": 3.0513462742642457e-05, "loss": 0.8099, "step": 7999, "task_loss": 0.5828117728233337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6271239519119263, "epoch": 6.76, "learning_rate": 3.051033187226049e-05, "loss": 0.7429, "step": 8000, "task_loss": 0.6456625461578369 }, { "epoch": 6.76, "eval_accuracy": 0.8897821782178218, "eval_loss": 0.48297107219696045, "eval_runtime": 206.9371, "eval_samples_per_second": 122.018, "eval_steps_per_second": 0.957, "step": 8000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9237237572669983, "epoch": 6.76, "learning_rate": 3.050720100187852e-05, "loss": 0.8892, "step": 8001, "task_loss": 1.4226126670837402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9296812415122986, "epoch": 6.76, "learning_rate": 3.050407013149656e-05, "loss": 0.6759, "step": 8002, "task_loss": 0.8763551115989685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2386844158172607, "epoch": 6.76, "learning_rate": 3.050093926111459e-05, "loss": 0.7346, "step": 8003, "task_loss": 1.128830909729004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6211484670639038, "epoch": 6.77, "learning_rate": 3.0497808390732622e-05, "loss": 0.786, "step": 8004, "task_loss": 0.8451536297798157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0221596956253052, "epoch": 6.77, "learning_rate": 3.0494677520350657e-05, "loss": 1.286, "step": 8005, "task_loss": 1.702332615852356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.818361759185791, "epoch": 6.77, "learning_rate": 3.0491546649968693e-05, "loss": 0.7531, "step": 8006, "task_loss": 1.2960338592529297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48814719915390015, "epoch": 6.77, "learning_rate": 3.0488415779586728e-05, "loss": 0.6556, "step": 8007, "task_loss": 0.9726532101631165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8058367967605591, "epoch": 6.77, "learning_rate": 3.048528490920476e-05, "loss": 0.7201, "step": 8008, "task_loss": 0.4125058948993683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8616460561752319, "epoch": 6.77, "learning_rate": 3.0482154038822798e-05, "loss": 0.8986, "step": 8009, "task_loss": 0.9930617809295654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6578724384307861, "epoch": 6.77, "learning_rate": 3.047902316844083e-05, "loss": 0.651, "step": 8010, "task_loss": 0.48877382278442383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5587406754493713, "epoch": 6.77, "learning_rate": 3.0475892298058862e-05, "loss": 0.7004, "step": 8011, "task_loss": 0.656059980392456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6491477489471436, "epoch": 6.77, "learning_rate": 3.0472761427676893e-05, "loss": 0.7857, "step": 8012, "task_loss": 0.9329527616500854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7311059832572937, "epoch": 6.77, "learning_rate": 3.0469630557294932e-05, "loss": 0.763, "step": 8013, "task_loss": 0.6288560032844543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4082827866077423, "epoch": 6.77, "learning_rate": 3.0466499686912964e-05, "loss": 0.6077, "step": 8014, "task_loss": 1.280840277671814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8919941186904907, "epoch": 6.77, "learning_rate": 3.0463368816530996e-05, "loss": 0.828, "step": 8015, "task_loss": 1.0084186792373657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.026086688041687, "epoch": 6.78, "learning_rate": 3.0460237946149027e-05, "loss": 0.8432, "step": 8016, "task_loss": 0.9882441759109497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5751664638519287, "epoch": 6.78, "learning_rate": 3.0457107075767066e-05, "loss": 0.8981, "step": 8017, "task_loss": 0.9066872000694275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8816987872123718, "epoch": 6.78, "learning_rate": 3.0453976205385098e-05, "loss": 0.7058, "step": 8018, "task_loss": 2.026719808578491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6403158903121948, "epoch": 6.78, "learning_rate": 3.045084533500313e-05, "loss": 0.7685, "step": 8019, "task_loss": 0.5499211549758911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8560663461685181, "epoch": 6.78, "learning_rate": 3.0447714464621168e-05, "loss": 0.7037, "step": 8020, "task_loss": 0.7716274857521057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6223645210266113, "epoch": 6.78, "learning_rate": 3.04445835942392e-05, "loss": 0.9315, "step": 8021, "task_loss": 0.5726089477539062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2201869487762451, "epoch": 6.78, "learning_rate": 3.0441452723857235e-05, "loss": 0.9437, "step": 8022, "task_loss": 1.5313386917114258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3757723569869995, "epoch": 6.78, "learning_rate": 3.0438321853475267e-05, "loss": 0.7427, "step": 8023, "task_loss": 0.131608247756958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7249765396118164, "epoch": 6.78, "learning_rate": 3.0435190983093302e-05, "loss": 0.647, "step": 8024, "task_loss": 0.7709258794784546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38551822304725647, "epoch": 6.78, "learning_rate": 3.0432060112711337e-05, "loss": 0.6214, "step": 8025, "task_loss": 0.5775653123855591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7031338810920715, "epoch": 6.78, "learning_rate": 3.042892924232937e-05, "loss": 0.6141, "step": 8026, "task_loss": 0.9211567640304565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6687856912612915, "epoch": 6.78, "learning_rate": 3.04257983719474e-05, "loss": 0.6799, "step": 8027, "task_loss": 0.9018639326095581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5962198972702026, "epoch": 6.79, "learning_rate": 3.042266750156544e-05, "loss": 0.6452, "step": 8028, "task_loss": 0.7247738838195801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6173904538154602, "epoch": 6.79, "learning_rate": 3.041953663118347e-05, "loss": 0.6996, "step": 8029, "task_loss": 1.1695126295089722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46292927861213684, "epoch": 6.79, "learning_rate": 3.0416405760801503e-05, "loss": 0.6325, "step": 8030, "task_loss": 0.8175317049026489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0278488397598267, "epoch": 6.79, "learning_rate": 3.041327489041954e-05, "loss": 0.8013, "step": 8031, "task_loss": 1.6823773384094238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8410451412200928, "epoch": 6.79, "learning_rate": 3.0410144020037573e-05, "loss": 0.6783, "step": 8032, "task_loss": 0.6385701298713684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6807041764259338, "epoch": 6.79, "learning_rate": 3.0407013149655605e-05, "loss": 0.7153, "step": 8033, "task_loss": 0.7458494901657104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7340676784515381, "epoch": 6.79, "learning_rate": 3.0403882279273637e-05, "loss": 0.9188, "step": 8034, "task_loss": 0.5032667517662048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5733375549316406, "epoch": 6.79, "learning_rate": 3.0400751408891675e-05, "loss": 0.6578, "step": 8035, "task_loss": 0.7333895564079285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5219876766204834, "epoch": 6.79, "learning_rate": 3.0397620538509707e-05, "loss": 0.7543, "step": 8036, "task_loss": 0.49722516536712646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6091053485870361, "epoch": 6.79, "learning_rate": 3.039448966812774e-05, "loss": 0.7147, "step": 8037, "task_loss": 0.3772902488708496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8725963830947876, "epoch": 6.79, "learning_rate": 3.0391358797745774e-05, "loss": 0.7597, "step": 8038, "task_loss": 0.9452968835830688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5433263182640076, "epoch": 6.79, "learning_rate": 3.038822792736381e-05, "loss": 0.5589, "step": 8039, "task_loss": 0.898946225643158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40997517108917236, "epoch": 6.8, "learning_rate": 3.0385097056981844e-05, "loss": 0.5757, "step": 8040, "task_loss": 0.26627999544143677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6392544507980347, "epoch": 6.8, "learning_rate": 3.0381966186599876e-05, "loss": 0.9314, "step": 8041, "task_loss": 0.8716665506362915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5395082235336304, "epoch": 6.8, "learning_rate": 3.0378835316217908e-05, "loss": 0.9453, "step": 8042, "task_loss": 0.2828959822654724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5009576678276062, "epoch": 6.8, "learning_rate": 3.0375704445835946e-05, "loss": 0.8073, "step": 8043, "task_loss": 0.13211873173713684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22153258323669434, "epoch": 6.8, "learning_rate": 3.0372573575453978e-05, "loss": 0.7349, "step": 8044, "task_loss": 0.022708099335432053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7816877365112305, "epoch": 6.8, "learning_rate": 3.036944270507201e-05, "loss": 0.5989, "step": 8045, "task_loss": 0.69509357213974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.660279393196106, "epoch": 6.8, "learning_rate": 3.036631183469005e-05, "loss": 0.7, "step": 8046, "task_loss": 0.8400467038154602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48076891899108887, "epoch": 6.8, "learning_rate": 3.036318096430808e-05, "loss": 0.4893, "step": 8047, "task_loss": 0.33777496218681335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7523253560066223, "epoch": 6.8, "learning_rate": 3.0360050093926112e-05, "loss": 0.9199, "step": 8048, "task_loss": 0.6466416120529175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8455725908279419, "epoch": 6.8, "learning_rate": 3.0356919223544144e-05, "loss": 0.9154, "step": 8049, "task_loss": 0.9433245062828064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5085881948471069, "epoch": 6.8, "learning_rate": 3.0353788353162182e-05, "loss": 0.671, "step": 8050, "task_loss": 0.9551513195037842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6140590906143188, "epoch": 6.81, "learning_rate": 3.0350657482780214e-05, "loss": 0.4832, "step": 8051, "task_loss": 0.5808295607566833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5937479138374329, "epoch": 6.81, "learning_rate": 3.0347526612398246e-05, "loss": 0.7683, "step": 8052, "task_loss": 0.278334379196167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1208174228668213, "epoch": 6.81, "learning_rate": 3.0344395742016278e-05, "loss": 0.7439, "step": 8053, "task_loss": 0.35838836431503296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5164705514907837, "epoch": 6.81, "learning_rate": 3.0341264871634316e-05, "loss": 0.5578, "step": 8054, "task_loss": 0.41655007004737854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1441199779510498, "epoch": 6.81, "learning_rate": 3.0338134001252348e-05, "loss": 0.7915, "step": 8055, "task_loss": 1.2529029846191406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7532901167869568, "epoch": 6.81, "learning_rate": 3.0335003130870383e-05, "loss": 0.6824, "step": 8056, "task_loss": 0.6256493330001831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7235993146896362, "epoch": 6.81, "learning_rate": 3.033187226048842e-05, "loss": 0.6404, "step": 8057, "task_loss": 1.219488501548767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7330193519592285, "epoch": 6.81, "learning_rate": 3.0328741390106454e-05, "loss": 0.6425, "step": 8058, "task_loss": 0.9485900402069092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7305917739868164, "epoch": 6.81, "learning_rate": 3.0325610519724485e-05, "loss": 0.8978, "step": 8059, "task_loss": 0.9874294996261597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43914830684661865, "epoch": 6.81, "learning_rate": 3.0322479649342517e-05, "loss": 0.7434, "step": 8060, "task_loss": 0.5397192239761353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9254952073097229, "epoch": 6.81, "learning_rate": 3.0319348778960556e-05, "loss": 0.6407, "step": 8061, "task_loss": 0.4717887043952942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5657081604003906, "epoch": 6.81, "learning_rate": 3.0316217908578588e-05, "loss": 0.5823, "step": 8062, "task_loss": 0.4082193374633789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8663035035133362, "epoch": 6.82, "learning_rate": 3.031308703819662e-05, "loss": 0.7511, "step": 8063, "task_loss": 0.333367258310318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5338727831840515, "epoch": 6.82, "learning_rate": 3.030995616781465e-05, "loss": 0.6935, "step": 8064, "task_loss": 0.26723653078079224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.835747241973877, "epoch": 6.82, "learning_rate": 3.030682529743269e-05, "loss": 0.7865, "step": 8065, "task_loss": 0.30668482184410095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5296400785446167, "epoch": 6.82, "learning_rate": 3.030369442705072e-05, "loss": 0.5643, "step": 8066, "task_loss": 0.7632284164428711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0935466289520264, "epoch": 6.82, "learning_rate": 3.0300563556668753e-05, "loss": 0.8399, "step": 8067, "task_loss": 1.2540957927703857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6823038458824158, "epoch": 6.82, "learning_rate": 3.0297432686286792e-05, "loss": 0.7128, "step": 8068, "task_loss": 0.5011739134788513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8162461519241333, "epoch": 6.82, "learning_rate": 3.0294301815904824e-05, "loss": 0.7604, "step": 8069, "task_loss": 0.8056613206863403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8828133344650269, "epoch": 6.82, "learning_rate": 3.0291170945522855e-05, "loss": 0.7626, "step": 8070, "task_loss": 0.6984111666679382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7993190288543701, "epoch": 6.82, "learning_rate": 3.0288040075140887e-05, "loss": 0.8338, "step": 8071, "task_loss": 1.2695109844207764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6803023219108582, "epoch": 6.82, "learning_rate": 3.0284909204758926e-05, "loss": 0.5586, "step": 8072, "task_loss": 0.889339804649353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9348542094230652, "epoch": 6.82, "learning_rate": 3.0281778334376957e-05, "loss": 0.5859, "step": 8073, "task_loss": 1.1599030494689941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6622951030731201, "epoch": 6.82, "learning_rate": 3.0278647463994993e-05, "loss": 0.704, "step": 8074, "task_loss": 0.6952185034751892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4221195876598358, "epoch": 6.83, "learning_rate": 3.0275516593613024e-05, "loss": 0.7246, "step": 8075, "task_loss": 0.09200258553028107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6214673519134521, "epoch": 6.83, "learning_rate": 3.0272385723231063e-05, "loss": 0.5805, "step": 8076, "task_loss": 0.2958487868309021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6279910206794739, "epoch": 6.83, "learning_rate": 3.0269254852849095e-05, "loss": 0.7605, "step": 8077, "task_loss": 1.1846754550933838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5997483134269714, "epoch": 6.83, "learning_rate": 3.0266123982467127e-05, "loss": 0.5975, "step": 8078, "task_loss": 0.8923358917236328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41756176948547363, "epoch": 6.83, "learning_rate": 3.026299311208516e-05, "loss": 0.517, "step": 8079, "task_loss": 0.4644472002983093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8176606893539429, "epoch": 6.83, "learning_rate": 3.0259862241703197e-05, "loss": 0.8192, "step": 8080, "task_loss": 1.1819168329238892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.496995210647583, "epoch": 6.83, "learning_rate": 3.025673137132123e-05, "loss": 0.8047, "step": 8081, "task_loss": 1.1424791812896729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5884836912155151, "epoch": 6.83, "learning_rate": 3.025360050093926e-05, "loss": 0.721, "step": 8082, "task_loss": 0.16091178357601166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7213292121887207, "epoch": 6.83, "learning_rate": 3.02504696305573e-05, "loss": 0.7508, "step": 8083, "task_loss": 0.9519731402397156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7509251832962036, "epoch": 6.83, "learning_rate": 3.024733876017533e-05, "loss": 0.8102, "step": 8084, "task_loss": 0.32930588722229004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6221021413803101, "epoch": 6.83, "learning_rate": 3.0244207889793363e-05, "loss": 0.7495, "step": 8085, "task_loss": 0.47216105461120605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1614248752593994, "epoch": 6.83, "learning_rate": 3.0241077019411394e-05, "loss": 0.7163, "step": 8086, "task_loss": 0.3392457365989685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7467561364173889, "epoch": 6.84, "learning_rate": 3.0237946149029433e-05, "loss": 0.7562, "step": 8087, "task_loss": 0.8653516173362732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8870803117752075, "epoch": 6.84, "learning_rate": 3.0234815278647465e-05, "loss": 0.7209, "step": 8088, "task_loss": 1.201331377029419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1293883323669434, "epoch": 6.84, "learning_rate": 3.02316844082655e-05, "loss": 0.8251, "step": 8089, "task_loss": 0.6805856227874756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44928741455078125, "epoch": 6.84, "learning_rate": 3.022855353788353e-05, "loss": 0.5991, "step": 8090, "task_loss": 0.3053816854953766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9375143647193909, "epoch": 6.84, "learning_rate": 3.0225422667501567e-05, "loss": 0.678, "step": 8091, "task_loss": 0.5342874526977539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.095868706703186, "epoch": 6.84, "learning_rate": 3.0222291797119602e-05, "loss": 0.6937, "step": 8092, "task_loss": 1.068865180015564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1896491050720215, "epoch": 6.84, "learning_rate": 3.0219160926737634e-05, "loss": 0.8264, "step": 8093, "task_loss": 1.094529151916504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3949718475341797, "epoch": 6.84, "learning_rate": 3.0216030056355672e-05, "loss": 0.6326, "step": 8094, "task_loss": 0.26016366481781006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8403114080429077, "epoch": 6.84, "learning_rate": 3.0212899185973704e-05, "loss": 0.6749, "step": 8095, "task_loss": 0.9010891318321228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.218662977218628, "epoch": 6.84, "learning_rate": 3.0209768315591736e-05, "loss": 0.7953, "step": 8096, "task_loss": 1.3011195659637451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6501017212867737, "epoch": 6.84, "learning_rate": 3.0206637445209768e-05, "loss": 0.767, "step": 8097, "task_loss": 1.2222459316253662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37962090969085693, "epoch": 6.84, "learning_rate": 3.0203506574827806e-05, "loss": 0.8464, "step": 8098, "task_loss": 0.46347737312316895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4342881441116333, "epoch": 6.85, "learning_rate": 3.0200375704445838e-05, "loss": 0.575, "step": 8099, "task_loss": 0.5659269690513611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6970057487487793, "epoch": 6.85, "learning_rate": 3.019724483406387e-05, "loss": 0.7103, "step": 8100, "task_loss": 0.7415133714675903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.750408411026001, "epoch": 6.85, "learning_rate": 3.01941139636819e-05, "loss": 0.706, "step": 8101, "task_loss": 1.009749412536621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48656606674194336, "epoch": 6.85, "learning_rate": 3.019098309329994e-05, "loss": 0.7165, "step": 8102, "task_loss": 0.10548720508813858 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4877772331237793, "epoch": 6.85, "learning_rate": 3.0187852222917972e-05, "loss": 0.7569, "step": 8103, "task_loss": 0.46117982268333435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8384486436843872, "epoch": 6.85, "learning_rate": 3.0184721352536004e-05, "loss": 0.8825, "step": 8104, "task_loss": 1.1143828630447388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48359155654907227, "epoch": 6.85, "learning_rate": 3.0181590482154042e-05, "loss": 0.6549, "step": 8105, "task_loss": 1.1718418598175049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6476954221725464, "epoch": 6.85, "learning_rate": 3.0178459611772074e-05, "loss": 0.69, "step": 8106, "task_loss": 1.077156662940979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5791364312171936, "epoch": 6.85, "learning_rate": 3.017532874139011e-05, "loss": 0.5217, "step": 8107, "task_loss": 1.1798210144042969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49661535024642944, "epoch": 6.85, "learning_rate": 3.017219787100814e-05, "loss": 0.6446, "step": 8108, "task_loss": 0.47959649562835693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8737447261810303, "epoch": 6.85, "learning_rate": 3.0169067000626176e-05, "loss": 0.8257, "step": 8109, "task_loss": 0.605752170085907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6892181634902954, "epoch": 6.85, "learning_rate": 3.016593613024421e-05, "loss": 0.8022, "step": 8110, "task_loss": 1.0716580152511597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.533786416053772, "epoch": 6.86, "learning_rate": 3.0162805259862243e-05, "loss": 0.5668, "step": 8111, "task_loss": 0.5489543676376343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.358297199010849, "epoch": 6.86, "learning_rate": 3.0159674389480275e-05, "loss": 0.4699, "step": 8112, "task_loss": 0.21350440382957458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7502490878105164, "epoch": 6.86, "learning_rate": 3.0156543519098313e-05, "loss": 0.8523, "step": 8113, "task_loss": 1.2638576030731201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1174057722091675, "epoch": 6.86, "learning_rate": 3.0153412648716345e-05, "loss": 0.9291, "step": 8114, "task_loss": 0.6267266869544983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6000711917877197, "epoch": 6.86, "learning_rate": 3.0150281778334377e-05, "loss": 0.7008, "step": 8115, "task_loss": 0.4242357611656189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4282722473144531, "epoch": 6.86, "learning_rate": 3.014715090795241e-05, "loss": 0.5128, "step": 8116, "task_loss": 0.20809897780418396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5432948470115662, "epoch": 6.86, "learning_rate": 3.0144020037570447e-05, "loss": 0.5729, "step": 8117, "task_loss": 0.6642740368843079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6813332438468933, "epoch": 6.86, "learning_rate": 3.014088916718848e-05, "loss": 0.5339, "step": 8118, "task_loss": 0.6386902928352356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7495070695877075, "epoch": 6.86, "learning_rate": 3.013775829680651e-05, "loss": 0.691, "step": 8119, "task_loss": 1.0432173013687134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.867213249206543, "epoch": 6.86, "learning_rate": 3.013462742642455e-05, "loss": 0.8185, "step": 8120, "task_loss": 0.5247995853424072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6872918605804443, "epoch": 6.86, "learning_rate": 3.013149655604258e-05, "loss": 0.851, "step": 8121, "task_loss": 1.199689269065857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1007130146026611, "epoch": 6.87, "learning_rate": 3.0128365685660613e-05, "loss": 0.9843, "step": 8122, "task_loss": 1.346502661705017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1710999011993408, "epoch": 6.87, "learning_rate": 3.0125234815278648e-05, "loss": 0.7872, "step": 8123, "task_loss": 1.1223171949386597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5320358276367188, "epoch": 6.87, "learning_rate": 3.0122103944896683e-05, "loss": 0.8741, "step": 8124, "task_loss": 1.6718436479568481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8973795175552368, "epoch": 6.87, "learning_rate": 3.011897307451472e-05, "loss": 0.7065, "step": 8125, "task_loss": 1.1133089065551758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.511543869972229, "epoch": 6.87, "learning_rate": 3.011584220413275e-05, "loss": 0.7786, "step": 8126, "task_loss": 0.8603947162628174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3682222366333008, "epoch": 6.87, "learning_rate": 3.0112711333750782e-05, "loss": 0.9627, "step": 8127, "task_loss": 1.8120564222335815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8137875199317932, "epoch": 6.87, "learning_rate": 3.010958046336882e-05, "loss": 0.7197, "step": 8128, "task_loss": 0.32752174139022827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.500300645828247, "epoch": 6.87, "learning_rate": 3.0106449592986852e-05, "loss": 0.9798, "step": 8129, "task_loss": 0.8861093521118164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.052286982536316, "epoch": 6.87, "learning_rate": 3.0103318722604884e-05, "loss": 0.7519, "step": 8130, "task_loss": 0.22675620019435883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5926276445388794, "epoch": 6.87, "learning_rate": 3.0100187852222923e-05, "loss": 0.7451, "step": 8131, "task_loss": 0.9217538237571716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44961124658584595, "epoch": 6.87, "learning_rate": 3.0097056981840955e-05, "loss": 0.6769, "step": 8132, "task_loss": 1.1424063444137573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7701340913772583, "epoch": 6.87, "learning_rate": 3.0093926111458986e-05, "loss": 0.6738, "step": 8133, "task_loss": 0.5107067823410034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7953731417655945, "epoch": 6.88, "learning_rate": 3.0090795241077018e-05, "loss": 0.8647, "step": 8134, "task_loss": 1.1185576915740967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8900073766708374, "epoch": 6.88, "learning_rate": 3.0087664370695057e-05, "loss": 0.9396, "step": 8135, "task_loss": 1.2536497116088867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.290150910615921, "epoch": 6.88, "learning_rate": 3.008453350031309e-05, "loss": 0.4616, "step": 8136, "task_loss": 0.03775976225733757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7434871196746826, "epoch": 6.88, "learning_rate": 3.008140262993112e-05, "loss": 0.6697, "step": 8137, "task_loss": 1.5524439811706543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5607860088348389, "epoch": 6.88, "learning_rate": 3.0078271759549152e-05, "loss": 0.629, "step": 8138, "task_loss": 1.0289041996002197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.68821120262146, "epoch": 6.88, "learning_rate": 3.007514088916719e-05, "loss": 0.7939, "step": 8139, "task_loss": 0.7378551363945007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4222099184989929, "epoch": 6.88, "learning_rate": 3.0072010018785222e-05, "loss": 0.5754, "step": 8140, "task_loss": 0.11441431939601898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6319394707679749, "epoch": 6.88, "learning_rate": 3.0068879148403258e-05, "loss": 0.6199, "step": 8141, "task_loss": 0.33578261733055115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40427738428115845, "epoch": 6.88, "learning_rate": 3.0065748278021293e-05, "loss": 0.683, "step": 8142, "task_loss": 0.33525505661964417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1200283765792847, "epoch": 6.88, "learning_rate": 3.0062617407639328e-05, "loss": 0.7952, "step": 8143, "task_loss": 0.7649474740028381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.717497706413269, "epoch": 6.88, "learning_rate": 3.005948653725736e-05, "loss": 0.6227, "step": 8144, "task_loss": 0.4384651780128479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6021300554275513, "epoch": 6.88, "learning_rate": 3.005635566687539e-05, "loss": 0.5159, "step": 8145, "task_loss": 0.7214975953102112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3607124090194702, "epoch": 6.89, "learning_rate": 3.005322479649343e-05, "loss": 1.3385, "step": 8146, "task_loss": 0.8320221900939941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.760191023349762, "epoch": 6.89, "learning_rate": 3.0050093926111462e-05, "loss": 0.907, "step": 8147, "task_loss": 0.4087485373020172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5911930799484253, "epoch": 6.89, "learning_rate": 3.0046963055729494e-05, "loss": 0.7048, "step": 8148, "task_loss": 0.8148401379585266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2542433738708496, "epoch": 6.89, "learning_rate": 3.0043832185347525e-05, "loss": 0.5525, "step": 8149, "task_loss": 0.1834975630044937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7280491590499878, "epoch": 6.89, "learning_rate": 3.0040701314965564e-05, "loss": 0.7083, "step": 8150, "task_loss": 0.43500208854675293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1002658605575562, "epoch": 6.89, "learning_rate": 3.0037570444583596e-05, "loss": 0.7584, "step": 8151, "task_loss": 1.1981797218322754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5944511890411377, "epoch": 6.89, "learning_rate": 3.0034439574201627e-05, "loss": 0.6571, "step": 8152, "task_loss": 0.3961486518383026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.925670862197876, "epoch": 6.89, "learning_rate": 3.003130870381966e-05, "loss": 0.7483, "step": 8153, "task_loss": 1.40053129196167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8908116221427917, "epoch": 6.89, "learning_rate": 3.0028177833437698e-05, "loss": 0.6894, "step": 8154, "task_loss": 0.8426689505577087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7713844180107117, "epoch": 6.89, "learning_rate": 3.002504696305573e-05, "loss": 0.7339, "step": 8155, "task_loss": 0.5364183187484741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8727474808692932, "epoch": 6.89, "learning_rate": 3.002191609267376e-05, "loss": 0.7457, "step": 8156, "task_loss": 1.03933584690094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9055630564689636, "epoch": 6.89, "learning_rate": 3.00187852222918e-05, "loss": 0.8085, "step": 8157, "task_loss": 0.40826719999313354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7499237656593323, "epoch": 6.9, "learning_rate": 3.001565435190983e-05, "loss": 0.6406, "step": 8158, "task_loss": 0.6746458411216736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9516542553901672, "epoch": 6.9, "learning_rate": 3.0012523481527867e-05, "loss": 0.9132, "step": 8159, "task_loss": 1.0802339315414429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6690530180931091, "epoch": 6.9, "learning_rate": 3.00093926111459e-05, "loss": 0.7874, "step": 8160, "task_loss": 0.5628272294998169 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4926019310951233, "epoch": 6.9, "learning_rate": 3.0006261740763937e-05, "loss": 0.704, "step": 8161, "task_loss": 0.3775377571582794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5091313123703003, "epoch": 6.9, "learning_rate": 3.000313087038197e-05, "loss": 0.8095, "step": 8162, "task_loss": 0.43810921907424927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7398466467857361, "epoch": 6.9, "learning_rate": 3e-05, "loss": 0.935, "step": 8163, "task_loss": 0.5929480195045471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6470872759819031, "epoch": 6.9, "learning_rate": 2.9996869129618033e-05, "loss": 0.8357, "step": 8164, "task_loss": 0.8429592847824097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3146915137767792, "epoch": 6.9, "learning_rate": 2.999373825923607e-05, "loss": 0.8045, "step": 8165, "task_loss": 0.3597647547721863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5323036909103394, "epoch": 6.9, "learning_rate": 2.9990607388854103e-05, "loss": 0.8836, "step": 8166, "task_loss": 0.8081330060958862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5580652356147766, "epoch": 6.9, "learning_rate": 2.9987476518472135e-05, "loss": 0.675, "step": 8167, "task_loss": 0.5136520862579346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4934103488922119, "epoch": 6.9, "learning_rate": 2.9984345648090173e-05, "loss": 0.6918, "step": 8168, "task_loss": 0.6341977715492249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6579170227050781, "epoch": 6.9, "learning_rate": 2.9981214777708205e-05, "loss": 0.673, "step": 8169, "task_loss": 0.36020103096961975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7303563356399536, "epoch": 6.91, "learning_rate": 2.9978083907326237e-05, "loss": 0.7982, "step": 8170, "task_loss": 0.8560694456100464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7556244730949402, "epoch": 6.91, "learning_rate": 2.997495303694427e-05, "loss": 0.8048, "step": 8171, "task_loss": 1.2690606117248535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6643710136413574, "epoch": 6.91, "learning_rate": 2.9971822166562307e-05, "loss": 0.7317, "step": 8172, "task_loss": 1.2174793481826782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9286907911300659, "epoch": 6.91, "learning_rate": 2.996869129618034e-05, "loss": 0.759, "step": 8173, "task_loss": 1.6725319623947144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.751425862312317, "epoch": 6.91, "learning_rate": 2.9965560425798374e-05, "loss": 1.0685, "step": 8174, "task_loss": 1.7011208534240723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1028659343719482, "epoch": 6.91, "learning_rate": 2.9962429555416406e-05, "loss": 0.7433, "step": 8175, "task_loss": 0.8444010615348816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.641750693321228, "epoch": 6.91, "learning_rate": 2.995929868503444e-05, "loss": 0.5713, "step": 8176, "task_loss": 0.20041301846504211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5542474389076233, "epoch": 6.91, "learning_rate": 2.9956167814652476e-05, "loss": 0.6544, "step": 8177, "task_loss": 2.169489622116089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23805931210517883, "epoch": 6.91, "learning_rate": 2.9953036944270508e-05, "loss": 0.6769, "step": 8178, "task_loss": 0.17125765979290009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5415460467338562, "epoch": 6.91, "learning_rate": 2.9949906073888547e-05, "loss": 0.6884, "step": 8179, "task_loss": 0.856005847454071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43035173416137695, "epoch": 6.91, "learning_rate": 2.994677520350658e-05, "loss": 0.587, "step": 8180, "task_loss": 0.47426703572273254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.645392656326294, "epoch": 6.91, "learning_rate": 2.994364433312461e-05, "loss": 0.7043, "step": 8181, "task_loss": 0.9160101413726807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7390706539154053, "epoch": 6.92, "learning_rate": 2.9940513462742642e-05, "loss": 0.7232, "step": 8182, "task_loss": 1.2234392166137695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0875803232192993, "epoch": 6.92, "learning_rate": 2.993738259236068e-05, "loss": 0.9898, "step": 8183, "task_loss": 1.7948362827301025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.828230619430542, "epoch": 6.92, "learning_rate": 2.9934251721978712e-05, "loss": 0.7672, "step": 8184, "task_loss": 1.5015982389450073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6170579195022583, "epoch": 6.92, "learning_rate": 2.9931120851596744e-05, "loss": 0.5868, "step": 8185, "task_loss": 0.4825744330883026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6344759464263916, "epoch": 6.92, "learning_rate": 2.9927989981214776e-05, "loss": 0.6704, "step": 8186, "task_loss": 1.2865689992904663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5876890420913696, "epoch": 6.92, "learning_rate": 2.9924859110832814e-05, "loss": 0.6273, "step": 8187, "task_loss": 0.39338040351867676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8761193752288818, "epoch": 6.92, "learning_rate": 2.9921728240450846e-05, "loss": 0.5556, "step": 8188, "task_loss": 0.6660030484199524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9153162837028503, "epoch": 6.92, "learning_rate": 2.9918597370068878e-05, "loss": 0.7838, "step": 8189, "task_loss": 1.0175247192382812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48799681663513184, "epoch": 6.92, "learning_rate": 2.9915466499686916e-05, "loss": 0.6237, "step": 8190, "task_loss": 1.434092402458191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.690570592880249, "epoch": 6.92, "learning_rate": 2.9912335629304948e-05, "loss": 0.6039, "step": 8191, "task_loss": 0.5776435732841492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7775385975837708, "epoch": 6.92, "learning_rate": 2.9909204758922983e-05, "loss": 0.8117, "step": 8192, "task_loss": 0.6183419227600098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8063658475875854, "epoch": 6.93, "learning_rate": 2.9906073888541015e-05, "loss": 0.7125, "step": 8193, "task_loss": 0.9301208257675171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.043213129043579, "epoch": 6.93, "learning_rate": 2.990294301815905e-05, "loss": 0.7671, "step": 8194, "task_loss": 0.44110265374183655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9006463289260864, "epoch": 6.93, "learning_rate": 2.9899812147777086e-05, "loss": 0.6351, "step": 8195, "task_loss": 0.33605489134788513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0061187744140625, "epoch": 6.93, "learning_rate": 2.9896681277395117e-05, "loss": 0.851, "step": 8196, "task_loss": 0.979409396648407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2800246477127075, "epoch": 6.93, "learning_rate": 2.989355040701315e-05, "loss": 0.8479, "step": 8197, "task_loss": 0.9758844971656799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3742014765739441, "epoch": 6.93, "learning_rate": 2.9890419536631188e-05, "loss": 0.624, "step": 8198, "task_loss": 0.31285858154296875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5720573663711548, "epoch": 6.93, "learning_rate": 2.988728866624922e-05, "loss": 0.681, "step": 8199, "task_loss": 0.8196434378623962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7024637460708618, "epoch": 6.93, "learning_rate": 2.988415779586725e-05, "loss": 0.7405, "step": 8200, "task_loss": 0.6470819115638733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4066734313964844, "epoch": 6.93, "learning_rate": 2.9881026925485283e-05, "loss": 0.8415, "step": 8201, "task_loss": 0.56482994556427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6262194514274597, "epoch": 6.93, "learning_rate": 2.987789605510332e-05, "loss": 0.5962, "step": 8202, "task_loss": 0.932586669921875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.343085765838623, "epoch": 6.93, "learning_rate": 2.9874765184721353e-05, "loss": 0.784, "step": 8203, "task_loss": 1.5496526956558228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4097750186920166, "epoch": 6.93, "learning_rate": 2.9871634314339385e-05, "loss": 0.5638, "step": 8204, "task_loss": 0.2551794648170471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5221184492111206, "epoch": 6.94, "learning_rate": 2.9868503443957424e-05, "loss": 0.7039, "step": 8205, "task_loss": 0.13172177970409393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4865038990974426, "epoch": 6.94, "learning_rate": 2.9865372573575455e-05, "loss": 0.6925, "step": 8206, "task_loss": 1.3061773777008057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3441685438156128, "epoch": 6.94, "learning_rate": 2.9862241703193487e-05, "loss": 0.5736, "step": 8207, "task_loss": 0.3088807463645935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3105846047401428, "epoch": 6.94, "learning_rate": 2.9859110832811522e-05, "loss": 0.6129, "step": 8208, "task_loss": 0.380891889333725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8546575307846069, "epoch": 6.94, "learning_rate": 2.9855979962429558e-05, "loss": 0.9135, "step": 8209, "task_loss": 0.8510474562644958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6784946322441101, "epoch": 6.94, "learning_rate": 2.9852849092047593e-05, "loss": 0.8152, "step": 8210, "task_loss": 1.6679670810699463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8533382415771484, "epoch": 6.94, "learning_rate": 2.9849718221665625e-05, "loss": 0.7434, "step": 8211, "task_loss": 1.4586001634597778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5104595422744751, "epoch": 6.94, "learning_rate": 2.9846587351283656e-05, "loss": 0.7085, "step": 8212, "task_loss": 0.5304948091506958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37429988384246826, "epoch": 6.94, "learning_rate": 2.9843456480901695e-05, "loss": 0.6383, "step": 8213, "task_loss": 0.4654366374015808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.677433431148529, "epoch": 6.94, "learning_rate": 2.9840325610519727e-05, "loss": 0.6886, "step": 8214, "task_loss": 0.29723504185676575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0450596809387207, "epoch": 6.94, "learning_rate": 2.983719474013776e-05, "loss": 0.7517, "step": 8215, "task_loss": 0.929245114326477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5493571162223816, "epoch": 6.94, "learning_rate": 2.9834063869755797e-05, "loss": 0.6758, "step": 8216, "task_loss": 0.3659208118915558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.714011549949646, "epoch": 6.95, "learning_rate": 2.983093299937383e-05, "loss": 0.6651, "step": 8217, "task_loss": 0.9543878436088562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7503427863121033, "epoch": 6.95, "learning_rate": 2.982780212899186e-05, "loss": 0.6965, "step": 8218, "task_loss": 1.2316817045211792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.563043475151062, "epoch": 6.95, "learning_rate": 2.9824671258609892e-05, "loss": 0.6107, "step": 8219, "task_loss": 0.4723999798297882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6685945391654968, "epoch": 6.95, "learning_rate": 2.982154038822793e-05, "loss": 0.9951, "step": 8220, "task_loss": 0.45323818922042847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9790596961975098, "epoch": 6.95, "learning_rate": 2.9818409517845963e-05, "loss": 0.6466, "step": 8221, "task_loss": 1.1610949039459229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7536357641220093, "epoch": 6.95, "learning_rate": 2.9815278647463994e-05, "loss": 0.717, "step": 8222, "task_loss": 0.8096750378608704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7741275429725647, "epoch": 6.95, "learning_rate": 2.9812147777082026e-05, "loss": 0.7821, "step": 8223, "task_loss": 1.2198443412780762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5009340047836304, "epoch": 6.95, "learning_rate": 2.9809016906700065e-05, "loss": 0.6107, "step": 8224, "task_loss": 0.32648107409477234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.656183123588562, "epoch": 6.95, "learning_rate": 2.9805886036318097e-05, "loss": 0.8437, "step": 8225, "task_loss": 0.3111511766910553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7237969040870667, "epoch": 6.95, "learning_rate": 2.9802755165936132e-05, "loss": 0.8842, "step": 8226, "task_loss": 0.4510647654533386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0425610542297363, "epoch": 6.95, "learning_rate": 2.9799624295554167e-05, "loss": 0.8608, "step": 8227, "task_loss": 0.9519915580749512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7855206727981567, "epoch": 6.95, "learning_rate": 2.9796493425172202e-05, "loss": 0.6707, "step": 8228, "task_loss": 0.661739706993103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7616167664527893, "epoch": 6.96, "learning_rate": 2.9793362554790234e-05, "loss": 0.8077, "step": 8229, "task_loss": 0.6744396686553955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0198662281036377, "epoch": 6.96, "learning_rate": 2.9790231684408266e-05, "loss": 0.9171, "step": 8230, "task_loss": 0.6151405572891235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9492737650871277, "epoch": 6.96, "learning_rate": 2.9787100814026304e-05, "loss": 0.7138, "step": 8231, "task_loss": 1.049807071685791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7409478425979614, "epoch": 6.96, "learning_rate": 2.9783969943644336e-05, "loss": 0.726, "step": 8232, "task_loss": 1.4285073280334473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49430251121520996, "epoch": 6.96, "learning_rate": 2.9780839073262368e-05, "loss": 0.4874, "step": 8233, "task_loss": 0.24297265708446503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8631232380867004, "epoch": 6.96, "learning_rate": 2.97777082028804e-05, "loss": 0.6819, "step": 8234, "task_loss": 0.7193951606750488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9551374316215515, "epoch": 6.96, "learning_rate": 2.9774577332498438e-05, "loss": 0.8133, "step": 8235, "task_loss": 0.9354220628738403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6289761662483215, "epoch": 6.96, "learning_rate": 2.977144646211647e-05, "loss": 0.6106, "step": 8236, "task_loss": 0.41213613748550415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7536382675170898, "epoch": 6.96, "learning_rate": 2.97683155917345e-05, "loss": 0.6939, "step": 8237, "task_loss": 1.4889625310897827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.591346263885498, "epoch": 6.96, "learning_rate": 2.9765184721352533e-05, "loss": 0.7047, "step": 8238, "task_loss": 0.529362142086029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5472671985626221, "epoch": 6.96, "learning_rate": 2.9762053850970572e-05, "loss": 0.7262, "step": 8239, "task_loss": 0.36773988604545593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8629181385040283, "epoch": 6.96, "learning_rate": 2.9758922980588604e-05, "loss": 0.8885, "step": 8240, "task_loss": 0.34048375487327576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5611582398414612, "epoch": 6.97, "learning_rate": 2.975579211020664e-05, "loss": 0.7634, "step": 8241, "task_loss": 0.23405542969703674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5249444246292114, "epoch": 6.97, "learning_rate": 2.9752661239824674e-05, "loss": 0.5929, "step": 8242, "task_loss": 0.8978744745254517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4134767949581146, "epoch": 6.97, "learning_rate": 2.9749530369442706e-05, "loss": 0.8335, "step": 8243, "task_loss": 0.11002688854932785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4800195693969727, "epoch": 6.97, "learning_rate": 2.974639949906074e-05, "loss": 0.7747, "step": 8244, "task_loss": 1.490763545036316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.893944501876831, "epoch": 6.97, "learning_rate": 2.9743268628678773e-05, "loss": 0.7661, "step": 8245, "task_loss": 1.4952266216278076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8417910933494568, "epoch": 6.97, "learning_rate": 2.974013775829681e-05, "loss": 0.7907, "step": 8246, "task_loss": 1.0339423418045044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8535694479942322, "epoch": 6.97, "learning_rate": 2.9737006887914843e-05, "loss": 0.7274, "step": 8247, "task_loss": 0.5682619214057922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2043683528900146, "epoch": 6.97, "learning_rate": 2.9733876017532875e-05, "loss": 0.8688, "step": 8248, "task_loss": 0.928697407245636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.502396285533905, "epoch": 6.97, "learning_rate": 2.9730745147150907e-05, "loss": 0.6291, "step": 8249, "task_loss": 1.0411654710769653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0455117225646973, "epoch": 6.97, "learning_rate": 2.9727614276768945e-05, "loss": 0.9488, "step": 8250, "task_loss": 1.4214110374450684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5337565541267395, "epoch": 6.97, "learning_rate": 2.9724483406386977e-05, "loss": 0.5707, "step": 8251, "task_loss": 0.39270198345184326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.050272822380066, "epoch": 6.97, "learning_rate": 2.972135253600501e-05, "loss": 0.8119, "step": 8252, "task_loss": 1.0450643301010132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8007222414016724, "epoch": 6.98, "learning_rate": 2.9718221665623047e-05, "loss": 0.6158, "step": 8253, "task_loss": 0.19106417894363403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6590077877044678, "epoch": 6.98, "learning_rate": 2.971509079524108e-05, "loss": 0.6862, "step": 8254, "task_loss": 1.2598692178726196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8908135294914246, "epoch": 6.98, "learning_rate": 2.971195992485911e-05, "loss": 0.7687, "step": 8255, "task_loss": 1.0055832862854004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9464797973632812, "epoch": 6.98, "learning_rate": 2.9708829054477143e-05, "loss": 0.7208, "step": 8256, "task_loss": 1.1358431577682495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36312586069107056, "epoch": 6.98, "learning_rate": 2.970569818409518e-05, "loss": 0.7463, "step": 8257, "task_loss": 0.10128264129161835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0695358514785767, "epoch": 6.98, "learning_rate": 2.9702567313713213e-05, "loss": 0.9328, "step": 8258, "task_loss": 0.917073130607605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4656471610069275, "epoch": 6.98, "learning_rate": 2.9699436443331248e-05, "loss": 0.8154, "step": 8259, "task_loss": 2.1287946701049805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6003868579864502, "epoch": 6.98, "learning_rate": 2.969630557294928e-05, "loss": 1.2113, "step": 8260, "task_loss": 1.5413974523544312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7342080473899841, "epoch": 6.98, "learning_rate": 2.9693174702567315e-05, "loss": 0.7261, "step": 8261, "task_loss": 0.8838079571723938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8852670788764954, "epoch": 6.98, "learning_rate": 2.969004383218535e-05, "loss": 0.9114, "step": 8262, "task_loss": 1.424656629562378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33947497606277466, "epoch": 6.98, "learning_rate": 2.9686912961803382e-05, "loss": 0.6364, "step": 8263, "task_loss": 0.5398897528648376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.866936445236206, "epoch": 6.99, "learning_rate": 2.968378209142142e-05, "loss": 0.8708, "step": 8264, "task_loss": 0.8721164464950562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6497496366500854, "epoch": 6.99, "learning_rate": 2.9680651221039453e-05, "loss": 0.8662, "step": 8265, "task_loss": 0.6810654401779175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6576759219169617, "epoch": 6.99, "learning_rate": 2.9677520350657484e-05, "loss": 0.6315, "step": 8266, "task_loss": 0.42672035098075867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8447191715240479, "epoch": 6.99, "learning_rate": 2.9674389480275516e-05, "loss": 0.6699, "step": 8267, "task_loss": 0.9529343843460083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8710964322090149, "epoch": 6.99, "learning_rate": 2.9671258609893555e-05, "loss": 0.6928, "step": 8268, "task_loss": 0.9233689308166504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3517892360687256, "epoch": 6.99, "learning_rate": 2.9668127739511586e-05, "loss": 0.5104, "step": 8269, "task_loss": 0.07728548347949982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6382412314414978, "epoch": 6.99, "learning_rate": 2.9664996869129618e-05, "loss": 0.6561, "step": 8270, "task_loss": 0.8234065771102905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6031219363212585, "epoch": 6.99, "learning_rate": 2.966186599874765e-05, "loss": 0.6734, "step": 8271, "task_loss": 0.5661101341247559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0777146816253662, "epoch": 6.99, "learning_rate": 2.965873512836569e-05, "loss": 0.8744, "step": 8272, "task_loss": 1.0266369581222534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25994181632995605, "epoch": 6.99, "learning_rate": 2.965560425798372e-05, "loss": 0.7299, "step": 8273, "task_loss": 0.5192921757698059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3277108371257782, "epoch": 6.99, "learning_rate": 2.9652473387601752e-05, "loss": 0.6538, "step": 8274, "task_loss": 0.6769075989723206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9310105443000793, "epoch": 6.99, "learning_rate": 2.9649342517219787e-05, "loss": 0.8215, "step": 8275, "task_loss": 1.0044070482254028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6757031679153442, "epoch": 7.0, "learning_rate": 2.9646211646837822e-05, "loss": 0.8455, "step": 8276, "task_loss": 0.9293089509010315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3385341167449951, "epoch": 7.0, "learning_rate": 2.9643080776455858e-05, "loss": 0.72, "step": 8277, "task_loss": 1.0852843523025513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8529776930809021, "epoch": 7.0, "learning_rate": 2.963994990607389e-05, "loss": 0.6461, "step": 8278, "task_loss": 0.6830036640167236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.512182354927063, "epoch": 7.0, "learning_rate": 2.9636819035691925e-05, "loss": 0.7253, "step": 8279, "task_loss": 0.43033504486083984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6444985270500183, "epoch": 7.0, "learning_rate": 2.963368816530996e-05, "loss": 0.7773, "step": 8280, "task_loss": 0.35942596197128296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7205936908721924, "epoch": 7.0, "learning_rate": 2.963055729492799e-05, "loss": 0.9441, "step": 8281, "task_loss": 0.41733843088150024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.732996940612793, "epoch": 7.0, "learning_rate": 2.9627426424546023e-05, "loss": 1.12, "step": 8282, "task_loss": 1.7237622737884521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5007354617118835, "epoch": 7.0, "learning_rate": 2.9624295554164062e-05, "loss": 0.6771, "step": 8283, "task_loss": 0.5290696620941162 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7311705350875854, "epoch": 7.0, "learning_rate": 2.9621164683782094e-05, "loss": 0.7748, "step": 8284, "task_loss": 1.049642562866211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8961499929428101, "epoch": 7.0, "learning_rate": 2.9618033813400125e-05, "loss": 1.0042, "step": 8285, "task_loss": 0.6140655279159546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6043975353240967, "epoch": 7.0, "learning_rate": 2.9614902943018157e-05, "loss": 0.6335, "step": 8286, "task_loss": 1.2655041217803955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0003442764282227, "epoch": 7.01, "learning_rate": 2.9611772072636196e-05, "loss": 0.628, "step": 8287, "task_loss": 1.1508104801177979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4548141658306122, "epoch": 7.01, "learning_rate": 2.9608641202254228e-05, "loss": 0.5567, "step": 8288, "task_loss": 0.12835673987865448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5635116100311279, "epoch": 7.01, "learning_rate": 2.960551033187226e-05, "loss": 0.5521, "step": 8289, "task_loss": 0.8541569709777832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.501203715801239, "epoch": 7.01, "learning_rate": 2.9602379461490298e-05, "loss": 0.6824, "step": 8290, "task_loss": 0.4687481224536896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.305403709411621, "epoch": 7.01, "learning_rate": 2.959924859110833e-05, "loss": 0.9124, "step": 8291, "task_loss": 1.7101144790649414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7404775619506836, "epoch": 7.01, "learning_rate": 2.959611772072636e-05, "loss": 0.6127, "step": 8292, "task_loss": 0.787891685962677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3041313588619232, "epoch": 7.01, "learning_rate": 2.9592986850344397e-05, "loss": 0.5077, "step": 8293, "task_loss": 0.20365099608898163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6780048608779907, "epoch": 7.01, "learning_rate": 2.9589855979962432e-05, "loss": 0.698, "step": 8294, "task_loss": 1.301695466041565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45375165343284607, "epoch": 7.01, "learning_rate": 2.9586725109580467e-05, "loss": 0.5657, "step": 8295, "task_loss": 0.07930853962898254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7668604254722595, "epoch": 7.01, "learning_rate": 2.95835942391985e-05, "loss": 0.8544, "step": 8296, "task_loss": 0.24958106875419617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8030464053153992, "epoch": 7.01, "learning_rate": 2.958046336881653e-05, "loss": 0.6824, "step": 8297, "task_loss": 1.2833049297332764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6652414798736572, "epoch": 7.01, "learning_rate": 2.957733249843457e-05, "loss": 0.5697, "step": 8298, "task_loss": 0.2099601775407791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4663553237915039, "epoch": 7.02, "learning_rate": 2.95742016280526e-05, "loss": 0.8593, "step": 8299, "task_loss": 0.60401451587677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.175383448600769, "epoch": 7.02, "learning_rate": 2.9571070757670633e-05, "loss": 0.8234, "step": 8300, "task_loss": 1.2107056379318237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6785612106323242, "epoch": 7.02, "learning_rate": 2.956793988728867e-05, "loss": 0.9465, "step": 8301, "task_loss": 0.9491071105003357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1463512182235718, "epoch": 7.02, "learning_rate": 2.9564809016906703e-05, "loss": 0.6308, "step": 8302, "task_loss": 0.7252556085586548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7106937766075134, "epoch": 7.02, "learning_rate": 2.9561678146524735e-05, "loss": 0.6775, "step": 8303, "task_loss": 0.8139784336090088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5031838417053223, "epoch": 7.02, "learning_rate": 2.9558547276142767e-05, "loss": 0.5506, "step": 8304, "task_loss": 0.7641212344169617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2974848747253418, "epoch": 7.02, "learning_rate": 2.9555416405760805e-05, "loss": 0.7862, "step": 8305, "task_loss": 1.3614537715911865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8155326247215271, "epoch": 7.02, "learning_rate": 2.9552285535378837e-05, "loss": 0.6008, "step": 8306, "task_loss": 0.6444891691207886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9127753973007202, "epoch": 7.02, "learning_rate": 2.954915466499687e-05, "loss": 0.9032, "step": 8307, "task_loss": 0.7909960746765137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.578173041343689, "epoch": 7.02, "learning_rate": 2.95460237946149e-05, "loss": 0.755, "step": 8308, "task_loss": 0.7292129397392273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8236485719680786, "epoch": 7.02, "learning_rate": 2.954289292423294e-05, "loss": 0.6184, "step": 8309, "task_loss": 0.6418868899345398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5954790115356445, "epoch": 7.02, "learning_rate": 2.953976205385097e-05, "loss": 0.6159, "step": 8310, "task_loss": 0.5928274393081665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5394206047058105, "epoch": 7.03, "learning_rate": 2.9536631183469006e-05, "loss": 0.5195, "step": 8311, "task_loss": 0.5669342875480652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4728372097015381, "epoch": 7.03, "learning_rate": 2.9533500313087038e-05, "loss": 0.5634, "step": 8312, "task_loss": 0.7026256322860718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6519684195518494, "epoch": 7.03, "learning_rate": 2.9530369442705076e-05, "loss": 0.7188, "step": 8313, "task_loss": 0.5890498757362366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4128437042236328, "epoch": 7.03, "learning_rate": 2.9527238572323108e-05, "loss": 0.5977, "step": 8314, "task_loss": 1.1083765029907227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1980838775634766, "epoch": 7.03, "learning_rate": 2.952410770194114e-05, "loss": 0.7537, "step": 8315, "task_loss": 2.2525205612182617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9620577096939087, "epoch": 7.03, "learning_rate": 2.952097683155918e-05, "loss": 0.753, "step": 8316, "task_loss": 0.8067613244056702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6200445890426636, "epoch": 7.03, "learning_rate": 2.951784596117721e-05, "loss": 0.7674, "step": 8317, "task_loss": 0.36662977933883667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26172181963920593, "epoch": 7.03, "learning_rate": 2.9514715090795242e-05, "loss": 0.6551, "step": 8318, "task_loss": 0.08137420564889908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6700535416603088, "epoch": 7.03, "learning_rate": 2.9511584220413274e-05, "loss": 0.7798, "step": 8319, "task_loss": 0.3775390088558197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6234840154647827, "epoch": 7.03, "learning_rate": 2.9508453350031312e-05, "loss": 0.6224, "step": 8320, "task_loss": 0.6887292861938477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.883590579032898, "epoch": 7.03, "learning_rate": 2.9505322479649344e-05, "loss": 0.5723, "step": 8321, "task_loss": 0.9843320846557617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.863947331905365, "epoch": 7.03, "learning_rate": 2.9502191609267376e-05, "loss": 0.6382, "step": 8322, "task_loss": 1.0454102754592896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8319348096847534, "epoch": 7.04, "learning_rate": 2.9499060738885408e-05, "loss": 0.7675, "step": 8323, "task_loss": 1.7220377922058105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8638874292373657, "epoch": 7.04, "learning_rate": 2.9495929868503446e-05, "loss": 0.6589, "step": 8324, "task_loss": 0.7224755883216858 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6129441261291504, "epoch": 7.04, "learning_rate": 2.9492798998121478e-05, "loss": 0.8373, "step": 8325, "task_loss": 0.1667366772890091 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4078971743583679, "epoch": 7.04, "learning_rate": 2.9489668127739513e-05, "loss": 0.4557, "step": 8326, "task_loss": 0.5577316880226135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7338078022003174, "epoch": 7.04, "learning_rate": 2.9486537257357548e-05, "loss": 0.62, "step": 8327, "task_loss": 1.645006537437439 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7689014673233032, "epoch": 7.04, "learning_rate": 2.948340638697558e-05, "loss": 0.6687, "step": 8328, "task_loss": 1.5956709384918213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47535979747772217, "epoch": 7.04, "learning_rate": 2.9480275516593615e-05, "loss": 0.5085, "step": 8329, "task_loss": 0.49476632475852966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9519761204719543, "epoch": 7.04, "learning_rate": 2.9477144646211647e-05, "loss": 0.7381, "step": 8330, "task_loss": 1.0267894268035889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6099039316177368, "epoch": 7.04, "learning_rate": 2.9474013775829686e-05, "loss": 0.7073, "step": 8331, "task_loss": 0.3174267113208771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7507873773574829, "epoch": 7.04, "learning_rate": 2.9470882905447717e-05, "loss": 0.5932, "step": 8332, "task_loss": 1.1417258977890015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.586452066898346, "epoch": 7.04, "learning_rate": 2.946775203506575e-05, "loss": 0.5817, "step": 8333, "task_loss": 0.13202017545700073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6832911968231201, "epoch": 7.04, "learning_rate": 2.946462116468378e-05, "loss": 0.6446, "step": 8334, "task_loss": 0.5949323177337646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3632161617279053, "epoch": 7.05, "learning_rate": 2.946149029430182e-05, "loss": 0.8136, "step": 8335, "task_loss": 0.5176236629486084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7284740209579468, "epoch": 7.05, "learning_rate": 2.945835942391985e-05, "loss": 0.888, "step": 8336, "task_loss": 1.9483602046966553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6673047542572021, "epoch": 7.05, "learning_rate": 2.9455228553537883e-05, "loss": 0.5825, "step": 8337, "task_loss": 0.3788047730922699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7463514804840088, "epoch": 7.05, "learning_rate": 2.945209768315592e-05, "loss": 0.5966, "step": 8338, "task_loss": 0.719331681728363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9642539024353027, "epoch": 7.05, "learning_rate": 2.9448966812773953e-05, "loss": 0.7568, "step": 8339, "task_loss": 1.693311333656311 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5490775108337402, "epoch": 7.05, "learning_rate": 2.9445835942391985e-05, "loss": 0.6322, "step": 8340, "task_loss": 0.42148587107658386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5579831600189209, "epoch": 7.05, "learning_rate": 2.9442705072010017e-05, "loss": 0.7151, "step": 8341, "task_loss": 0.9865814447402954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.257936716079712, "epoch": 7.05, "learning_rate": 2.9439574201628056e-05, "loss": 0.9567, "step": 8342, "task_loss": 1.1076250076293945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6079378724098206, "epoch": 7.05, "learning_rate": 2.9436443331246087e-05, "loss": 0.5286, "step": 8343, "task_loss": 0.6095579862594604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37184450030326843, "epoch": 7.05, "learning_rate": 2.9433312460864122e-05, "loss": 0.5806, "step": 8344, "task_loss": 0.8754131197929382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5585755109786987, "epoch": 7.05, "learning_rate": 2.9430181590482154e-05, "loss": 0.6811, "step": 8345, "task_loss": 0.691850483417511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4130551218986511, "epoch": 7.05, "learning_rate": 2.942705072010019e-05, "loss": 0.6256, "step": 8346, "task_loss": 0.9425588846206665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.909311830997467, "epoch": 7.06, "learning_rate": 2.9423919849718225e-05, "loss": 0.7332, "step": 8347, "task_loss": 0.9789543151855469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.397413432598114, "epoch": 7.06, "learning_rate": 2.9420788979336256e-05, "loss": 0.786, "step": 8348, "task_loss": 0.36813005805015564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.805423378944397, "epoch": 7.06, "learning_rate": 2.9417658108954288e-05, "loss": 0.657, "step": 8349, "task_loss": 0.9609295725822449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0089960098266602, "epoch": 7.06, "learning_rate": 2.9414527238572327e-05, "loss": 0.604, "step": 8350, "task_loss": 1.1505802869796753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8323147296905518, "epoch": 7.06, "learning_rate": 2.941139636819036e-05, "loss": 0.6513, "step": 8351, "task_loss": 0.9915481209754944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8365718126296997, "epoch": 7.06, "learning_rate": 2.940826549780839e-05, "loss": 0.7836, "step": 8352, "task_loss": 0.7775787711143494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4506862163543701, "epoch": 7.06, "learning_rate": 2.940513462742643e-05, "loss": 0.495, "step": 8353, "task_loss": 0.5949183702468872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.670773983001709, "epoch": 7.06, "learning_rate": 2.940200375704446e-05, "loss": 0.652, "step": 8354, "task_loss": 0.6408771872520447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6360841393470764, "epoch": 7.06, "learning_rate": 2.9398872886662492e-05, "loss": 0.7203, "step": 8355, "task_loss": 0.2654729187488556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6759210228919983, "epoch": 7.06, "learning_rate": 2.9395742016280524e-05, "loss": 0.6082, "step": 8356, "task_loss": 1.006055474281311 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5370023250579834, "epoch": 7.06, "learning_rate": 2.9392611145898563e-05, "loss": 0.7687, "step": 8357, "task_loss": 1.113146424293518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44977694749832153, "epoch": 7.07, "learning_rate": 2.9389480275516595e-05, "loss": 0.5649, "step": 8358, "task_loss": 1.1804192066192627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6669475436210632, "epoch": 7.07, "learning_rate": 2.9386349405134626e-05, "loss": 0.8261, "step": 8359, "task_loss": 1.0568162202835083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5299201607704163, "epoch": 7.07, "learning_rate": 2.938321853475266e-05, "loss": 0.5451, "step": 8360, "task_loss": 0.8611422777175903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6079142093658447, "epoch": 7.07, "learning_rate": 2.9380087664370697e-05, "loss": 0.6003, "step": 8361, "task_loss": 1.270713448524475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45716309547424316, "epoch": 7.07, "learning_rate": 2.9376956793988732e-05, "loss": 0.6947, "step": 8362, "task_loss": 0.19058650732040405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5224007964134216, "epoch": 7.07, "learning_rate": 2.9373825923606764e-05, "loss": 0.7694, "step": 8363, "task_loss": 0.14001256227493286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5531502366065979, "epoch": 7.07, "learning_rate": 2.93706950532248e-05, "loss": 0.7021, "step": 8364, "task_loss": 0.7022714018821716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.535232663154602, "epoch": 7.07, "learning_rate": 2.9367564182842834e-05, "loss": 0.5145, "step": 8365, "task_loss": 0.6700655817985535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4933353066444397, "epoch": 7.07, "learning_rate": 2.9364433312460866e-05, "loss": 0.4529, "step": 8366, "task_loss": 0.304262638092041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36001619696617126, "epoch": 7.07, "learning_rate": 2.9361302442078897e-05, "loss": 0.4356, "step": 8367, "task_loss": 0.09022704511880875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4008488655090332, "epoch": 7.07, "learning_rate": 2.9358171571696936e-05, "loss": 0.5768, "step": 8368, "task_loss": 0.8851678371429443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6072884798049927, "epoch": 7.07, "learning_rate": 2.9355040701314968e-05, "loss": 0.6185, "step": 8369, "task_loss": 0.653599739074707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9832698106765747, "epoch": 7.08, "learning_rate": 2.9351909830933e-05, "loss": 0.7962, "step": 8370, "task_loss": 0.8289151191711426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9965639114379883, "epoch": 7.08, "learning_rate": 2.934877896055103e-05, "loss": 0.6321, "step": 8371, "task_loss": 1.637667179107666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6466943025588989, "epoch": 7.08, "learning_rate": 2.934564809016907e-05, "loss": 0.7185, "step": 8372, "task_loss": 1.939172625541687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7484586238861084, "epoch": 7.08, "learning_rate": 2.9342517219787102e-05, "loss": 0.8285, "step": 8373, "task_loss": 1.6172107458114624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4636559784412384, "epoch": 7.08, "learning_rate": 2.9339386349405134e-05, "loss": 0.7622, "step": 8374, "task_loss": 0.39617785811424255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6572720408439636, "epoch": 7.08, "learning_rate": 2.9336255479023172e-05, "loss": 0.5843, "step": 8375, "task_loss": 0.09491807222366333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5226950645446777, "epoch": 7.08, "learning_rate": 2.9333124608641204e-05, "loss": 0.6463, "step": 8376, "task_loss": 0.9544906616210938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6478410959243774, "epoch": 7.08, "learning_rate": 2.9329993738259236e-05, "loss": 0.6844, "step": 8377, "task_loss": 0.8824155926704407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7192875742912292, "epoch": 7.08, "learning_rate": 2.932686286787727e-05, "loss": 0.6068, "step": 8378, "task_loss": 1.0658384561538696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7030977010726929, "epoch": 7.08, "learning_rate": 2.9323731997495306e-05, "loss": 0.7623, "step": 8379, "task_loss": 1.1669188737869263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8288366794586182, "epoch": 7.08, "learning_rate": 2.932060112711334e-05, "loss": 0.7088, "step": 8380, "task_loss": 2.2414560317993164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7514395713806152, "epoch": 7.08, "learning_rate": 2.9317470256731373e-05, "loss": 0.6024, "step": 8381, "task_loss": 1.1083250045776367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1825751066207886, "epoch": 7.09, "learning_rate": 2.9314339386349405e-05, "loss": 0.8349, "step": 8382, "task_loss": 1.11164391040802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3609377145767212, "epoch": 7.09, "learning_rate": 2.9311208515967443e-05, "loss": 0.6968, "step": 8383, "task_loss": 1.25571870803833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.769501805305481, "epoch": 7.09, "learning_rate": 2.9308077645585475e-05, "loss": 0.5197, "step": 8384, "task_loss": 1.4435622692108154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46155059337615967, "epoch": 7.09, "learning_rate": 2.9304946775203507e-05, "loss": 0.5953, "step": 8385, "task_loss": 0.8137142658233643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5700051784515381, "epoch": 7.09, "learning_rate": 2.930181590482154e-05, "loss": 0.7934, "step": 8386, "task_loss": 0.4249722957611084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0433627367019653, "epoch": 7.09, "learning_rate": 2.9298685034439577e-05, "loss": 0.8478, "step": 8387, "task_loss": 0.6368272304534912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5494570732116699, "epoch": 7.09, "learning_rate": 2.929555416405761e-05, "loss": 0.6659, "step": 8388, "task_loss": 0.4476962983608246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6525925397872925, "epoch": 7.09, "learning_rate": 2.929242329367564e-05, "loss": 0.5141, "step": 8389, "task_loss": 1.2259167432785034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0710070133209229, "epoch": 7.09, "learning_rate": 2.928929242329368e-05, "loss": 0.7469, "step": 8390, "task_loss": 0.9697255492210388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0444247722625732, "epoch": 7.09, "learning_rate": 2.928616155291171e-05, "loss": 0.7506, "step": 8391, "task_loss": 1.4048583507537842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1489598751068115, "epoch": 7.09, "learning_rate": 2.9283030682529743e-05, "loss": 0.9545, "step": 8392, "task_loss": 1.3230764865875244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7074766755104065, "epoch": 7.09, "learning_rate": 2.9279899812147778e-05, "loss": 0.7927, "step": 8393, "task_loss": 0.7097452282905579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3774504065513611, "epoch": 7.1, "learning_rate": 2.9276768941765813e-05, "loss": 0.6719, "step": 8394, "task_loss": 0.967192530632019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4025534391403198, "epoch": 7.1, "learning_rate": 2.9273638071383845e-05, "loss": 0.5521, "step": 8395, "task_loss": 0.4960082769393921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8260334730148315, "epoch": 7.1, "learning_rate": 2.927050720100188e-05, "loss": 0.82, "step": 8396, "task_loss": 0.7959938049316406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45791202783584595, "epoch": 7.1, "learning_rate": 2.9267376330619912e-05, "loss": 0.6592, "step": 8397, "task_loss": 1.0171210765838623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9668603539466858, "epoch": 7.1, "learning_rate": 2.926424546023795e-05, "loss": 0.7642, "step": 8398, "task_loss": 1.804939866065979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6358003616333008, "epoch": 7.1, "learning_rate": 2.9261114589855982e-05, "loss": 0.9043, "step": 8399, "task_loss": 1.1619596481323242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7378454208374023, "epoch": 7.1, "learning_rate": 2.9257983719474014e-05, "loss": 0.7915, "step": 8400, "task_loss": 0.4507439136505127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0098702907562256, "epoch": 7.1, "learning_rate": 2.9254852849092053e-05, "loss": 0.7411, "step": 8401, "task_loss": 0.6980133652687073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7672312259674072, "epoch": 7.1, "learning_rate": 2.9251721978710084e-05, "loss": 0.7011, "step": 8402, "task_loss": 1.281646728515625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45192790031433105, "epoch": 7.1, "learning_rate": 2.9248591108328116e-05, "loss": 0.6997, "step": 8403, "task_loss": 0.7668692469596863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5556004047393799, "epoch": 7.1, "learning_rate": 2.9245460237946148e-05, "loss": 0.6969, "step": 8404, "task_loss": 0.2961523234844208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5933274626731873, "epoch": 7.1, "learning_rate": 2.9242329367564186e-05, "loss": 0.6935, "step": 8405, "task_loss": 0.7611309885978699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4950964152812958, "epoch": 7.11, "learning_rate": 2.9239198497182218e-05, "loss": 0.876, "step": 8406, "task_loss": 0.8504976630210876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6239908933639526, "epoch": 7.11, "learning_rate": 2.923606762680025e-05, "loss": 0.6772, "step": 8407, "task_loss": 1.0180819034576416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7426325678825378, "epoch": 7.11, "learning_rate": 2.9232936756418282e-05, "loss": 0.8633, "step": 8408, "task_loss": 1.7080402374267578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5002589225769043, "epoch": 7.11, "learning_rate": 2.922980588603632e-05, "loss": 0.5745, "step": 8409, "task_loss": 1.1212892532348633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3810337781906128, "epoch": 7.11, "learning_rate": 2.9226675015654352e-05, "loss": 0.475, "step": 8410, "task_loss": 0.2896597981452942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8141155242919922, "epoch": 7.11, "learning_rate": 2.9223544145272387e-05, "loss": 0.6559, "step": 8411, "task_loss": 0.29846805334091187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7127605676651001, "epoch": 7.11, "learning_rate": 2.9220413274890423e-05, "loss": 0.5718, "step": 8412, "task_loss": 0.5656725764274597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.657057523727417, "epoch": 7.11, "learning_rate": 2.9217282404508454e-05, "loss": 0.7406, "step": 8413, "task_loss": 0.9394055604934692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.835350513458252, "epoch": 7.11, "learning_rate": 2.921415153412649e-05, "loss": 0.8557, "step": 8414, "task_loss": 0.6517860293388367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4607698917388916, "epoch": 7.11, "learning_rate": 2.921102066374452e-05, "loss": 0.6641, "step": 8415, "task_loss": 0.297634094953537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.432747483253479, "epoch": 7.11, "learning_rate": 2.920788979336256e-05, "loss": 0.4862, "step": 8416, "task_loss": 0.7287262678146362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6560622453689575, "epoch": 7.11, "learning_rate": 2.920475892298059e-05, "loss": 0.5433, "step": 8417, "task_loss": 0.6486332416534424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.62632155418396, "epoch": 7.12, "learning_rate": 2.9201628052598623e-05, "loss": 0.6921, "step": 8418, "task_loss": 0.32609447836875916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7859692573547363, "epoch": 7.12, "learning_rate": 2.9198497182216655e-05, "loss": 0.7793, "step": 8419, "task_loss": 0.6105086803436279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6712459325790405, "epoch": 7.12, "learning_rate": 2.9195366311834694e-05, "loss": 0.713, "step": 8420, "task_loss": 0.813909649848938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7596548199653625, "epoch": 7.12, "learning_rate": 2.9192235441452725e-05, "loss": 0.8128, "step": 8421, "task_loss": 0.9687859416007996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0775786638259888, "epoch": 7.12, "learning_rate": 2.9189104571070757e-05, "loss": 0.7493, "step": 8422, "task_loss": 1.224519968032837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6293587684631348, "epoch": 7.12, "learning_rate": 2.918597370068879e-05, "loss": 0.7208, "step": 8423, "task_loss": 0.5532387495040894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7482473254203796, "epoch": 7.12, "learning_rate": 2.9182842830306828e-05, "loss": 0.7144, "step": 8424, "task_loss": 0.22691401839256287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5333763360977173, "epoch": 7.12, "learning_rate": 2.917971195992486e-05, "loss": 0.7015, "step": 8425, "task_loss": 1.2896177768707275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5384747385978699, "epoch": 7.12, "learning_rate": 2.917658108954289e-05, "loss": 0.6517, "step": 8426, "task_loss": 0.44909995794296265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5467045307159424, "epoch": 7.12, "learning_rate": 2.917345021916093e-05, "loss": 0.5937, "step": 8427, "task_loss": 0.6755611896514893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4871315658092499, "epoch": 7.12, "learning_rate": 2.917031934877896e-05, "loss": 0.7896, "step": 8428, "task_loss": 0.47339949011802673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0428558588027954, "epoch": 7.13, "learning_rate": 2.9167188478396997e-05, "loss": 0.8464, "step": 8429, "task_loss": 1.4831196069717407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4885130524635315, "epoch": 7.13, "learning_rate": 2.916405760801503e-05, "loss": 0.6141, "step": 8430, "task_loss": 0.8234672546386719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5767615437507629, "epoch": 7.13, "learning_rate": 2.9160926737633064e-05, "loss": 0.7512, "step": 8431, "task_loss": 0.200160950422287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7338234186172485, "epoch": 7.13, "learning_rate": 2.91577958672511e-05, "loss": 0.5182, "step": 8432, "task_loss": 0.7331130504608154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9372183084487915, "epoch": 7.13, "learning_rate": 2.915466499686913e-05, "loss": 0.6776, "step": 8433, "task_loss": 0.8087535500526428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5369473695755005, "epoch": 7.13, "learning_rate": 2.9151534126487162e-05, "loss": 0.5728, "step": 8434, "task_loss": 1.1643040180206299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3704882860183716, "epoch": 7.13, "learning_rate": 2.91484032561052e-05, "loss": 0.6093, "step": 8435, "task_loss": 0.49414217472076416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7320324182510376, "epoch": 7.13, "learning_rate": 2.9145272385723233e-05, "loss": 0.685, "step": 8436, "task_loss": 0.42644914984703064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.680963397026062, "epoch": 7.13, "learning_rate": 2.9142141515341264e-05, "loss": 0.726, "step": 8437, "task_loss": 0.5905590057373047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4458405077457428, "epoch": 7.13, "learning_rate": 2.9139010644959303e-05, "loss": 0.8117, "step": 8438, "task_loss": 0.9344676733016968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7073341012001038, "epoch": 7.13, "learning_rate": 2.9135879774577335e-05, "loss": 0.5159, "step": 8439, "task_loss": 0.5714548230171204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8468595147132874, "epoch": 7.13, "learning_rate": 2.9132748904195367e-05, "loss": 0.6911, "step": 8440, "task_loss": 0.8470777273178101 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6917262077331543, "epoch": 7.14, "learning_rate": 2.91296180338134e-05, "loss": 0.9045, "step": 8441, "task_loss": 1.0598526000976562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9073841571807861, "epoch": 7.14, "learning_rate": 2.9126487163431437e-05, "loss": 0.7491, "step": 8442, "task_loss": 1.443321704864502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7316235899925232, "epoch": 7.14, "learning_rate": 2.912335629304947e-05, "loss": 0.7873, "step": 8443, "task_loss": 0.9511703848838806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22676518559455872, "epoch": 7.14, "learning_rate": 2.91202254226675e-05, "loss": 0.656, "step": 8444, "task_loss": 0.3447759747505188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8178427815437317, "epoch": 7.14, "learning_rate": 2.9117094552285536e-05, "loss": 0.7211, "step": 8445, "task_loss": 1.5397361516952515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5460292100906372, "epoch": 7.14, "learning_rate": 2.911396368190357e-05, "loss": 0.7662, "step": 8446, "task_loss": 0.6354162693023682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0405540466308594, "epoch": 7.14, "learning_rate": 2.9110832811521606e-05, "loss": 0.7107, "step": 8447, "task_loss": 0.9243070483207703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6672730445861816, "epoch": 7.14, "learning_rate": 2.9107701941139638e-05, "loss": 0.64, "step": 8448, "task_loss": 1.2669674158096313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3988312780857086, "epoch": 7.14, "learning_rate": 2.9104571070757673e-05, "loss": 0.5405, "step": 8449, "task_loss": 0.2477552592754364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.626816987991333, "epoch": 7.14, "learning_rate": 2.9101440200375708e-05, "loss": 0.7124, "step": 8450, "task_loss": 1.1811225414276123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5239540934562683, "epoch": 7.14, "learning_rate": 2.909830932999374e-05, "loss": 0.5216, "step": 8451, "task_loss": 0.4284743666648865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4397891759872437, "epoch": 7.14, "learning_rate": 2.909517845961177e-05, "loss": 0.7303, "step": 8452, "task_loss": 0.9282432794570923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5073019862174988, "epoch": 7.15, "learning_rate": 2.909204758922981e-05, "loss": 0.5231, "step": 8453, "task_loss": 0.1532255858182907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43695682287216187, "epoch": 7.15, "learning_rate": 2.9088916718847842e-05, "loss": 0.6329, "step": 8454, "task_loss": 0.8769485950469971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.900570273399353, "epoch": 7.15, "learning_rate": 2.9085785848465874e-05, "loss": 0.8067, "step": 8455, "task_loss": 0.8334083557128906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34340232610702515, "epoch": 7.15, "learning_rate": 2.9082654978083906e-05, "loss": 0.6261, "step": 8456, "task_loss": 0.3603319525718689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3870607614517212, "epoch": 7.15, "learning_rate": 2.9079524107701944e-05, "loss": 0.9039, "step": 8457, "task_loss": 0.9872869849205017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.667678713798523, "epoch": 7.15, "learning_rate": 2.9076393237319976e-05, "loss": 0.6988, "step": 8458, "task_loss": 0.3934127986431122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7701703310012817, "epoch": 7.15, "learning_rate": 2.9073262366938008e-05, "loss": 0.7724, "step": 8459, "task_loss": 0.5397769212722778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7447539567947388, "epoch": 7.15, "learning_rate": 2.9070131496556046e-05, "loss": 0.9174, "step": 8460, "task_loss": 0.41293102502822876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4428827464580536, "epoch": 7.15, "learning_rate": 2.9067000626174078e-05, "loss": 0.5962, "step": 8461, "task_loss": 0.6082631945610046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9418858289718628, "epoch": 7.15, "learning_rate": 2.906386975579211e-05, "loss": 0.6971, "step": 8462, "task_loss": 0.30975010991096497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.8305273056030273, "epoch": 7.15, "learning_rate": 2.9060738885410145e-05, "loss": 1.1781, "step": 8463, "task_loss": 1.665222406387329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0035510063171387, "epoch": 7.15, "learning_rate": 2.905760801502818e-05, "loss": 0.6248, "step": 8464, "task_loss": 0.7656728029251099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36049604415893555, "epoch": 7.16, "learning_rate": 2.9054477144646215e-05, "loss": 0.6226, "step": 8465, "task_loss": 0.7675876021385193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5711262822151184, "epoch": 7.16, "learning_rate": 2.9051346274264247e-05, "loss": 0.6191, "step": 8466, "task_loss": 0.3874540627002716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3112791180610657, "epoch": 7.16, "learning_rate": 2.904821540388228e-05, "loss": 0.4258, "step": 8467, "task_loss": 0.492402046918869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5945789813995361, "epoch": 7.16, "learning_rate": 2.9045084533500317e-05, "loss": 0.6672, "step": 8468, "task_loss": 0.2650187015533447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.344730257987976, "epoch": 7.16, "learning_rate": 2.904195366311835e-05, "loss": 0.7544, "step": 8469, "task_loss": 1.241779088973999 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31430190801620483, "epoch": 7.16, "learning_rate": 2.903882279273638e-05, "loss": 0.5576, "step": 8470, "task_loss": 0.21608427166938782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0115684270858765, "epoch": 7.16, "learning_rate": 2.9035691922354413e-05, "loss": 0.8615, "step": 8471, "task_loss": 0.4113961160182953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7125699520111084, "epoch": 7.16, "learning_rate": 2.903256105197245e-05, "loss": 0.6906, "step": 8472, "task_loss": 0.8146806955337524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5456189513206482, "epoch": 7.16, "learning_rate": 2.9029430181590483e-05, "loss": 0.689, "step": 8473, "task_loss": 0.3977053463459015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9667834043502808, "epoch": 7.16, "learning_rate": 2.9026299311208515e-05, "loss": 0.8636, "step": 8474, "task_loss": 1.616628885269165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7576895952224731, "epoch": 7.16, "learning_rate": 2.9023168440826553e-05, "loss": 0.6685, "step": 8475, "task_loss": 0.888723611831665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7220460772514343, "epoch": 7.16, "learning_rate": 2.9020037570444585e-05, "loss": 0.9097, "step": 8476, "task_loss": 0.9115903973579407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6380133032798767, "epoch": 7.17, "learning_rate": 2.9016906700062617e-05, "loss": 0.8118, "step": 8477, "task_loss": 1.0588537454605103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7529407143592834, "epoch": 7.17, "learning_rate": 2.9013775829680652e-05, "loss": 0.8688, "step": 8478, "task_loss": 1.2142201662063599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45243969559669495, "epoch": 7.17, "learning_rate": 2.9010644959298687e-05, "loss": 0.672, "step": 8479, "task_loss": 0.7506532669067383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5928336381912231, "epoch": 7.17, "learning_rate": 2.900751408891672e-05, "loss": 0.6921, "step": 8480, "task_loss": 0.44101956486701965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5681997537612915, "epoch": 7.17, "learning_rate": 2.9004383218534754e-05, "loss": 0.6181, "step": 8481, "task_loss": 0.5914584994316101 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5081037282943726, "epoch": 7.17, "learning_rate": 2.9001252348152786e-05, "loss": 0.7344, "step": 8482, "task_loss": 0.9676117897033691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9501303434371948, "epoch": 7.17, "learning_rate": 2.8998121477770825e-05, "loss": 0.7678, "step": 8483, "task_loss": 2.2637195587158203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1668572425842285, "epoch": 7.17, "learning_rate": 2.8994990607388856e-05, "loss": 0.8538, "step": 8484, "task_loss": 0.5309388637542725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6079646944999695, "epoch": 7.17, "learning_rate": 2.8991859737006888e-05, "loss": 0.8673, "step": 8485, "task_loss": 0.789552628993988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.724735677242279, "epoch": 7.17, "learning_rate": 2.8988728866624927e-05, "loss": 0.7105, "step": 8486, "task_loss": 0.4593677222728729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3990280330181122, "epoch": 7.17, "learning_rate": 2.898559799624296e-05, "loss": 0.678, "step": 8487, "task_loss": 0.5321353673934937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.470764696598053, "epoch": 7.17, "learning_rate": 2.898246712586099e-05, "loss": 0.7046, "step": 8488, "task_loss": 0.505430281162262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8476776480674744, "epoch": 7.18, "learning_rate": 2.8979336255479022e-05, "loss": 0.7409, "step": 8489, "task_loss": 0.5985147356987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.734761655330658, "epoch": 7.18, "learning_rate": 2.897620538509706e-05, "loss": 0.6901, "step": 8490, "task_loss": 0.5357382297515869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.05806303024292, "epoch": 7.18, "learning_rate": 2.8973074514715092e-05, "loss": 0.8784, "step": 8491, "task_loss": 1.0840471982955933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.336719810962677, "epoch": 7.18, "learning_rate": 2.8969943644333124e-05, "loss": 0.6815, "step": 8492, "task_loss": 0.5295722484588623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.555902361869812, "epoch": 7.18, "learning_rate": 2.8966812773951156e-05, "loss": 0.8378, "step": 8493, "task_loss": 0.7135011553764343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6154170036315918, "epoch": 7.18, "learning_rate": 2.8963681903569195e-05, "loss": 0.5791, "step": 8494, "task_loss": 0.8280824422836304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5383581519126892, "epoch": 7.18, "learning_rate": 2.8960551033187226e-05, "loss": 0.4512, "step": 8495, "task_loss": 0.310337632894516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4695965051651001, "epoch": 7.18, "learning_rate": 2.895742016280526e-05, "loss": 0.5653, "step": 8496, "task_loss": 0.25509247183799744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3477236032485962, "epoch": 7.18, "learning_rate": 2.8954289292423297e-05, "loss": 0.7507, "step": 8497, "task_loss": 0.06938290596008301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9905865788459778, "epoch": 7.18, "learning_rate": 2.895115842204133e-05, "loss": 0.681, "step": 8498, "task_loss": 0.4209606349468231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6659349203109741, "epoch": 7.18, "learning_rate": 2.8948027551659364e-05, "loss": 0.7803, "step": 8499, "task_loss": 1.0511351823806763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7859931588172913, "epoch": 7.19, "learning_rate": 2.8944896681277395e-05, "loss": 0.6759, "step": 8500, "task_loss": 0.7900732755661011 }, { "epoch": 7.19, "eval_accuracy": 0.8925940594059406, "eval_loss": 0.4681185483932495, "eval_runtime": 207.4777, "eval_samples_per_second": 121.7, "eval_steps_per_second": 0.954, "step": 8500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3277024030685425, "epoch": 7.19, "learning_rate": 2.8941765810895434e-05, "loss": 0.9429, "step": 8501, "task_loss": 1.782302975654602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6645029783248901, "epoch": 7.19, "learning_rate": 2.8938634940513466e-05, "loss": 0.8263, "step": 8502, "task_loss": 0.45978352427482605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43341535329818726, "epoch": 7.19, "learning_rate": 2.8935504070131498e-05, "loss": 0.5745, "step": 8503, "task_loss": 0.10814512521028519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7202534079551697, "epoch": 7.19, "learning_rate": 2.893237319974953e-05, "loss": 0.7292, "step": 8504, "task_loss": 1.259522795677185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1052767038345337, "epoch": 7.19, "learning_rate": 2.8929242329367568e-05, "loss": 0.8795, "step": 8505, "task_loss": 0.647347092628479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0731115341186523, "epoch": 7.19, "learning_rate": 2.89261114589856e-05, "loss": 0.7004, "step": 8506, "task_loss": 0.9314419031143188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9268348217010498, "epoch": 7.19, "learning_rate": 2.892298058860363e-05, "loss": 0.8548, "step": 8507, "task_loss": 1.020493745803833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4610937833786011, "epoch": 7.19, "learning_rate": 2.8919849718221663e-05, "loss": 0.6173, "step": 8508, "task_loss": 0.025457588955760002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.695428729057312, "epoch": 7.19, "learning_rate": 2.8916718847839702e-05, "loss": 0.6977, "step": 8509, "task_loss": 0.5885156989097595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5256944298744202, "epoch": 7.19, "learning_rate": 2.8913587977457734e-05, "loss": 0.5892, "step": 8510, "task_loss": 0.9806305170059204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9952657222747803, "epoch": 7.19, "learning_rate": 2.8910457107075765e-05, "loss": 1.005, "step": 8511, "task_loss": 1.8869261741638184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9615174531936646, "epoch": 7.2, "learning_rate": 2.8907326236693804e-05, "loss": 0.7454, "step": 8512, "task_loss": 1.0767946243286133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8628499507904053, "epoch": 7.2, "learning_rate": 2.8904195366311836e-05, "loss": 0.7568, "step": 8513, "task_loss": 0.48640450835227966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0952292680740356, "epoch": 7.2, "learning_rate": 2.890106449592987e-05, "loss": 0.8394, "step": 8514, "task_loss": 1.017411470413208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3835301995277405, "epoch": 7.2, "learning_rate": 2.8897933625547903e-05, "loss": 0.5673, "step": 8515, "task_loss": 0.7085422873497009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6088808178901672, "epoch": 7.2, "learning_rate": 2.8894802755165938e-05, "loss": 0.7571, "step": 8516, "task_loss": 0.6759728789329529 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7588416337966919, "epoch": 7.2, "learning_rate": 2.8891671884783973e-05, "loss": 0.555, "step": 8517, "task_loss": 0.9012501835823059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.616608738899231, "epoch": 7.2, "learning_rate": 2.8888541014402005e-05, "loss": 0.6008, "step": 8518, "task_loss": 0.24728742241859436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41181379556655884, "epoch": 7.2, "learning_rate": 2.8885410144020037e-05, "loss": 0.4826, "step": 8519, "task_loss": 0.4661071300506592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5657882690429688, "epoch": 7.2, "learning_rate": 2.8882279273638075e-05, "loss": 0.6705, "step": 8520, "task_loss": 0.6164771318435669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8003948926925659, "epoch": 7.2, "learning_rate": 2.8879148403256107e-05, "loss": 0.7084, "step": 8521, "task_loss": 0.4315544068813324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6952469348907471, "epoch": 7.2, "learning_rate": 2.887601753287414e-05, "loss": 0.8236, "step": 8522, "task_loss": 0.7669837474822998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7853109240531921, "epoch": 7.2, "learning_rate": 2.8872886662492177e-05, "loss": 0.7232, "step": 8523, "task_loss": 0.8910977840423584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5437528491020203, "epoch": 7.21, "learning_rate": 2.886975579211021e-05, "loss": 0.5924, "step": 8524, "task_loss": 0.7517634630203247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7424126267433167, "epoch": 7.21, "learning_rate": 2.886662492172824e-05, "loss": 0.7227, "step": 8525, "task_loss": 0.19784097373485565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9888759255409241, "epoch": 7.21, "learning_rate": 2.8863494051346273e-05, "loss": 0.6339, "step": 8526, "task_loss": 1.0070085525512695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9545824527740479, "epoch": 7.21, "learning_rate": 2.886036318096431e-05, "loss": 0.6815, "step": 8527, "task_loss": 0.5094006061553955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4287201464176178, "epoch": 7.21, "learning_rate": 2.8857232310582343e-05, "loss": 0.7248, "step": 8528, "task_loss": 0.5410181879997253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6023178100585938, "epoch": 7.21, "learning_rate": 2.8854101440200375e-05, "loss": 0.6076, "step": 8529, "task_loss": 0.12837500870227814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6872901320457458, "epoch": 7.21, "learning_rate": 2.885097056981841e-05, "loss": 0.7372, "step": 8530, "task_loss": 1.489762783050537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6673352718353271, "epoch": 7.21, "learning_rate": 2.8847839699436445e-05, "loss": 0.8286, "step": 8531, "task_loss": 0.5886353850364685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7891655564308167, "epoch": 7.21, "learning_rate": 2.884470882905448e-05, "loss": 0.6106, "step": 8532, "task_loss": 0.8517605662345886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9494723081588745, "epoch": 7.21, "learning_rate": 2.8841577958672512e-05, "loss": 0.8706, "step": 8533, "task_loss": 1.5992943048477173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41990435123443604, "epoch": 7.21, "learning_rate": 2.883844708829055e-05, "loss": 0.62, "step": 8534, "task_loss": 0.12419299781322479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1201229095458984, "epoch": 7.21, "learning_rate": 2.8835316217908582e-05, "loss": 0.7099, "step": 8535, "task_loss": 0.6658726334571838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.698501706123352, "epoch": 7.22, "learning_rate": 2.8832185347526614e-05, "loss": 0.7413, "step": 8536, "task_loss": 0.4935651123523712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.6590805053710938, "epoch": 7.22, "learning_rate": 2.8829054477144646e-05, "loss": 0.8073, "step": 8537, "task_loss": 0.8711647391319275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8418596386909485, "epoch": 7.22, "learning_rate": 2.8825923606762684e-05, "loss": 0.7221, "step": 8538, "task_loss": 0.8303602337837219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6755740642547607, "epoch": 7.22, "learning_rate": 2.8822792736380716e-05, "loss": 0.7884, "step": 8539, "task_loss": 1.5591596364974976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6425141096115112, "epoch": 7.22, "learning_rate": 2.8819661865998748e-05, "loss": 0.5688, "step": 8540, "task_loss": 0.7017989754676819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6662578582763672, "epoch": 7.22, "learning_rate": 2.881653099561678e-05, "loss": 0.6663, "step": 8541, "task_loss": 0.36611446738243103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6353689432144165, "epoch": 7.22, "learning_rate": 2.881340012523482e-05, "loss": 0.7029, "step": 8542, "task_loss": 0.876526951789856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9469740986824036, "epoch": 7.22, "learning_rate": 2.881026925485285e-05, "loss": 0.798, "step": 8543, "task_loss": 0.5206006765365601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8707156181335449, "epoch": 7.22, "learning_rate": 2.8807138384470882e-05, "loss": 0.7124, "step": 8544, "task_loss": 0.551642894744873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4961901307106018, "epoch": 7.22, "learning_rate": 2.8804007514088917e-05, "loss": 0.5875, "step": 8545, "task_loss": 0.24061128497123718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8031316995620728, "epoch": 7.22, "learning_rate": 2.8800876643706952e-05, "loss": 0.7292, "step": 8546, "task_loss": 1.6597280502319336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4622085392475128, "epoch": 7.22, "learning_rate": 2.8797745773324984e-05, "loss": 0.6205, "step": 8547, "task_loss": 0.764066219329834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40466898679733276, "epoch": 7.23, "learning_rate": 2.879461490294302e-05, "loss": 0.5354, "step": 8548, "task_loss": 0.8652380108833313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.905463457107544, "epoch": 7.23, "learning_rate": 2.8791484032561054e-05, "loss": 0.7787, "step": 8549, "task_loss": 0.7638587951660156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43188029527664185, "epoch": 7.23, "learning_rate": 2.878835316217909e-05, "loss": 0.5996, "step": 8550, "task_loss": 0.45867919921875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5483752489089966, "epoch": 7.23, "learning_rate": 2.878522229179712e-05, "loss": 0.6033, "step": 8551, "task_loss": 0.9483766555786133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5407836437225342, "epoch": 7.23, "learning_rate": 2.8782091421415153e-05, "loss": 0.7058, "step": 8552, "task_loss": 1.0136126279830933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.781080961227417, "epoch": 7.23, "learning_rate": 2.877896055103319e-05, "loss": 0.6439, "step": 8553, "task_loss": 0.1750170737504959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48610514402389526, "epoch": 7.23, "learning_rate": 2.8775829680651223e-05, "loss": 0.6416, "step": 8554, "task_loss": 0.2671133577823639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8188432455062866, "epoch": 7.23, "learning_rate": 2.8772698810269255e-05, "loss": 0.8145, "step": 8555, "task_loss": 0.6973693370819092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5930465459823608, "epoch": 7.23, "learning_rate": 2.8769567939887287e-05, "loss": 0.7252, "step": 8556, "task_loss": 0.6312101483345032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44140899181365967, "epoch": 7.23, "learning_rate": 2.8766437069505326e-05, "loss": 0.6576, "step": 8557, "task_loss": 1.2635352611541748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5833554267883301, "epoch": 7.23, "learning_rate": 2.8763306199123357e-05, "loss": 0.5322, "step": 8558, "task_loss": 0.5675053000450134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.783807635307312, "epoch": 7.23, "learning_rate": 2.876017532874139e-05, "loss": 0.656, "step": 8559, "task_loss": 0.8085671067237854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6929118037223816, "epoch": 7.24, "learning_rate": 2.8757044458359428e-05, "loss": 0.6761, "step": 8560, "task_loss": 0.47090795636177063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6358998417854309, "epoch": 7.24, "learning_rate": 2.875391358797746e-05, "loss": 0.6954, "step": 8561, "task_loss": 0.9831289052963257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7212762832641602, "epoch": 7.24, "learning_rate": 2.875078271759549e-05, "loss": 0.8575, "step": 8562, "task_loss": 0.5489591956138611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6051925420761108, "epoch": 7.24, "learning_rate": 2.8747651847213526e-05, "loss": 0.5902, "step": 8563, "task_loss": 0.40009844303131104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8441596031188965, "epoch": 7.24, "learning_rate": 2.874452097683156e-05, "loss": 0.746, "step": 8564, "task_loss": 1.2800731658935547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6737219095230103, "epoch": 7.24, "learning_rate": 2.8741390106449593e-05, "loss": 0.8575, "step": 8565, "task_loss": 0.6792552471160889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6885948777198792, "epoch": 7.24, "learning_rate": 2.873825923606763e-05, "loss": 0.7695, "step": 8566, "task_loss": 0.4158936142921448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2824878692626953, "epoch": 7.24, "learning_rate": 2.873512836568566e-05, "loss": 0.4979, "step": 8567, "task_loss": 0.507425844669342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42805615067481995, "epoch": 7.24, "learning_rate": 2.87319974953037e-05, "loss": 0.6108, "step": 8568, "task_loss": 0.23912085592746735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6319940090179443, "epoch": 7.24, "learning_rate": 2.872886662492173e-05, "loss": 0.5127, "step": 8569, "task_loss": 0.6711313724517822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8848299384117126, "epoch": 7.24, "learning_rate": 2.8725735754539762e-05, "loss": 0.635, "step": 8570, "task_loss": 0.79273521900177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.606544017791748, "epoch": 7.24, "learning_rate": 2.87226048841578e-05, "loss": 0.6605, "step": 8571, "task_loss": 0.9383052587509155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5888261795043945, "epoch": 7.25, "learning_rate": 2.8719474013775833e-05, "loss": 0.5737, "step": 8572, "task_loss": 0.27636510133743286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5159081220626831, "epoch": 7.25, "learning_rate": 2.8716343143393865e-05, "loss": 0.5438, "step": 8573, "task_loss": 1.3573548793792725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.206367015838623, "epoch": 7.25, "learning_rate": 2.8713212273011896e-05, "loss": 0.9802, "step": 8574, "task_loss": 1.0719966888427734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.626648485660553, "epoch": 7.25, "learning_rate": 2.8710081402629935e-05, "loss": 0.7844, "step": 8575, "task_loss": 0.6997456550598145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6875728964805603, "epoch": 7.25, "learning_rate": 2.8706950532247967e-05, "loss": 0.9363, "step": 8576, "task_loss": 0.16816899180412292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7307252883911133, "epoch": 7.25, "learning_rate": 2.8703819661866e-05, "loss": 0.6178, "step": 8577, "task_loss": 0.16525757312774658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9495736360549927, "epoch": 7.25, "learning_rate": 2.870068879148403e-05, "loss": 0.705, "step": 8578, "task_loss": 1.8565824031829834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3775675892829895, "epoch": 7.25, "learning_rate": 2.869755792110207e-05, "loss": 0.7291, "step": 8579, "task_loss": 0.6085376143455505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2472071647644043, "epoch": 7.25, "learning_rate": 2.86944270507201e-05, "loss": 0.8601, "step": 8580, "task_loss": 1.4585493803024292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7883668541908264, "epoch": 7.25, "learning_rate": 2.8691296180338136e-05, "loss": 0.7706, "step": 8581, "task_loss": 0.6570387482643127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0099949836730957, "epoch": 7.25, "learning_rate": 2.8688165309956168e-05, "loss": 0.7893, "step": 8582, "task_loss": 1.5854765176773071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4164952337741852, "epoch": 7.26, "learning_rate": 2.8685034439574203e-05, "loss": 0.6529, "step": 8583, "task_loss": 0.3964788317680359 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0894207954406738, "epoch": 7.26, "learning_rate": 2.8681903569192238e-05, "loss": 0.7001, "step": 8584, "task_loss": 1.3682140111923218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21114501357078552, "epoch": 7.26, "learning_rate": 2.867877269881027e-05, "loss": 0.4646, "step": 8585, "task_loss": 0.1588110625743866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7123158574104309, "epoch": 7.26, "learning_rate": 2.8675641828428308e-05, "loss": 0.7487, "step": 8586, "task_loss": 2.1601977348327637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41852647066116333, "epoch": 7.26, "learning_rate": 2.867251095804634e-05, "loss": 0.5308, "step": 8587, "task_loss": 0.12753839790821075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5859882831573486, "epoch": 7.26, "learning_rate": 2.8669380087664372e-05, "loss": 0.4887, "step": 8588, "task_loss": 0.6195494532585144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1636831760406494, "epoch": 7.26, "learning_rate": 2.8666249217282404e-05, "loss": 0.5949, "step": 8589, "task_loss": 1.6840858459472656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9895098209381104, "epoch": 7.26, "learning_rate": 2.8663118346900442e-05, "loss": 0.7433, "step": 8590, "task_loss": 0.7413873672485352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3906145691871643, "epoch": 7.26, "learning_rate": 2.8659987476518474e-05, "loss": 0.5211, "step": 8591, "task_loss": 0.5584813356399536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7019374370574951, "epoch": 7.26, "learning_rate": 2.8656856606136506e-05, "loss": 0.7515, "step": 8592, "task_loss": 0.4632842540740967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4681006073951721, "epoch": 7.26, "learning_rate": 2.8653725735754537e-05, "loss": 0.4514, "step": 8593, "task_loss": 0.4330950379371643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6910254955291748, "epoch": 7.26, "learning_rate": 2.8650594865372576e-05, "loss": 0.6996, "step": 8594, "task_loss": 1.2528157234191895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49568241834640503, "epoch": 7.27, "learning_rate": 2.8647463994990608e-05, "loss": 0.6496, "step": 8595, "task_loss": 0.4488910138607025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8397867679595947, "epoch": 7.27, "learning_rate": 2.864433312460864e-05, "loss": 0.7036, "step": 8596, "task_loss": 0.7110047936439514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7903435230255127, "epoch": 7.27, "learning_rate": 2.8641202254226678e-05, "loss": 0.9848, "step": 8597, "task_loss": 0.7439081072807312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37117475271224976, "epoch": 7.27, "learning_rate": 2.863807138384471e-05, "loss": 0.5159, "step": 8598, "task_loss": 0.16953514516353607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6604053974151611, "epoch": 7.27, "learning_rate": 2.8634940513462745e-05, "loss": 0.6401, "step": 8599, "task_loss": 0.9172648787498474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6670671105384827, "epoch": 7.27, "learning_rate": 2.8631809643080777e-05, "loss": 0.718, "step": 8600, "task_loss": 1.6450773477554321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5137817859649658, "epoch": 7.27, "learning_rate": 2.8628678772698815e-05, "loss": 0.6417, "step": 8601, "task_loss": 0.18667013943195343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5208088755607605, "epoch": 7.27, "learning_rate": 2.8625547902316847e-05, "loss": 0.613, "step": 8602, "task_loss": 0.2738474905490875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9315210580825806, "epoch": 7.27, "learning_rate": 2.862241703193488e-05, "loss": 0.7548, "step": 8603, "task_loss": 1.2455719709396362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7936480641365051, "epoch": 7.27, "learning_rate": 2.861928616155291e-05, "loss": 0.5376, "step": 8604, "task_loss": 0.6613881587982178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.019195318222046, "epoch": 7.27, "learning_rate": 2.861615529117095e-05, "loss": 0.7874, "step": 8605, "task_loss": 1.7265456914901733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0540739297866821, "epoch": 7.27, "learning_rate": 2.861302442078898e-05, "loss": 0.7681, "step": 8606, "task_loss": 1.22652006149292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3233312964439392, "epoch": 7.28, "learning_rate": 2.8609893550407013e-05, "loss": 0.5011, "step": 8607, "task_loss": 0.5957701802253723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7260701656341553, "epoch": 7.28, "learning_rate": 2.860676268002505e-05, "loss": 0.7732, "step": 8608, "task_loss": 1.1861320734024048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35099080204963684, "epoch": 7.28, "learning_rate": 2.8603631809643083e-05, "loss": 0.4767, "step": 8609, "task_loss": 0.46697306632995605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6344455480575562, "epoch": 7.28, "learning_rate": 2.8600500939261115e-05, "loss": 0.622, "step": 8610, "task_loss": 0.5536139011383057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4957481622695923, "epoch": 7.28, "learning_rate": 2.8597370068879147e-05, "loss": 0.5493, "step": 8611, "task_loss": 0.6833500862121582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4445546269416809, "epoch": 7.28, "learning_rate": 2.8594239198497185e-05, "loss": 0.644, "step": 8612, "task_loss": 0.2237693965435028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.679714560508728, "epoch": 7.28, "learning_rate": 2.8591108328115217e-05, "loss": 0.678, "step": 8613, "task_loss": 0.7906001806259155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7219349145889282, "epoch": 7.28, "learning_rate": 2.858797745773325e-05, "loss": 0.6821, "step": 8614, "task_loss": 0.99933260679245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9113666415214539, "epoch": 7.28, "learning_rate": 2.8584846587351284e-05, "loss": 0.8287, "step": 8615, "task_loss": 0.40912845730781555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6997302770614624, "epoch": 7.28, "learning_rate": 2.858171571696932e-05, "loss": 0.595, "step": 8616, "task_loss": 1.802162766456604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5504604578018188, "epoch": 7.28, "learning_rate": 2.8578584846587354e-05, "loss": 0.6563, "step": 8617, "task_loss": 0.4735666513442993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.853141188621521, "epoch": 7.28, "learning_rate": 2.8575453976205386e-05, "loss": 0.7537, "step": 8618, "task_loss": 0.9317866563796997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2762336730957031, "epoch": 7.29, "learning_rate": 2.8572323105823418e-05, "loss": 0.6504, "step": 8619, "task_loss": 0.24718131124973297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5327799916267395, "epoch": 7.29, "learning_rate": 2.8569192235441457e-05, "loss": 0.5895, "step": 8620, "task_loss": 1.659619688987732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5108404159545898, "epoch": 7.29, "learning_rate": 2.8566061365059488e-05, "loss": 0.6467, "step": 8621, "task_loss": 0.19818004965782166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7444647550582886, "epoch": 7.29, "learning_rate": 2.856293049467752e-05, "loss": 0.6393, "step": 8622, "task_loss": 0.8534947633743286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7461051940917969, "epoch": 7.29, "learning_rate": 2.855979962429556e-05, "loss": 0.7311, "step": 8623, "task_loss": 1.230838418006897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.911177396774292, "epoch": 7.29, "learning_rate": 2.855666875391359e-05, "loss": 0.6868, "step": 8624, "task_loss": 0.5235562324523926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7193633317947388, "epoch": 7.29, "learning_rate": 2.8553537883531622e-05, "loss": 0.8315, "step": 8625, "task_loss": 0.30210885405540466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5662848949432373, "epoch": 7.29, "learning_rate": 2.8550407013149654e-05, "loss": 0.6089, "step": 8626, "task_loss": 0.6963916420936584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0736348628997803, "epoch": 7.29, "learning_rate": 2.8547276142767693e-05, "loss": 0.8686, "step": 8627, "task_loss": 1.0708379745483398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.702079176902771, "epoch": 7.29, "learning_rate": 2.8544145272385724e-05, "loss": 0.6407, "step": 8628, "task_loss": 0.5556609034538269 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0152673721313477, "epoch": 7.29, "learning_rate": 2.8541014402003756e-05, "loss": 0.8141, "step": 8629, "task_loss": 1.732312560081482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9794276356697083, "epoch": 7.29, "learning_rate": 2.853788353162179e-05, "loss": 0.7798, "step": 8630, "task_loss": 0.5193533897399902 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49087685346603394, "epoch": 7.3, "learning_rate": 2.8534752661239826e-05, "loss": 0.5915, "step": 8631, "task_loss": 0.36664682626724243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.522433876991272, "epoch": 7.3, "learning_rate": 2.8531621790857858e-05, "loss": 0.5683, "step": 8632, "task_loss": 0.2835423946380615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4012669324874878, "epoch": 7.3, "learning_rate": 2.8528490920475893e-05, "loss": 0.4802, "step": 8633, "task_loss": 0.44125574827194214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8514754176139832, "epoch": 7.3, "learning_rate": 2.852536005009393e-05, "loss": 0.7541, "step": 8634, "task_loss": 0.6810735464096069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.520588219165802, "epoch": 7.3, "learning_rate": 2.8522229179711964e-05, "loss": 0.6276, "step": 8635, "task_loss": 1.0220717191696167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5086417198181152, "epoch": 7.3, "learning_rate": 2.8519098309329996e-05, "loss": 0.6674, "step": 8636, "task_loss": 1.0254924297332764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5787479877471924, "epoch": 7.3, "learning_rate": 2.8515967438948027e-05, "loss": 0.7554, "step": 8637, "task_loss": 0.49894484877586365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8624910116195679, "epoch": 7.3, "learning_rate": 2.8512836568566066e-05, "loss": 0.5454, "step": 8638, "task_loss": 0.48530498147010803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8402595520019531, "epoch": 7.3, "learning_rate": 2.8509705698184098e-05, "loss": 0.7231, "step": 8639, "task_loss": 0.3541012704372406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7706143856048584, "epoch": 7.3, "learning_rate": 2.850657482780213e-05, "loss": 0.9454, "step": 8640, "task_loss": 0.8278295397758484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5422072410583496, "epoch": 7.3, "learning_rate": 2.850344395742016e-05, "loss": 0.5554, "step": 8641, "task_loss": 0.3702258765697479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40319639444351196, "epoch": 7.3, "learning_rate": 2.85003130870382e-05, "loss": 0.4579, "step": 8642, "task_loss": 0.49643954634666443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7089477777481079, "epoch": 7.31, "learning_rate": 2.849718221665623e-05, "loss": 0.5219, "step": 8643, "task_loss": 0.40868648886680603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5833612680435181, "epoch": 7.31, "learning_rate": 2.8494051346274263e-05, "loss": 0.717, "step": 8644, "task_loss": 0.22646091878414154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7593744993209839, "epoch": 7.31, "learning_rate": 2.8490920475892302e-05, "loss": 0.7358, "step": 8645, "task_loss": 0.43097084760665894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9237546920776367, "epoch": 7.31, "learning_rate": 2.8487789605510334e-05, "loss": 0.7706, "step": 8646, "task_loss": 1.6945515871047974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46302247047424316, "epoch": 7.31, "learning_rate": 2.8484658735128365e-05, "loss": 0.6656, "step": 8647, "task_loss": 0.8735542297363281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3166213035583496, "epoch": 7.31, "learning_rate": 2.84815278647464e-05, "loss": 0.4841, "step": 8648, "task_loss": 0.47853073477745056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7373340129852295, "epoch": 7.31, "learning_rate": 2.8478396994364436e-05, "loss": 0.7174, "step": 8649, "task_loss": 0.9698644876480103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6602089405059814, "epoch": 7.31, "learning_rate": 2.8475266123982468e-05, "loss": 0.6708, "step": 8650, "task_loss": 0.4596869647502899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.196053147315979, "epoch": 7.31, "learning_rate": 2.8472135253600503e-05, "loss": 0.9111, "step": 8651, "task_loss": 1.2404510974884033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5022478103637695, "epoch": 7.31, "learning_rate": 2.8469004383218534e-05, "loss": 0.6609, "step": 8652, "task_loss": 0.5830695629119873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5460759401321411, "epoch": 7.31, "learning_rate": 2.8465873512836573e-05, "loss": 0.6349, "step": 8653, "task_loss": 1.0080026388168335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.813552975654602, "epoch": 7.32, "learning_rate": 2.8462742642454605e-05, "loss": 0.5614, "step": 8654, "task_loss": 0.9982833862304688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6778571605682373, "epoch": 7.32, "learning_rate": 2.8459611772072637e-05, "loss": 0.5248, "step": 8655, "task_loss": 0.8319441080093384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.615939199924469, "epoch": 7.32, "learning_rate": 2.845648090169067e-05, "loss": 0.6282, "step": 8656, "task_loss": 0.7912004590034485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8702715635299683, "epoch": 7.32, "learning_rate": 2.8453350031308707e-05, "loss": 0.8366, "step": 8657, "task_loss": 0.2056104838848114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4774315357208252, "epoch": 7.32, "learning_rate": 2.845021916092674e-05, "loss": 0.6032, "step": 8658, "task_loss": 0.9941402673721313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8213778734207153, "epoch": 7.32, "learning_rate": 2.844708829054477e-05, "loss": 0.9074, "step": 8659, "task_loss": 0.7557512521743774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6155924201011658, "epoch": 7.32, "learning_rate": 2.844395742016281e-05, "loss": 0.7847, "step": 8660, "task_loss": 0.7359055876731873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7528572082519531, "epoch": 7.32, "learning_rate": 2.844082654978084e-05, "loss": 0.657, "step": 8661, "task_loss": 1.0408761501312256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9101608991622925, "epoch": 7.32, "learning_rate": 2.8437695679398873e-05, "loss": 0.7137, "step": 8662, "task_loss": 0.5490474700927734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36038684844970703, "epoch": 7.32, "learning_rate": 2.8434564809016904e-05, "loss": 0.618, "step": 8663, "task_loss": 0.8551596403121948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5387612581253052, "epoch": 7.32, "learning_rate": 2.8431433938634943e-05, "loss": 0.8086, "step": 8664, "task_loss": 0.5462741255760193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5528877973556519, "epoch": 7.32, "learning_rate": 2.8428303068252975e-05, "loss": 0.6805, "step": 8665, "task_loss": 1.2509362697601318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6959391832351685, "epoch": 7.33, "learning_rate": 2.842517219787101e-05, "loss": 0.6468, "step": 8666, "task_loss": 0.40695253014564514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8267828822135925, "epoch": 7.33, "learning_rate": 2.8422041327489042e-05, "loss": 0.7827, "step": 8667, "task_loss": 0.5403761863708496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7104264497756958, "epoch": 7.33, "learning_rate": 2.8418910457107077e-05, "loss": 0.8054, "step": 8668, "task_loss": 1.1819618940353394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25265926122665405, "epoch": 7.33, "learning_rate": 2.8415779586725112e-05, "loss": 0.6607, "step": 8669, "task_loss": 0.3533862829208374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6659880876541138, "epoch": 7.33, "learning_rate": 2.8412648716343144e-05, "loss": 0.5993, "step": 8670, "task_loss": 1.1192617416381836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7255817651748657, "epoch": 7.33, "learning_rate": 2.8409517845961182e-05, "loss": 1.0335, "step": 8671, "task_loss": 1.5877655744552612 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46622607111930847, "epoch": 7.33, "learning_rate": 2.8406386975579214e-05, "loss": 0.5733, "step": 8672, "task_loss": 0.16803529858589172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5867037773132324, "epoch": 7.33, "learning_rate": 2.8403256105197246e-05, "loss": 0.7274, "step": 8673, "task_loss": 0.7516364455223083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8302671909332275, "epoch": 7.33, "learning_rate": 2.8400125234815278e-05, "loss": 0.6773, "step": 8674, "task_loss": 0.6512947678565979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4477478563785553, "epoch": 7.33, "learning_rate": 2.8396994364433316e-05, "loss": 0.5545, "step": 8675, "task_loss": 1.0099594593048096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7254599928855896, "epoch": 7.33, "learning_rate": 2.8393863494051348e-05, "loss": 0.663, "step": 8676, "task_loss": 1.0637034177780151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6115039587020874, "epoch": 7.33, "learning_rate": 2.839073262366938e-05, "loss": 0.7306, "step": 8677, "task_loss": 0.477742999792099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7099313735961914, "epoch": 7.34, "learning_rate": 2.838760175328741e-05, "loss": 0.6945, "step": 8678, "task_loss": 0.8870000243186951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6966025829315186, "epoch": 7.34, "learning_rate": 2.838447088290545e-05, "loss": 0.7117, "step": 8679, "task_loss": 1.34494149684906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5724956393241882, "epoch": 7.34, "learning_rate": 2.8381340012523482e-05, "loss": 0.6406, "step": 8680, "task_loss": 0.3406558334827423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5493896007537842, "epoch": 7.34, "learning_rate": 2.8378209142141514e-05, "loss": 0.767, "step": 8681, "task_loss": 0.9012842178344727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6395965814590454, "epoch": 7.34, "learning_rate": 2.8375078271759552e-05, "loss": 0.733, "step": 8682, "task_loss": 1.1161426305770874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1632899045944214, "epoch": 7.34, "learning_rate": 2.8371947401377584e-05, "loss": 0.785, "step": 8683, "task_loss": 0.9169334769248962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5504445433616638, "epoch": 7.34, "learning_rate": 2.836881653099562e-05, "loss": 0.5163, "step": 8684, "task_loss": 0.7636939287185669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8667901754379272, "epoch": 7.34, "learning_rate": 2.836568566061365e-05, "loss": 0.8055, "step": 8685, "task_loss": 0.6262723803520203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2702341675758362, "epoch": 7.34, "learning_rate": 2.836255479023169e-05, "loss": 0.5833, "step": 8686, "task_loss": 0.359015554189682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6818777918815613, "epoch": 7.34, "learning_rate": 2.835942391984972e-05, "loss": 0.7734, "step": 8687, "task_loss": 0.6865376234054565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.652172327041626, "epoch": 7.34, "learning_rate": 2.8356293049467753e-05, "loss": 0.7082, "step": 8688, "task_loss": 0.36470121145248413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6683883666992188, "epoch": 7.34, "learning_rate": 2.8353162179085785e-05, "loss": 0.749, "step": 8689, "task_loss": 0.41065987944602966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8586405515670776, "epoch": 7.35, "learning_rate": 2.8350031308703824e-05, "loss": 0.6724, "step": 8690, "task_loss": 0.5837189555168152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5282970666885376, "epoch": 7.35, "learning_rate": 2.8346900438321855e-05, "loss": 0.845, "step": 8691, "task_loss": 0.6772649884223938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5409913063049316, "epoch": 7.35, "learning_rate": 2.8343769567939887e-05, "loss": 0.6949, "step": 8692, "task_loss": 0.2651866376399994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5407718420028687, "epoch": 7.35, "learning_rate": 2.8340638697557926e-05, "loss": 0.5473, "step": 8693, "task_loss": 0.14122101664543152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6948367953300476, "epoch": 7.35, "learning_rate": 2.8337507827175957e-05, "loss": 0.763, "step": 8694, "task_loss": 1.181789517402649 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5234971642494202, "epoch": 7.35, "learning_rate": 2.833437695679399e-05, "loss": 0.7147, "step": 8695, "task_loss": 0.5053929686546326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6904911994934082, "epoch": 7.35, "learning_rate": 2.833124608641202e-05, "loss": 0.757, "step": 8696, "task_loss": 0.3554937541484833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.546612024307251, "epoch": 7.35, "learning_rate": 2.832811521603006e-05, "loss": 0.5715, "step": 8697, "task_loss": 0.8234674334526062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4455993175506592, "epoch": 7.35, "learning_rate": 2.832498434564809e-05, "loss": 0.7385, "step": 8698, "task_loss": 0.8174681067466736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9225709438323975, "epoch": 7.35, "learning_rate": 2.8321853475266123e-05, "loss": 1.0329, "step": 8699, "task_loss": 1.0808393955230713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8496356010437012, "epoch": 7.35, "learning_rate": 2.8318722604884158e-05, "loss": 0.7539, "step": 8700, "task_loss": 0.5998347997665405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7472780346870422, "epoch": 7.35, "learning_rate": 2.8315591734502193e-05, "loss": 0.6013, "step": 8701, "task_loss": 2.901012659072876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8117891550064087, "epoch": 7.36, "learning_rate": 2.831246086412023e-05, "loss": 0.6527, "step": 8702, "task_loss": 0.7235847115516663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4216265082359314, "epoch": 7.36, "learning_rate": 2.830932999373826e-05, "loss": 0.6022, "step": 8703, "task_loss": 0.1849815547466278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6287831664085388, "epoch": 7.36, "learning_rate": 2.8306199123356292e-05, "loss": 0.744, "step": 8704, "task_loss": 0.8400972485542297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8073284029960632, "epoch": 7.36, "learning_rate": 2.830306825297433e-05, "loss": 0.6374, "step": 8705, "task_loss": 0.9699385166168213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8281658887863159, "epoch": 7.36, "learning_rate": 2.8299937382592362e-05, "loss": 0.8206, "step": 8706, "task_loss": 0.36759576201438904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.649684488773346, "epoch": 7.36, "learning_rate": 2.8296806512210394e-05, "loss": 0.5669, "step": 8707, "task_loss": 0.6057938933372498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8331478834152222, "epoch": 7.36, "learning_rate": 2.8293675641828433e-05, "loss": 0.6389, "step": 8708, "task_loss": 0.7456074357032776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6102719306945801, "epoch": 7.36, "learning_rate": 2.8290544771446465e-05, "loss": 0.6837, "step": 8709, "task_loss": 0.7579582333564758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.669179379940033, "epoch": 7.36, "learning_rate": 2.8287413901064496e-05, "loss": 0.7715, "step": 8710, "task_loss": 0.5618059635162354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6799174547195435, "epoch": 7.36, "learning_rate": 2.8284283030682528e-05, "loss": 0.6388, "step": 8711, "task_loss": 0.48855751752853394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9695017337799072, "epoch": 7.36, "learning_rate": 2.8281152160300567e-05, "loss": 0.8974, "step": 8712, "task_loss": 0.6907163858413696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.008392095565796, "epoch": 7.36, "learning_rate": 2.82780212899186e-05, "loss": 0.7208, "step": 8713, "task_loss": 0.46296849846839905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6042251586914062, "epoch": 7.37, "learning_rate": 2.827489041953663e-05, "loss": 0.5907, "step": 8714, "task_loss": 0.21405747532844543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4050444960594177, "epoch": 7.37, "learning_rate": 2.8271759549154665e-05, "loss": 0.6828, "step": 8715, "task_loss": 0.7362436056137085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7017986178398132, "epoch": 7.37, "learning_rate": 2.82686286787727e-05, "loss": 0.5543, "step": 8716, "task_loss": 0.5097084045410156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.036901593208313, "epoch": 7.37, "learning_rate": 2.8265497808390732e-05, "loss": 0.7115, "step": 8717, "task_loss": 0.2339293658733368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0199949741363525, "epoch": 7.37, "learning_rate": 2.8262366938008768e-05, "loss": 0.8205, "step": 8718, "task_loss": 0.3011426329612732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6754268407821655, "epoch": 7.37, "learning_rate": 2.8259236067626803e-05, "loss": 0.6217, "step": 8719, "task_loss": 0.16464506089687347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.019136667251587, "epoch": 7.37, "learning_rate": 2.8256105197244838e-05, "loss": 0.8991, "step": 8720, "task_loss": 0.622413694858551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5463671088218689, "epoch": 7.37, "learning_rate": 2.825297432686287e-05, "loss": 0.6298, "step": 8721, "task_loss": 0.2082112580537796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4698333144187927, "epoch": 7.37, "learning_rate": 2.82498434564809e-05, "loss": 0.7249, "step": 8722, "task_loss": 0.6398137807846069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6278135180473328, "epoch": 7.37, "learning_rate": 2.824671258609894e-05, "loss": 0.6993, "step": 8723, "task_loss": 0.4139755368232727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3523227572441101, "epoch": 7.37, "learning_rate": 2.8243581715716972e-05, "loss": 0.6578, "step": 8724, "task_loss": 0.5073292255401611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6609482765197754, "epoch": 7.38, "learning_rate": 2.8240450845335004e-05, "loss": 0.6858, "step": 8725, "task_loss": 0.4792857766151428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6024028062820435, "epoch": 7.38, "learning_rate": 2.8237319974953035e-05, "loss": 0.6626, "step": 8726, "task_loss": 0.3469833433628082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7680933475494385, "epoch": 7.38, "learning_rate": 2.8234189104571074e-05, "loss": 0.645, "step": 8727, "task_loss": 0.2925221920013428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7418922185897827, "epoch": 7.38, "learning_rate": 2.8231058234189106e-05, "loss": 0.8073, "step": 8728, "task_loss": 1.2363967895507812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.913530707359314, "epoch": 7.38, "learning_rate": 2.8227927363807138e-05, "loss": 0.6555, "step": 8729, "task_loss": 1.100874900817871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35989701747894287, "epoch": 7.38, "learning_rate": 2.8224796493425176e-05, "loss": 0.7538, "step": 8730, "task_loss": 0.733830988407135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43583977222442627, "epoch": 7.38, "learning_rate": 2.8221665623043208e-05, "loss": 0.6859, "step": 8731, "task_loss": 1.107914924621582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5965237617492676, "epoch": 7.38, "learning_rate": 2.821853475266124e-05, "loss": 0.5953, "step": 8732, "task_loss": 0.5497075319290161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0495258569717407, "epoch": 7.38, "learning_rate": 2.8215403882279275e-05, "loss": 0.7212, "step": 8733, "task_loss": 1.3129991292953491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6533822417259216, "epoch": 7.38, "learning_rate": 2.821227301189731e-05, "loss": 0.5702, "step": 8734, "task_loss": 0.4768957495689392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5164860486984253, "epoch": 7.38, "learning_rate": 2.8209142141515342e-05, "loss": 0.6167, "step": 8735, "task_loss": 1.7156665325164795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40979883074760437, "epoch": 7.38, "learning_rate": 2.8206011271133377e-05, "loss": 0.7757, "step": 8736, "task_loss": 0.10094466060400009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2154157161712646, "epoch": 7.39, "learning_rate": 2.820288040075141e-05, "loss": 0.804, "step": 8737, "task_loss": 0.8376672863960266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5440839529037476, "epoch": 7.39, "learning_rate": 2.8199749530369447e-05, "loss": 0.5903, "step": 8738, "task_loss": 0.7369547486305237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5277293920516968, "epoch": 7.39, "learning_rate": 2.819661865998748e-05, "loss": 0.6524, "step": 8739, "task_loss": 0.4011024236679077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33726656436920166, "epoch": 7.39, "learning_rate": 2.819348778960551e-05, "loss": 0.523, "step": 8740, "task_loss": 0.3628687262535095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8878815174102783, "epoch": 7.39, "learning_rate": 2.8190356919223543e-05, "loss": 0.818, "step": 8741, "task_loss": 0.30085140466690063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8479537963867188, "epoch": 7.39, "learning_rate": 2.818722604884158e-05, "loss": 0.576, "step": 8742, "task_loss": 1.0505235195159912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5981613993644714, "epoch": 7.39, "learning_rate": 2.8184095178459613e-05, "loss": 0.7304, "step": 8743, "task_loss": 1.3986998796463013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5609970092773438, "epoch": 7.39, "learning_rate": 2.8180964308077645e-05, "loss": 0.6479, "step": 8744, "task_loss": 0.8494614362716675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7561981678009033, "epoch": 7.39, "learning_rate": 2.8177833437695683e-05, "loss": 0.9993, "step": 8745, "task_loss": 0.9566130042076111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7072064280509949, "epoch": 7.39, "learning_rate": 2.8174702567313715e-05, "loss": 0.6553, "step": 8746, "task_loss": 0.38347485661506653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5993257164955139, "epoch": 7.39, "learning_rate": 2.8171571696931747e-05, "loss": 0.5311, "step": 8747, "task_loss": 0.6305227875709534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4937189519405365, "epoch": 7.39, "learning_rate": 2.816844082654978e-05, "loss": 0.6643, "step": 8748, "task_loss": 1.2561657428741455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8303518891334534, "epoch": 7.4, "learning_rate": 2.8165309956167817e-05, "loss": 0.7158, "step": 8749, "task_loss": 1.220591425895691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40555283427238464, "epoch": 7.4, "learning_rate": 2.816217908578585e-05, "loss": 0.4857, "step": 8750, "task_loss": 0.3062984049320221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5500526428222656, "epoch": 7.4, "learning_rate": 2.8159048215403884e-05, "loss": 0.5282, "step": 8751, "task_loss": 0.3149394989013672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4951242208480835, "epoch": 7.4, "learning_rate": 2.8155917345021916e-05, "loss": 0.6992, "step": 8752, "task_loss": 0.43961429595947266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.403460294008255, "epoch": 7.4, "learning_rate": 2.8152786474639954e-05, "loss": 0.6582, "step": 8753, "task_loss": 0.3072225749492645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0304667949676514, "epoch": 7.4, "learning_rate": 2.8149655604257986e-05, "loss": 0.6719, "step": 8754, "task_loss": 0.9349584579467773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5876983404159546, "epoch": 7.4, "learning_rate": 2.8146524733876018e-05, "loss": 0.5313, "step": 8755, "task_loss": 0.5104329586029053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7532891035079956, "epoch": 7.4, "learning_rate": 2.8143393863494057e-05, "loss": 0.7602, "step": 8756, "task_loss": 0.6858499050140381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5554585456848145, "epoch": 7.4, "learning_rate": 2.814026299311209e-05, "loss": 0.6068, "step": 8757, "task_loss": 0.5668289661407471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8183461427688599, "epoch": 7.4, "learning_rate": 2.813713212273012e-05, "loss": 0.8234, "step": 8758, "task_loss": 0.9885321259498596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9959775805473328, "epoch": 7.4, "learning_rate": 2.8134001252348152e-05, "loss": 0.8982, "step": 8759, "task_loss": 2.30863094329834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7859461307525635, "epoch": 7.4, "learning_rate": 2.813087038196619e-05, "loss": 0.6838, "step": 8760, "task_loss": 1.5350921154022217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6158106327056885, "epoch": 7.41, "learning_rate": 2.8127739511584222e-05, "loss": 0.6528, "step": 8761, "task_loss": 0.8665949702262878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7590534687042236, "epoch": 7.41, "learning_rate": 2.8124608641202254e-05, "loss": 0.7804, "step": 8762, "task_loss": 0.7687414288520813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9065198302268982, "epoch": 7.41, "learning_rate": 2.8121477770820286e-05, "loss": 0.6938, "step": 8763, "task_loss": 1.2923884391784668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7673044204711914, "epoch": 7.41, "learning_rate": 2.8118346900438324e-05, "loss": 0.8704, "step": 8764, "task_loss": 0.6381109952926636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5836144685745239, "epoch": 7.41, "learning_rate": 2.8115216030056356e-05, "loss": 0.8061, "step": 8765, "task_loss": 0.6048976182937622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7787831425666809, "epoch": 7.41, "learning_rate": 2.8112085159674388e-05, "loss": 0.713, "step": 8766, "task_loss": 1.054921269416809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.181098222732544, "epoch": 7.41, "learning_rate": 2.8108954289292427e-05, "loss": 0.7747, "step": 8767, "task_loss": 1.2563267946243286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43193119764328003, "epoch": 7.41, "learning_rate": 2.8105823418910458e-05, "loss": 0.6092, "step": 8768, "task_loss": 0.5129494071006775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3594672679901123, "epoch": 7.41, "learning_rate": 2.8102692548528493e-05, "loss": 0.5606, "step": 8769, "task_loss": 0.32786133885383606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4575199484825134, "epoch": 7.41, "learning_rate": 2.8099561678146525e-05, "loss": 0.7541, "step": 8770, "task_loss": 0.3065391480922699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43765515089035034, "epoch": 7.41, "learning_rate": 2.8096430807764564e-05, "loss": 0.4801, "step": 8771, "task_loss": 0.5472772717475891 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5160094499588013, "epoch": 7.41, "learning_rate": 2.8093299937382596e-05, "loss": 0.6125, "step": 8772, "task_loss": 1.5261750221252441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5440532565116882, "epoch": 7.42, "learning_rate": 2.8090169067000627e-05, "loss": 0.5259, "step": 8773, "task_loss": 0.4931740164756775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42767882347106934, "epoch": 7.42, "learning_rate": 2.808703819661866e-05, "loss": 0.7162, "step": 8774, "task_loss": 0.30984702706336975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8581374883651733, "epoch": 7.42, "learning_rate": 2.8083907326236698e-05, "loss": 0.8329, "step": 8775, "task_loss": 0.2951958179473877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6098191738128662, "epoch": 7.42, "learning_rate": 2.808077645585473e-05, "loss": 0.5665, "step": 8776, "task_loss": 1.0712279081344604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45149874687194824, "epoch": 7.42, "learning_rate": 2.807764558547276e-05, "loss": 0.6898, "step": 8777, "task_loss": 1.0855058431625366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7814539074897766, "epoch": 7.42, "learning_rate": 2.8074514715090793e-05, "loss": 0.7366, "step": 8778, "task_loss": 1.1221083402633667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3750661611557007, "epoch": 7.42, "learning_rate": 2.807138384470883e-05, "loss": 0.5371, "step": 8779, "task_loss": 0.23768866062164307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.812508761882782, "epoch": 7.42, "learning_rate": 2.8068252974326863e-05, "loss": 0.7138, "step": 8780, "task_loss": 0.5685204863548279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.631499707698822, "epoch": 7.42, "learning_rate": 2.8065122103944895e-05, "loss": 0.7139, "step": 8781, "task_loss": 0.4607557952404022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9398317933082581, "epoch": 7.42, "learning_rate": 2.8061991233562934e-05, "loss": 0.7167, "step": 8782, "task_loss": 0.7672886252403259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4100750982761383, "epoch": 7.42, "learning_rate": 2.8058860363180966e-05, "loss": 0.6413, "step": 8783, "task_loss": 0.3345680832862854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6734197735786438, "epoch": 7.42, "learning_rate": 2.8055729492798997e-05, "loss": 0.5644, "step": 8784, "task_loss": 0.5919389724731445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5595947504043579, "epoch": 7.43, "learning_rate": 2.8052598622417032e-05, "loss": 0.6964, "step": 8785, "task_loss": 0.7438595294952393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4685544967651367, "epoch": 7.43, "learning_rate": 2.8049467752035068e-05, "loss": 0.6219, "step": 8786, "task_loss": 0.6929187178611755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.703748881816864, "epoch": 7.43, "learning_rate": 2.8046336881653103e-05, "loss": 0.755, "step": 8787, "task_loss": 0.9299463033676147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5516213178634644, "epoch": 7.43, "learning_rate": 2.8043206011271135e-05, "loss": 0.6818, "step": 8788, "task_loss": 0.44678303599357605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7688393592834473, "epoch": 7.43, "learning_rate": 2.8040075140889166e-05, "loss": 0.6139, "step": 8789, "task_loss": 0.9909530878067017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5975785255432129, "epoch": 7.43, "learning_rate": 2.8036944270507205e-05, "loss": 0.6589, "step": 8790, "task_loss": 0.4738157391548157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4342321753501892, "epoch": 7.43, "learning_rate": 2.8033813400125237e-05, "loss": 0.6378, "step": 8791, "task_loss": 0.25811901688575745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4748522937297821, "epoch": 7.43, "learning_rate": 2.803068252974327e-05, "loss": 0.4793, "step": 8792, "task_loss": 0.8404300808906555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48181551694869995, "epoch": 7.43, "learning_rate": 2.8027551659361307e-05, "loss": 0.5645, "step": 8793, "task_loss": 0.16662679612636566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45101988315582275, "epoch": 7.43, "learning_rate": 2.802442078897934e-05, "loss": 0.7006, "step": 8794, "task_loss": 0.6702026724815369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8319796919822693, "epoch": 7.43, "learning_rate": 2.802128991859737e-05, "loss": 0.7109, "step": 8795, "task_loss": 1.1450470685958862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.407101571559906, "epoch": 7.44, "learning_rate": 2.8018159048215402e-05, "loss": 0.6234, "step": 8796, "task_loss": 0.23016445338726044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.120642066001892, "epoch": 7.44, "learning_rate": 2.801502817783344e-05, "loss": 0.7221, "step": 8797, "task_loss": 0.4159664511680603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4295078217983246, "epoch": 7.44, "learning_rate": 2.8011897307451473e-05, "loss": 0.8079, "step": 8798, "task_loss": 0.0964450091123581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6340898871421814, "epoch": 7.44, "learning_rate": 2.8008766437069504e-05, "loss": 0.5964, "step": 8799, "task_loss": 0.9807495474815369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4612695574760437, "epoch": 7.44, "learning_rate": 2.800563556668754e-05, "loss": 0.6225, "step": 8800, "task_loss": 0.3714570701122284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4207022190093994, "epoch": 7.44, "learning_rate": 2.8002504696305575e-05, "loss": 1.0454, "step": 8801, "task_loss": 1.7001583576202393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5477580428123474, "epoch": 7.44, "learning_rate": 2.7999373825923607e-05, "loss": 0.8119, "step": 8802, "task_loss": 1.2847486734390259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42922502756118774, "epoch": 7.44, "learning_rate": 2.7996242955541642e-05, "loss": 0.4855, "step": 8803, "task_loss": 0.5238210558891296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.860409140586853, "epoch": 7.44, "learning_rate": 2.7993112085159677e-05, "loss": 0.5753, "step": 8804, "task_loss": 0.3088688552379608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.333641916513443, "epoch": 7.44, "learning_rate": 2.7989981214777712e-05, "loss": 0.5121, "step": 8805, "task_loss": 0.3712344467639923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4421110153198242, "epoch": 7.44, "learning_rate": 2.7986850344395744e-05, "loss": 0.6154, "step": 8806, "task_loss": 0.6552814841270447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1284995079040527, "epoch": 7.44, "learning_rate": 2.7983719474013776e-05, "loss": 0.7507, "step": 8807, "task_loss": 0.987210750579834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6407496929168701, "epoch": 7.45, "learning_rate": 2.7980588603631814e-05, "loss": 0.7273, "step": 8808, "task_loss": 1.2387347221374512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6481969356536865, "epoch": 7.45, "learning_rate": 2.7977457733249846e-05, "loss": 0.5724, "step": 8809, "task_loss": 0.8668972253799438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4183013439178467, "epoch": 7.45, "learning_rate": 2.7974326862867878e-05, "loss": 0.6274, "step": 8810, "task_loss": 0.20484508574008942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7269802093505859, "epoch": 7.45, "learning_rate": 2.797119599248591e-05, "loss": 0.6353, "step": 8811, "task_loss": 0.5995336174964905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8299833536148071, "epoch": 7.45, "learning_rate": 2.7968065122103948e-05, "loss": 0.7038, "step": 8812, "task_loss": 0.989422082901001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8846255540847778, "epoch": 7.45, "learning_rate": 2.796493425172198e-05, "loss": 0.5529, "step": 8813, "task_loss": 1.2069478034973145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0477914810180664, "epoch": 7.45, "learning_rate": 2.7961803381340012e-05, "loss": 0.7911, "step": 8814, "task_loss": 1.316024899482727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3210374116897583, "epoch": 7.45, "learning_rate": 2.7958672510958043e-05, "loss": 0.6042, "step": 8815, "task_loss": 0.6727946996688843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35982444882392883, "epoch": 7.45, "learning_rate": 2.7955541640576082e-05, "loss": 0.6111, "step": 8816, "task_loss": 0.794288158416748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6954720616340637, "epoch": 7.45, "learning_rate": 2.7952410770194114e-05, "loss": 0.7388, "step": 8817, "task_loss": 1.392090082168579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6814044117927551, "epoch": 7.45, "learning_rate": 2.794927989981215e-05, "loss": 0.6326, "step": 8818, "task_loss": 0.9974490404129028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.586689829826355, "epoch": 7.45, "learning_rate": 2.7946149029430184e-05, "loss": 0.8002, "step": 8819, "task_loss": 0.581546425819397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4403482675552368, "epoch": 7.46, "learning_rate": 2.7943018159048216e-05, "loss": 0.6368, "step": 8820, "task_loss": 0.29727253317832947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4192628264427185, "epoch": 7.46, "learning_rate": 2.793988728866625e-05, "loss": 0.7166, "step": 8821, "task_loss": 0.6607955694198608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1285805702209473, "epoch": 7.46, "learning_rate": 2.7936756418284283e-05, "loss": 0.7103, "step": 8822, "task_loss": 0.5228863954544067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7411041259765625, "epoch": 7.46, "learning_rate": 2.793362554790232e-05, "loss": 0.5581, "step": 8823, "task_loss": 0.13978183269500732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6713605523109436, "epoch": 7.46, "learning_rate": 2.7930494677520353e-05, "loss": 0.7107, "step": 8824, "task_loss": 1.295412540435791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5203004479408264, "epoch": 7.46, "learning_rate": 2.7927363807138385e-05, "loss": 0.6783, "step": 8825, "task_loss": 1.0964449644088745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5880818963050842, "epoch": 7.46, "learning_rate": 2.7924232936756417e-05, "loss": 0.6202, "step": 8826, "task_loss": 0.5840260982513428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0902953147888184, "epoch": 7.46, "learning_rate": 2.7921102066374455e-05, "loss": 0.6872, "step": 8827, "task_loss": 1.0698812007904053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.681069016456604, "epoch": 7.46, "learning_rate": 2.7917971195992487e-05, "loss": 0.5614, "step": 8828, "task_loss": 0.14443495869636536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1090134382247925, "epoch": 7.46, "learning_rate": 2.791484032561052e-05, "loss": 0.768, "step": 8829, "task_loss": 0.5864757299423218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7818779945373535, "epoch": 7.46, "learning_rate": 2.7911709455228557e-05, "loss": 0.7638, "step": 8830, "task_loss": 1.5806022882461548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7092574238777161, "epoch": 7.46, "learning_rate": 2.790857858484659e-05, "loss": 0.7283, "step": 8831, "task_loss": 0.6833672523498535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15776681900024414, "epoch": 7.47, "learning_rate": 2.790544771446462e-05, "loss": 0.5907, "step": 8832, "task_loss": 0.17077116668224335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7315012812614441, "epoch": 7.47, "learning_rate": 2.7902316844082653e-05, "loss": 0.5811, "step": 8833, "task_loss": 1.1257193088531494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6757863759994507, "epoch": 7.47, "learning_rate": 2.789918597370069e-05, "loss": 0.7313, "step": 8834, "task_loss": 0.7668935656547546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3643103837966919, "epoch": 7.47, "learning_rate": 2.7896055103318723e-05, "loss": 0.6305, "step": 8835, "task_loss": 0.4031587839126587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5588780641555786, "epoch": 7.47, "learning_rate": 2.789292423293676e-05, "loss": 0.5787, "step": 8836, "task_loss": 0.7323093414306641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2874821424484253, "epoch": 7.47, "learning_rate": 2.788979336255479e-05, "loss": 0.9615, "step": 8837, "task_loss": 0.5754157900810242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1571381092071533, "epoch": 7.47, "learning_rate": 2.788666249217283e-05, "loss": 0.782, "step": 8838, "task_loss": 1.0394586324691772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48742619156837463, "epoch": 7.47, "learning_rate": 2.788353162179086e-05, "loss": 0.4822, "step": 8839, "task_loss": 0.6790181994438171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5946010947227478, "epoch": 7.47, "learning_rate": 2.7880400751408892e-05, "loss": 0.4061, "step": 8840, "task_loss": 0.5514503121376038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6389089822769165, "epoch": 7.47, "learning_rate": 2.787726988102693e-05, "loss": 0.6007, "step": 8841, "task_loss": 0.8563059568405151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4459320902824402, "epoch": 7.47, "learning_rate": 2.7874139010644963e-05, "loss": 0.5663, "step": 8842, "task_loss": 1.2268357276916504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9017670154571533, "epoch": 7.47, "learning_rate": 2.7871008140262994e-05, "loss": 0.8433, "step": 8843, "task_loss": 1.6468980312347412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7712960243225098, "epoch": 7.48, "learning_rate": 2.7867877269881026e-05, "loss": 0.9062, "step": 8844, "task_loss": 1.8324816226959229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8349848985671997, "epoch": 7.48, "learning_rate": 2.7864746399499065e-05, "loss": 0.6907, "step": 8845, "task_loss": 0.7089641094207764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3186090588569641, "epoch": 7.48, "learning_rate": 2.7861615529117096e-05, "loss": 0.8044, "step": 8846, "task_loss": 0.7154178023338318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6857525110244751, "epoch": 7.48, "learning_rate": 2.7858484658735128e-05, "loss": 0.8201, "step": 8847, "task_loss": 1.1292932033538818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9675085544586182, "epoch": 7.48, "learning_rate": 2.785535378835316e-05, "loss": 0.7992, "step": 8848, "task_loss": 1.4347161054611206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5075402855873108, "epoch": 7.48, "learning_rate": 2.78522229179712e-05, "loss": 0.8955, "step": 8849, "task_loss": 0.605917751789093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7644408941268921, "epoch": 7.48, "learning_rate": 2.784909204758923e-05, "loss": 0.6442, "step": 8850, "task_loss": 0.6165237426757812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6717935800552368, "epoch": 7.48, "learning_rate": 2.7845961177207262e-05, "loss": 0.7148, "step": 8851, "task_loss": 1.019392728805542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4922497868537903, "epoch": 7.48, "learning_rate": 2.7842830306825297e-05, "loss": 0.5166, "step": 8852, "task_loss": 0.90252685546875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8218318819999695, "epoch": 7.48, "learning_rate": 2.7839699436443332e-05, "loss": 0.5942, "step": 8853, "task_loss": 0.6900655627250671 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25402048230171204, "epoch": 7.48, "learning_rate": 2.7836568566061368e-05, "loss": 0.6946, "step": 8854, "task_loss": 0.2919767200946808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8974696397781372, "epoch": 7.48, "learning_rate": 2.78334376956794e-05, "loss": 0.8387, "step": 8855, "task_loss": 1.2610074281692505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0227079391479492, "epoch": 7.49, "learning_rate": 2.7830306825297438e-05, "loss": 0.671, "step": 8856, "task_loss": 0.9339841604232788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.470958411693573, "epoch": 7.49, "learning_rate": 2.782717595491547e-05, "loss": 0.4925, "step": 8857, "task_loss": 0.414824515581131 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.495168000459671, "epoch": 7.49, "learning_rate": 2.78240450845335e-05, "loss": 0.6279, "step": 8858, "task_loss": 0.5650363564491272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4172421097755432, "epoch": 7.49, "learning_rate": 2.7820914214151533e-05, "loss": 0.6447, "step": 8859, "task_loss": 0.19549964368343353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6954474449157715, "epoch": 7.49, "learning_rate": 2.7817783343769572e-05, "loss": 0.8002, "step": 8860, "task_loss": 0.8795031309127808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7621569633483887, "epoch": 7.49, "learning_rate": 2.7814652473387604e-05, "loss": 0.563, "step": 8861, "task_loss": 0.6552727222442627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8916828632354736, "epoch": 7.49, "learning_rate": 2.7811521603005635e-05, "loss": 0.8194, "step": 8862, "task_loss": 1.1523269414901733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4542318880558014, "epoch": 7.49, "learning_rate": 2.7808390732623667e-05, "loss": 0.6688, "step": 8863, "task_loss": 0.9505837559700012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.153666615486145, "epoch": 7.49, "learning_rate": 2.7805259862241706e-05, "loss": 0.6819, "step": 8864, "task_loss": 0.6711534857749939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4368741810321808, "epoch": 7.49, "learning_rate": 2.7802128991859738e-05, "loss": 0.6763, "step": 8865, "task_loss": 0.1741897016763687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4314463138580322, "epoch": 7.49, "learning_rate": 2.779899812147777e-05, "loss": 0.6362, "step": 8866, "task_loss": 0.31251904368400574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5735291838645935, "epoch": 7.5, "learning_rate": 2.7795867251095808e-05, "loss": 0.6009, "step": 8867, "task_loss": 0.35348787903785706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6740708351135254, "epoch": 7.5, "learning_rate": 2.779273638071384e-05, "loss": 0.5799, "step": 8868, "task_loss": 1.2136934995651245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5912533402442932, "epoch": 7.5, "learning_rate": 2.778960551033187e-05, "loss": 0.6616, "step": 8869, "task_loss": 0.3323783576488495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.586917519569397, "epoch": 7.5, "learning_rate": 2.7786474639949907e-05, "loss": 0.6413, "step": 8870, "task_loss": 0.5449810028076172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3835616409778595, "epoch": 7.5, "learning_rate": 2.7783343769567942e-05, "loss": 0.585, "step": 8871, "task_loss": 0.29183581471443176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36324742436408997, "epoch": 7.5, "learning_rate": 2.7780212899185977e-05, "loss": 0.5342, "step": 8872, "task_loss": 0.3628849685192108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4239441454410553, "epoch": 7.5, "learning_rate": 2.777708202880401e-05, "loss": 0.5365, "step": 8873, "task_loss": 0.79250568151474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6779139637947083, "epoch": 7.5, "learning_rate": 2.777395115842204e-05, "loss": 0.8201, "step": 8874, "task_loss": 1.3842679262161255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41660404205322266, "epoch": 7.5, "learning_rate": 2.777082028804008e-05, "loss": 0.5753, "step": 8875, "task_loss": 0.34002232551574707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40680819749832153, "epoch": 7.5, "learning_rate": 2.776768941765811e-05, "loss": 0.6343, "step": 8876, "task_loss": 0.22283755242824554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8476375341415405, "epoch": 7.5, "learning_rate": 2.7764558547276143e-05, "loss": 0.7047, "step": 8877, "task_loss": 0.6939153671264648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7010583877563477, "epoch": 7.5, "learning_rate": 2.776142767689418e-05, "loss": 0.6307, "step": 8878, "task_loss": 0.27121445536613464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4452442228794098, "epoch": 7.51, "learning_rate": 2.7758296806512213e-05, "loss": 0.5459, "step": 8879, "task_loss": 0.2766035497188568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5221630334854126, "epoch": 7.51, "learning_rate": 2.7755165936130245e-05, "loss": 0.7134, "step": 8880, "task_loss": 0.7477262020111084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7782728672027588, "epoch": 7.51, "learning_rate": 2.7752035065748277e-05, "loss": 0.6077, "step": 8881, "task_loss": 0.9199982285499573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5955924391746521, "epoch": 7.51, "learning_rate": 2.7748904195366315e-05, "loss": 0.9854, "step": 8882, "task_loss": 0.23325540125370026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6426391005516052, "epoch": 7.51, "learning_rate": 2.7745773324984347e-05, "loss": 0.6154, "step": 8883, "task_loss": 0.7717761397361755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9633383750915527, "epoch": 7.51, "learning_rate": 2.774264245460238e-05, "loss": 0.8353, "step": 8884, "task_loss": 1.3567430973052979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7771105766296387, "epoch": 7.51, "learning_rate": 2.7739511584220414e-05, "loss": 0.5856, "step": 8885, "task_loss": 0.8314599394798279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46642422676086426, "epoch": 7.51, "learning_rate": 2.773638071383845e-05, "loss": 0.5241, "step": 8886, "task_loss": 1.2365704774856567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6254347562789917, "epoch": 7.51, "learning_rate": 2.773324984345648e-05, "loss": 0.732, "step": 8887, "task_loss": 0.24277329444885254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9156547784805298, "epoch": 7.51, "learning_rate": 2.7730118973074516e-05, "loss": 0.6307, "step": 8888, "task_loss": 0.958961009979248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4069744050502777, "epoch": 7.51, "learning_rate": 2.7726988102692548e-05, "loss": 0.575, "step": 8889, "task_loss": 0.23684550821781158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7926462888717651, "epoch": 7.51, "learning_rate": 2.7723857232310586e-05, "loss": 0.6381, "step": 8890, "task_loss": 0.6162194609642029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28588706254959106, "epoch": 7.52, "learning_rate": 2.7720726361928618e-05, "loss": 0.5794, "step": 8891, "task_loss": 0.17895661294460297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4703579843044281, "epoch": 7.52, "learning_rate": 2.771759549154665e-05, "loss": 0.5624, "step": 8892, "task_loss": 0.1737464964389801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6404953002929688, "epoch": 7.52, "learning_rate": 2.771446462116469e-05, "loss": 0.6538, "step": 8893, "task_loss": 0.9414440393447876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5610682368278503, "epoch": 7.52, "learning_rate": 2.771133375078272e-05, "loss": 0.5766, "step": 8894, "task_loss": 0.639207124710083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0036556720733643, "epoch": 7.52, "learning_rate": 2.7708202880400752e-05, "loss": 0.8815, "step": 8895, "task_loss": 1.0366026163101196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8443719744682312, "epoch": 7.52, "learning_rate": 2.7705072010018784e-05, "loss": 0.6364, "step": 8896, "task_loss": 0.8146626949310303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7869648933410645, "epoch": 7.52, "learning_rate": 2.7701941139636822e-05, "loss": 0.6134, "step": 8897, "task_loss": 0.881245493888855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6698192954063416, "epoch": 7.52, "learning_rate": 2.7698810269254854e-05, "loss": 0.5076, "step": 8898, "task_loss": 1.0186127424240112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8638697862625122, "epoch": 7.52, "learning_rate": 2.7695679398872886e-05, "loss": 0.7399, "step": 8899, "task_loss": 1.0903170108795166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6105559468269348, "epoch": 7.52, "learning_rate": 2.7692548528490918e-05, "loss": 0.5778, "step": 8900, "task_loss": 1.2111175060272217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6490325927734375, "epoch": 7.52, "learning_rate": 2.7689417658108956e-05, "loss": 0.7742, "step": 8901, "task_loss": 0.5405369400978088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7843350172042847, "epoch": 7.52, "learning_rate": 2.7686286787726988e-05, "loss": 0.5162, "step": 8902, "task_loss": 0.9883390665054321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5447616577148438, "epoch": 7.53, "learning_rate": 2.7683155917345023e-05, "loss": 0.7361, "step": 8903, "task_loss": 0.9327481985092163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3944104313850403, "epoch": 7.53, "learning_rate": 2.768002504696306e-05, "loss": 0.6119, "step": 8904, "task_loss": 0.4954860806465149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.599523663520813, "epoch": 7.53, "learning_rate": 2.7676894176581094e-05, "loss": 0.6979, "step": 8905, "task_loss": 0.4132433533668518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5537995100021362, "epoch": 7.53, "learning_rate": 2.7673763306199125e-05, "loss": 0.6551, "step": 8906, "task_loss": 0.3478572368621826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5768522024154663, "epoch": 7.53, "learning_rate": 2.7670632435817157e-05, "loss": 0.6606, "step": 8907, "task_loss": 0.8405965566635132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2960614562034607, "epoch": 7.53, "learning_rate": 2.7667501565435196e-05, "loss": 0.6128, "step": 8908, "task_loss": 0.1926761120557785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.536874532699585, "epoch": 7.53, "learning_rate": 2.7664370695053227e-05, "loss": 0.5525, "step": 8909, "task_loss": 0.43825411796569824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39426273107528687, "epoch": 7.53, "learning_rate": 2.766123982467126e-05, "loss": 0.5423, "step": 8910, "task_loss": 0.16923609375953674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4464220106601715, "epoch": 7.53, "learning_rate": 2.765810895428929e-05, "loss": 0.4005, "step": 8911, "task_loss": 0.6438937783241272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.061429500579834, "epoch": 7.53, "learning_rate": 2.765497808390733e-05, "loss": 0.7042, "step": 8912, "task_loss": 0.8864091634750366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44453394412994385, "epoch": 7.53, "learning_rate": 2.765184721352536e-05, "loss": 0.5922, "step": 8913, "task_loss": 1.446808099746704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6103577017784119, "epoch": 7.53, "learning_rate": 2.7648716343143393e-05, "loss": 0.7321, "step": 8914, "task_loss": 0.3099899888038635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34666088223457336, "epoch": 7.54, "learning_rate": 2.764558547276143e-05, "loss": 0.6508, "step": 8915, "task_loss": 1.0368236303329468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46396172046661377, "epoch": 7.54, "learning_rate": 2.7642454602379463e-05, "loss": 0.4667, "step": 8916, "task_loss": 0.5881316661834717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.954359769821167, "epoch": 7.54, "learning_rate": 2.7639323731997495e-05, "loss": 0.6388, "step": 8917, "task_loss": 1.0911415815353394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5896576642990112, "epoch": 7.54, "learning_rate": 2.7636192861615527e-05, "loss": 0.5984, "step": 8918, "task_loss": 0.7172211408615112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5846594572067261, "epoch": 7.54, "learning_rate": 2.7633061991233566e-05, "loss": 0.8266, "step": 8919, "task_loss": 0.7047592997550964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1020228862762451, "epoch": 7.54, "learning_rate": 2.7629931120851597e-05, "loss": 0.7547, "step": 8920, "task_loss": 1.8042447566986084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7113831639289856, "epoch": 7.54, "learning_rate": 2.7626800250469633e-05, "loss": 0.8926, "step": 8921, "task_loss": 1.2243740558624268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47615402936935425, "epoch": 7.54, "learning_rate": 2.7623669380087664e-05, "loss": 0.473, "step": 8922, "task_loss": 0.7355588674545288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1924290657043457, "epoch": 7.54, "learning_rate": 2.7620538509705703e-05, "loss": 0.8897, "step": 8923, "task_loss": 0.9614308476448059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3856412172317505, "epoch": 7.54, "learning_rate": 2.7617407639323735e-05, "loss": 0.6245, "step": 8924, "task_loss": 0.5958232283592224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9148974418640137, "epoch": 7.54, "learning_rate": 2.7614276768941766e-05, "loss": 0.6517, "step": 8925, "task_loss": 1.2139389514923096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.698992908000946, "epoch": 7.54, "learning_rate": 2.7611145898559805e-05, "loss": 0.6519, "step": 8926, "task_loss": 0.8846461772918701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5656126737594604, "epoch": 7.55, "learning_rate": 2.7608015028177837e-05, "loss": 0.799, "step": 8927, "task_loss": 0.5656106472015381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4970540404319763, "epoch": 7.55, "learning_rate": 2.760488415779587e-05, "loss": 0.6197, "step": 8928, "task_loss": 0.7635831236839294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8086221814155579, "epoch": 7.55, "learning_rate": 2.76017532874139e-05, "loss": 0.7734, "step": 8929, "task_loss": 0.5769930481910706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.916208028793335, "epoch": 7.55, "learning_rate": 2.759862241703194e-05, "loss": 0.8572, "step": 8930, "task_loss": 1.2751905918121338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38610202074050903, "epoch": 7.55, "learning_rate": 2.759549154664997e-05, "loss": 0.5383, "step": 8931, "task_loss": 0.4414200186729431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.657019317150116, "epoch": 7.55, "learning_rate": 2.7592360676268002e-05, "loss": 0.737, "step": 8932, "task_loss": 0.668395459651947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42420363426208496, "epoch": 7.55, "learning_rate": 2.7589229805886034e-05, "loss": 0.6021, "step": 8933, "task_loss": 0.6997520327568054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7705491185188293, "epoch": 7.55, "learning_rate": 2.7586098935504073e-05, "loss": 0.6844, "step": 8934, "task_loss": 0.9108431935310364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8366838693618774, "epoch": 7.55, "learning_rate": 2.7582968065122105e-05, "loss": 0.625, "step": 8935, "task_loss": 1.1577550172805786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43314749002456665, "epoch": 7.55, "learning_rate": 2.7579837194740136e-05, "loss": 0.6589, "step": 8936, "task_loss": 0.2802475392818451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7549093961715698, "epoch": 7.55, "learning_rate": 2.757670632435817e-05, "loss": 0.5563, "step": 8937, "task_loss": 0.49648770689964294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5094294548034668, "epoch": 7.56, "learning_rate": 2.7573575453976207e-05, "loss": 0.5716, "step": 8938, "task_loss": 0.2903459668159485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6225179433822632, "epoch": 7.56, "learning_rate": 2.7570444583594242e-05, "loss": 0.7641, "step": 8939, "task_loss": 0.6109243035316467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5209974050521851, "epoch": 7.56, "learning_rate": 2.7567313713212274e-05, "loss": 0.7357, "step": 8940, "task_loss": 0.5212772488594055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38129663467407227, "epoch": 7.56, "learning_rate": 2.7564182842830312e-05, "loss": 0.5624, "step": 8941, "task_loss": 0.5728554129600525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42950281500816345, "epoch": 7.56, "learning_rate": 2.7561051972448344e-05, "loss": 0.4966, "step": 8942, "task_loss": 0.7538378238677979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8192622065544128, "epoch": 7.56, "learning_rate": 2.7557921102066376e-05, "loss": 0.6487, "step": 8943, "task_loss": 0.8644980788230896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5837259292602539, "epoch": 7.56, "learning_rate": 2.7554790231684408e-05, "loss": 0.5473, "step": 8944, "task_loss": 1.2237054109573364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5449843406677246, "epoch": 7.56, "learning_rate": 2.7551659361302446e-05, "loss": 0.6685, "step": 8945, "task_loss": 0.44731566309928894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7824638485908508, "epoch": 7.56, "learning_rate": 2.7548528490920478e-05, "loss": 0.6102, "step": 8946, "task_loss": 1.8488942384719849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7510493397712708, "epoch": 7.56, "learning_rate": 2.754539762053851e-05, "loss": 0.6628, "step": 8947, "task_loss": 0.41483360528945923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6366939544677734, "epoch": 7.56, "learning_rate": 2.754226675015654e-05, "loss": 0.6394, "step": 8948, "task_loss": 0.6321890354156494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.464306116104126, "epoch": 7.56, "learning_rate": 2.753913587977458e-05, "loss": 0.621, "step": 8949, "task_loss": 0.2168804258108139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6221418380737305, "epoch": 7.57, "learning_rate": 2.7536005009392612e-05, "loss": 0.6778, "step": 8950, "task_loss": 1.2634588479995728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6802655458450317, "epoch": 7.57, "learning_rate": 2.7532874139010644e-05, "loss": 0.7092, "step": 8951, "task_loss": 2.0564656257629395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5001466274261475, "epoch": 7.57, "learning_rate": 2.7529743268628682e-05, "loss": 0.5886, "step": 8952, "task_loss": 1.0723568201065063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0155131816864014, "epoch": 7.57, "learning_rate": 2.7526612398246714e-05, "loss": 0.7448, "step": 8953, "task_loss": 0.8294299840927124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6465198397636414, "epoch": 7.57, "learning_rate": 2.7523481527864746e-05, "loss": 0.6206, "step": 8954, "task_loss": 1.1105976104736328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7446368932723999, "epoch": 7.57, "learning_rate": 2.752035065748278e-05, "loss": 0.7179, "step": 8955, "task_loss": 1.3301059007644653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.803210973739624, "epoch": 7.57, "learning_rate": 2.7517219787100816e-05, "loss": 0.6762, "step": 8956, "task_loss": 1.0994250774383545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8462374210357666, "epoch": 7.57, "learning_rate": 2.751408891671885e-05, "loss": 0.6686, "step": 8957, "task_loss": 0.5027087926864624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31246569752693176, "epoch": 7.57, "learning_rate": 2.7510958046336883e-05, "loss": 0.5762, "step": 8958, "task_loss": 0.41692429780960083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43392086029052734, "epoch": 7.57, "learning_rate": 2.7507827175954915e-05, "loss": 0.5562, "step": 8959, "task_loss": 0.336427241563797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.72489333152771, "epoch": 7.57, "learning_rate": 2.7504696305572953e-05, "loss": 0.6262, "step": 8960, "task_loss": 0.7338352203369141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6216086745262146, "epoch": 7.57, "learning_rate": 2.7501565435190985e-05, "loss": 0.6962, "step": 8961, "task_loss": 0.4543043375015259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3157344162464142, "epoch": 7.58, "learning_rate": 2.7498434564809017e-05, "loss": 0.5164, "step": 8962, "task_loss": 0.28988200426101685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.715294361114502, "epoch": 7.58, "learning_rate": 2.7495303694427055e-05, "loss": 0.6867, "step": 8963, "task_loss": 0.9106261730194092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5449313521385193, "epoch": 7.58, "learning_rate": 2.7492172824045087e-05, "loss": 0.5301, "step": 8964, "task_loss": 0.19430138170719147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6488290429115295, "epoch": 7.58, "learning_rate": 2.748904195366312e-05, "loss": 0.6518, "step": 8965, "task_loss": 0.5688123106956482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8042439222335815, "epoch": 7.58, "learning_rate": 2.748591108328115e-05, "loss": 0.552, "step": 8966, "task_loss": 0.8395355939865112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6000596880912781, "epoch": 7.58, "learning_rate": 2.748278021289919e-05, "loss": 0.7041, "step": 8967, "task_loss": 0.5005080103874207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3057855367660522, "epoch": 7.58, "learning_rate": 2.747964934251722e-05, "loss": 0.9428, "step": 8968, "task_loss": 0.9319005608558655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5928449034690857, "epoch": 7.58, "learning_rate": 2.7476518472135253e-05, "loss": 0.8239, "step": 8969, "task_loss": 0.29400935769081116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4799942672252655, "epoch": 7.58, "learning_rate": 2.7473387601753288e-05, "loss": 0.7154, "step": 8970, "task_loss": 0.5471099019050598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17628273367881775, "epoch": 7.58, "learning_rate": 2.7470256731371323e-05, "loss": 0.5824, "step": 8971, "task_loss": 0.008879422210156918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7649878263473511, "epoch": 7.58, "learning_rate": 2.7467125860989355e-05, "loss": 0.775, "step": 8972, "task_loss": 0.9271711111068726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6579602360725403, "epoch": 7.58, "learning_rate": 2.746399499060739e-05, "loss": 0.5471, "step": 8973, "task_loss": 0.7659314274787903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4490913152694702, "epoch": 7.59, "learning_rate": 2.7460864120225422e-05, "loss": 0.5808, "step": 8974, "task_loss": 0.1274532973766327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.230480670928955, "epoch": 7.59, "learning_rate": 2.745773324984346e-05, "loss": 0.9306, "step": 8975, "task_loss": 0.6082167625427246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6733161211013794, "epoch": 7.59, "learning_rate": 2.7454602379461492e-05, "loss": 0.7211, "step": 8976, "task_loss": 0.5211902856826782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30328530073165894, "epoch": 7.59, "learning_rate": 2.7451471509079524e-05, "loss": 0.5687, "step": 8977, "task_loss": 0.39067089557647705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6034406423568726, "epoch": 7.59, "learning_rate": 2.7448340638697563e-05, "loss": 0.6091, "step": 8978, "task_loss": 0.7363659739494324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5380334854125977, "epoch": 7.59, "learning_rate": 2.7445209768315594e-05, "loss": 0.8331, "step": 8979, "task_loss": 0.7293322086334229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44974029064178467, "epoch": 7.59, "learning_rate": 2.7442078897933626e-05, "loss": 0.5531, "step": 8980, "task_loss": 0.24387188255786896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6355788111686707, "epoch": 7.59, "learning_rate": 2.7438948027551658e-05, "loss": 0.4686, "step": 8981, "task_loss": 0.7747161388397217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6032196879386902, "epoch": 7.59, "learning_rate": 2.7435817157169697e-05, "loss": 0.7654, "step": 8982, "task_loss": 0.4569885730743408 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8820475339889526, "epoch": 7.59, "learning_rate": 2.743268628678773e-05, "loss": 0.6582, "step": 8983, "task_loss": 0.5206196308135986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7819731831550598, "epoch": 7.59, "learning_rate": 2.742955541640576e-05, "loss": 0.7067, "step": 8984, "task_loss": 1.0394830703735352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.430209755897522, "epoch": 7.59, "learning_rate": 2.7426424546023792e-05, "loss": 0.5098, "step": 8985, "task_loss": 0.3782522678375244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9605779647827148, "epoch": 7.6, "learning_rate": 2.742329367564183e-05, "loss": 0.7014, "step": 8986, "task_loss": 0.5686838030815125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3502236604690552, "epoch": 7.6, "learning_rate": 2.7420162805259862e-05, "loss": 0.5812, "step": 8987, "task_loss": 0.3403714895248413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4587635397911072, "epoch": 7.6, "learning_rate": 2.7417031934877897e-05, "loss": 0.7804, "step": 8988, "task_loss": 0.6978436708450317 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6119934916496277, "epoch": 7.6, "learning_rate": 2.7413901064495933e-05, "loss": 0.6332, "step": 8989, "task_loss": 0.6507314443588257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6518468856811523, "epoch": 7.6, "learning_rate": 2.7410770194113968e-05, "loss": 0.6538, "step": 8990, "task_loss": 0.4380953311920166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4558527171611786, "epoch": 7.6, "learning_rate": 2.7407639323732e-05, "loss": 0.702, "step": 8991, "task_loss": 0.9231404066085815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.183954119682312, "epoch": 7.6, "learning_rate": 2.740450845335003e-05, "loss": 0.5953, "step": 8992, "task_loss": 0.35410624742507935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5388160943984985, "epoch": 7.6, "learning_rate": 2.740137758296807e-05, "loss": 0.652, "step": 8993, "task_loss": 0.7035973072052002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7143998146057129, "epoch": 7.6, "learning_rate": 2.73982467125861e-05, "loss": 0.6642, "step": 8994, "task_loss": 0.40524184703826904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9364047050476074, "epoch": 7.6, "learning_rate": 2.7395115842204133e-05, "loss": 0.6933, "step": 8995, "task_loss": 1.063331961631775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28107309341430664, "epoch": 7.6, "learning_rate": 2.7391984971822165e-05, "loss": 0.5263, "step": 8996, "task_loss": 0.5530312061309814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8153818249702454, "epoch": 7.6, "learning_rate": 2.7388854101440204e-05, "loss": 0.857, "step": 8997, "task_loss": 0.8958744406700134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.660413920879364, "epoch": 7.61, "learning_rate": 2.7385723231058236e-05, "loss": 0.5712, "step": 8998, "task_loss": 0.4465912878513336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40993309020996094, "epoch": 7.61, "learning_rate": 2.7382592360676267e-05, "loss": 0.5613, "step": 8999, "task_loss": 0.4795423746109009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3447607159614563, "epoch": 7.61, "learning_rate": 2.7379461490294306e-05, "loss": 0.5352, "step": 9000, "task_loss": 0.32322460412979126 }, { "epoch": 7.61, "eval_accuracy": 0.8956039603960396, "eval_loss": 0.43599167466163635, "eval_runtime": 207.1873, "eval_samples_per_second": 121.87, "eval_steps_per_second": 0.956, "step": 9000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7202310562133789, "epoch": 7.61, "learning_rate": 2.7376330619912338e-05, "loss": 0.615, "step": 9001, "task_loss": 0.7756620049476624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8583766222000122, "epoch": 7.61, "learning_rate": 2.737319974953037e-05, "loss": 0.5859, "step": 9002, "task_loss": 0.43935754895210266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6091982126235962, "epoch": 7.61, "learning_rate": 2.73700688791484e-05, "loss": 0.7449, "step": 9003, "task_loss": 1.0431554317474365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6967465281486511, "epoch": 7.61, "learning_rate": 2.736693800876644e-05, "loss": 0.6679, "step": 9004, "task_loss": 0.6833162307739258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4311971664428711, "epoch": 7.61, "learning_rate": 2.736380713838447e-05, "loss": 0.5284, "step": 9005, "task_loss": 0.08479667454957962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0950758457183838, "epoch": 7.61, "learning_rate": 2.7360676268002507e-05, "loss": 0.7658, "step": 9006, "task_loss": 0.7293604016304016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.112318992614746, "epoch": 7.61, "learning_rate": 2.735754539762054e-05, "loss": 0.8914, "step": 9007, "task_loss": 1.2525591850280762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28706562519073486, "epoch": 7.61, "learning_rate": 2.7354414527238577e-05, "loss": 0.4478, "step": 9008, "task_loss": 0.35179102420806885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46857258677482605, "epoch": 7.61, "learning_rate": 2.735128365685661e-05, "loss": 0.6593, "step": 9009, "task_loss": 0.8727076649665833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7051342725753784, "epoch": 7.62, "learning_rate": 2.734815278647464e-05, "loss": 0.6096, "step": 9010, "task_loss": 0.9365283846855164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5259950757026672, "epoch": 7.62, "learning_rate": 2.7345021916092672e-05, "loss": 0.6557, "step": 9011, "task_loss": 1.0656437873840332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1925135850906372, "epoch": 7.62, "learning_rate": 2.734189104571071e-05, "loss": 0.8085, "step": 9012, "task_loss": 1.7250317335128784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8004405498504639, "epoch": 7.62, "learning_rate": 2.7338760175328743e-05, "loss": 0.8325, "step": 9013, "task_loss": 1.2954504489898682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.50799959897995, "epoch": 7.62, "learning_rate": 2.7335629304946775e-05, "loss": 0.6045, "step": 9014, "task_loss": 0.4986036419868469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7376776933670044, "epoch": 7.62, "learning_rate": 2.7332498434564813e-05, "loss": 0.5882, "step": 9015, "task_loss": 1.284864902496338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7638691663742065, "epoch": 7.62, "learning_rate": 2.7329367564182845e-05, "loss": 0.5939, "step": 9016, "task_loss": 0.9938772916793823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5138095617294312, "epoch": 7.62, "learning_rate": 2.7326236693800877e-05, "loss": 0.8285, "step": 9017, "task_loss": 0.15245717763900757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5792665481567383, "epoch": 7.62, "learning_rate": 2.732310582341891e-05, "loss": 0.7462, "step": 9018, "task_loss": 0.4709542691707611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6692178249359131, "epoch": 7.62, "learning_rate": 2.7319974953036947e-05, "loss": 0.6398, "step": 9019, "task_loss": 0.813185453414917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5950918793678284, "epoch": 7.62, "learning_rate": 2.731684408265498e-05, "loss": 0.5813, "step": 9020, "task_loss": 0.5670608878135681 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.70735102891922, "epoch": 7.63, "learning_rate": 2.731371321227301e-05, "loss": 0.528, "step": 9021, "task_loss": 1.0493927001953125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8375440835952759, "epoch": 7.63, "learning_rate": 2.7310582341891046e-05, "loss": 0.7743, "step": 9022, "task_loss": 0.4641715884208679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7559237480163574, "epoch": 7.63, "learning_rate": 2.730745147150908e-05, "loss": 0.6857, "step": 9023, "task_loss": 1.1693031787872314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.518613338470459, "epoch": 7.63, "learning_rate": 2.7304320601127116e-05, "loss": 0.5973, "step": 9024, "task_loss": 1.2777090072631836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8691790103912354, "epoch": 7.63, "learning_rate": 2.7301189730745148e-05, "loss": 0.6319, "step": 9025, "task_loss": 0.9262452125549316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.737280011177063, "epoch": 7.63, "learning_rate": 2.7298058860363186e-05, "loss": 0.7229, "step": 9026, "task_loss": 0.5243698358535767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5139451622962952, "epoch": 7.63, "learning_rate": 2.7294927989981218e-05, "loss": 0.7629, "step": 9027, "task_loss": 0.7549276947975159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5766741633415222, "epoch": 7.63, "learning_rate": 2.729179711959925e-05, "loss": 0.6995, "step": 9028, "task_loss": 0.3982377350330353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6361104249954224, "epoch": 7.63, "learning_rate": 2.7288666249217282e-05, "loss": 0.6994, "step": 9029, "task_loss": 0.7479232549667358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.51141357421875, "epoch": 7.63, "learning_rate": 2.728553537883532e-05, "loss": 0.7137, "step": 9030, "task_loss": 1.2313200235366821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22702017426490784, "epoch": 7.63, "learning_rate": 2.7282404508453352e-05, "loss": 0.5149, "step": 9031, "task_loss": 0.18724285066127777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7985252737998962, "epoch": 7.63, "learning_rate": 2.7279273638071384e-05, "loss": 0.87, "step": 9032, "task_loss": 0.8498522639274597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5648787617683411, "epoch": 7.64, "learning_rate": 2.7276142767689416e-05, "loss": 0.7079, "step": 9033, "task_loss": 0.9290918111801147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3202856779098511, "epoch": 7.64, "learning_rate": 2.7273011897307454e-05, "loss": 0.5974, "step": 9034, "task_loss": 0.09156602621078491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6488555669784546, "epoch": 7.64, "learning_rate": 2.7269881026925486e-05, "loss": 0.7966, "step": 9035, "task_loss": 0.11520276963710785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6469608545303345, "epoch": 7.64, "learning_rate": 2.7266750156543518e-05, "loss": 0.5558, "step": 9036, "task_loss": 0.15248796343803406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5787816047668457, "epoch": 7.64, "learning_rate": 2.7263619286161556e-05, "loss": 0.5489, "step": 9037, "task_loss": 0.3309899866580963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4872792959213257, "epoch": 7.64, "learning_rate": 2.7260488415779588e-05, "loss": 0.6191, "step": 9038, "task_loss": 0.8850634098052979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40178221464157104, "epoch": 7.64, "learning_rate": 2.725735754539762e-05, "loss": 0.8154, "step": 9039, "task_loss": 0.5982359647750854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5226277112960815, "epoch": 7.64, "learning_rate": 2.7254226675015655e-05, "loss": 0.6668, "step": 9040, "task_loss": 0.8941110968589783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5097953081130981, "epoch": 7.64, "learning_rate": 2.725109580463369e-05, "loss": 0.6027, "step": 9041, "task_loss": 0.5206576585769653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.01516854763031, "epoch": 7.64, "learning_rate": 2.7247964934251725e-05, "loss": 0.9081, "step": 9042, "task_loss": 1.1039774417877197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5026754140853882, "epoch": 7.64, "learning_rate": 2.7244834063869757e-05, "loss": 0.5967, "step": 9043, "task_loss": 0.7683252096176147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5753145813941956, "epoch": 7.64, "learning_rate": 2.724170319348779e-05, "loss": 0.541, "step": 9044, "task_loss": 1.6632689237594604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9812200665473938, "epoch": 7.65, "learning_rate": 2.7238572323105828e-05, "loss": 0.8021, "step": 9045, "task_loss": 1.5812714099884033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.654583752155304, "epoch": 7.65, "learning_rate": 2.723544145272386e-05, "loss": 0.7933, "step": 9046, "task_loss": 0.8673695921897888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5649954080581665, "epoch": 7.65, "learning_rate": 2.723231058234189e-05, "loss": 0.665, "step": 9047, "task_loss": 0.6104059815406799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41008418798446655, "epoch": 7.65, "learning_rate": 2.7229179711959923e-05, "loss": 0.4575, "step": 9048, "task_loss": 0.8219320774078369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9152613878250122, "epoch": 7.65, "learning_rate": 2.722604884157796e-05, "loss": 0.698, "step": 9049, "task_loss": 0.7423321008682251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5565897226333618, "epoch": 7.65, "learning_rate": 2.7222917971195993e-05, "loss": 0.6405, "step": 9050, "task_loss": 0.34620094299316406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6551817059516907, "epoch": 7.65, "learning_rate": 2.7219787100814025e-05, "loss": 0.5298, "step": 9051, "task_loss": 0.5248398184776306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5036306977272034, "epoch": 7.65, "learning_rate": 2.7216656230432064e-05, "loss": 0.5331, "step": 9052, "task_loss": 0.31872451305389404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6308820247650146, "epoch": 7.65, "learning_rate": 2.7213525360050095e-05, "loss": 0.5468, "step": 9053, "task_loss": 0.620202898979187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8178375959396362, "epoch": 7.65, "learning_rate": 2.7210394489668127e-05, "loss": 0.6015, "step": 9054, "task_loss": 1.0340542793273926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8176068663597107, "epoch": 7.65, "learning_rate": 2.7207263619286162e-05, "loss": 0.775, "step": 9055, "task_loss": 1.0172107219696045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7397323846817017, "epoch": 7.65, "learning_rate": 2.7204132748904197e-05, "loss": 0.5697, "step": 9056, "task_loss": 1.0757213830947876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.023863434791565, "epoch": 7.66, "learning_rate": 2.7201001878522233e-05, "loss": 0.7438, "step": 9057, "task_loss": 1.097316861152649 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9110146760940552, "epoch": 7.66, "learning_rate": 2.7197871008140264e-05, "loss": 0.6574, "step": 9058, "task_loss": 0.7461894154548645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6675537824630737, "epoch": 7.66, "learning_rate": 2.7194740137758296e-05, "loss": 0.588, "step": 9059, "task_loss": 0.2416151762008667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.656948447227478, "epoch": 7.66, "learning_rate": 2.7191609267376335e-05, "loss": 0.6442, "step": 9060, "task_loss": 0.2007015347480774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6384002566337585, "epoch": 7.66, "learning_rate": 2.7188478396994366e-05, "loss": 0.6343, "step": 9061, "task_loss": 0.49866676330566406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6565188765525818, "epoch": 7.66, "learning_rate": 2.7185347526612398e-05, "loss": 0.613, "step": 9062, "task_loss": 0.6551436185836792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8952928781509399, "epoch": 7.66, "learning_rate": 2.7182216656230437e-05, "loss": 0.9246, "step": 9063, "task_loss": 1.0258947610855103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.794374942779541, "epoch": 7.66, "learning_rate": 2.717908578584847e-05, "loss": 0.7193, "step": 9064, "task_loss": 1.407500147819519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9081317186355591, "epoch": 7.66, "learning_rate": 2.71759549154665e-05, "loss": 0.6813, "step": 9065, "task_loss": 0.6833802461624146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.50393146276474, "epoch": 7.66, "learning_rate": 2.7172824045084532e-05, "loss": 0.6937, "step": 9066, "task_loss": 0.7177032232284546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8315527439117432, "epoch": 7.66, "learning_rate": 2.716969317470257e-05, "loss": 0.6658, "step": 9067, "task_loss": 0.8557356595993042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0852165222167969, "epoch": 7.66, "learning_rate": 2.7166562304320603e-05, "loss": 0.7838, "step": 9068, "task_loss": 2.129610061645508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0870864391326904, "epoch": 7.67, "learning_rate": 2.7163431433938634e-05, "loss": 0.7721, "step": 9069, "task_loss": 0.2841489613056183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9615383148193359, "epoch": 7.67, "learning_rate": 2.7160300563556666e-05, "loss": 0.8935, "step": 9070, "task_loss": 1.1024796962738037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7214419841766357, "epoch": 7.67, "learning_rate": 2.7157169693174705e-05, "loss": 0.7171, "step": 9071, "task_loss": 1.4930222034454346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7507506608963013, "epoch": 7.67, "learning_rate": 2.7154038822792736e-05, "loss": 0.582, "step": 9072, "task_loss": 0.49404898285865784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8709163665771484, "epoch": 7.67, "learning_rate": 2.715090795241077e-05, "loss": 0.7892, "step": 9073, "task_loss": 0.7357127070426941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0297917127609253, "epoch": 7.67, "learning_rate": 2.7147777082028807e-05, "loss": 0.9243, "step": 9074, "task_loss": 0.7619327306747437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3471616804599762, "epoch": 7.67, "learning_rate": 2.7144646211646842e-05, "loss": 0.5738, "step": 9075, "task_loss": 0.7350857257843018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8885467052459717, "epoch": 7.67, "learning_rate": 2.7141515341264874e-05, "loss": 0.9707, "step": 9076, "task_loss": 0.801266610622406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7417207956314087, "epoch": 7.67, "learning_rate": 2.7138384470882905e-05, "loss": 0.9145, "step": 9077, "task_loss": 1.1820377111434937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6719886660575867, "epoch": 7.67, "learning_rate": 2.7135253600500944e-05, "loss": 0.7545, "step": 9078, "task_loss": 1.0895042419433594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0616694688796997, "epoch": 7.67, "learning_rate": 2.7132122730118976e-05, "loss": 0.9436, "step": 9079, "task_loss": 1.4097150564193726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4300075173377991, "epoch": 7.67, "learning_rate": 2.7128991859737008e-05, "loss": 0.5255, "step": 9080, "task_loss": 0.26918861269950867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.974052906036377, "epoch": 7.68, "learning_rate": 2.712586098935504e-05, "loss": 0.6209, "step": 9081, "task_loss": 1.1401796340942383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5571780800819397, "epoch": 7.68, "learning_rate": 2.7122730118973078e-05, "loss": 0.4336, "step": 9082, "task_loss": 0.45195654034614563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6481809616088867, "epoch": 7.68, "learning_rate": 2.711959924859111e-05, "loss": 0.7239, "step": 9083, "task_loss": 1.0444782972335815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6535935401916504, "epoch": 7.68, "learning_rate": 2.711646837820914e-05, "loss": 0.6987, "step": 9084, "task_loss": 1.4963624477386475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5060843825340271, "epoch": 7.68, "learning_rate": 2.7113337507827173e-05, "loss": 0.695, "step": 9085, "task_loss": 0.7987533211708069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.543657660484314, "epoch": 7.68, "learning_rate": 2.7110206637445212e-05, "loss": 0.7021, "step": 9086, "task_loss": 0.9052963256835938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9547255039215088, "epoch": 7.68, "learning_rate": 2.7107075767063244e-05, "loss": 0.9537, "step": 9087, "task_loss": 0.6027315258979797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6986555457115173, "epoch": 7.68, "learning_rate": 2.7103944896681275e-05, "loss": 0.7303, "step": 9088, "task_loss": 0.8668461441993713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.456111341714859, "epoch": 7.68, "learning_rate": 2.7100814026299314e-05, "loss": 0.4549, "step": 9089, "task_loss": 0.2734018862247467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.427778959274292, "epoch": 7.68, "learning_rate": 2.7097683155917346e-05, "loss": 0.5175, "step": 9090, "task_loss": 0.23947983980178833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39282214641571045, "epoch": 7.68, "learning_rate": 2.709455228553538e-05, "loss": 0.5801, "step": 9091, "task_loss": 1.1931464672088623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6622781157493591, "epoch": 7.69, "learning_rate": 2.7091421415153413e-05, "loss": 0.5536, "step": 9092, "task_loss": 1.120289921760559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2418060302734375, "epoch": 7.69, "learning_rate": 2.708829054477145e-05, "loss": 0.936, "step": 9093, "task_loss": 1.758711576461792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6164236068725586, "epoch": 7.69, "learning_rate": 2.7085159674389483e-05, "loss": 0.4964, "step": 9094, "task_loss": 0.6728459596633911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6823778748512268, "epoch": 7.69, "learning_rate": 2.7082028804007515e-05, "loss": 0.8776, "step": 9095, "task_loss": 0.40003132820129395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8517938852310181, "epoch": 7.69, "learning_rate": 2.7078897933625547e-05, "loss": 0.7656, "step": 9096, "task_loss": 0.18181295692920685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9337392449378967, "epoch": 7.69, "learning_rate": 2.7075767063243585e-05, "loss": 0.8494, "step": 9097, "task_loss": 0.9348750114440918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7000215649604797, "epoch": 7.69, "learning_rate": 2.7072636192861617e-05, "loss": 0.786, "step": 9098, "task_loss": 0.839861273765564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7628123760223389, "epoch": 7.69, "learning_rate": 2.706950532247965e-05, "loss": 0.882, "step": 9099, "task_loss": 0.5566021203994751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47661837935447693, "epoch": 7.69, "learning_rate": 2.7066374452097687e-05, "loss": 0.5737, "step": 9100, "task_loss": 0.5094653367996216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37750351428985596, "epoch": 7.69, "learning_rate": 2.706324358171572e-05, "loss": 0.5036, "step": 9101, "task_loss": 0.19961479306221008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5275090932846069, "epoch": 7.69, "learning_rate": 2.706011271133375e-05, "loss": 0.753, "step": 9102, "task_loss": 0.49623599648475647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.320562481880188, "epoch": 7.69, "learning_rate": 2.7056981840951783e-05, "loss": 0.8326, "step": 9103, "task_loss": 1.296225666999817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29508012533187866, "epoch": 7.7, "learning_rate": 2.705385097056982e-05, "loss": 0.6426, "step": 9104, "task_loss": 1.279815912246704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5500090718269348, "epoch": 7.7, "learning_rate": 2.7050720100187853e-05, "loss": 0.5285, "step": 9105, "task_loss": 0.15011996030807495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3957902789115906, "epoch": 7.7, "learning_rate": 2.7047589229805885e-05, "loss": 0.5322, "step": 9106, "task_loss": 0.5886231064796448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7615779638290405, "epoch": 7.7, "learning_rate": 2.704445835942392e-05, "loss": 0.8466, "step": 9107, "task_loss": 0.8037652373313904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7726302146911621, "epoch": 7.7, "learning_rate": 2.7041327489041955e-05, "loss": 0.8143, "step": 9108, "task_loss": 0.9132171869277954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9139465093612671, "epoch": 7.7, "learning_rate": 2.703819661865999e-05, "loss": 0.6785, "step": 9109, "task_loss": 0.6418156027793884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6855244636535645, "epoch": 7.7, "learning_rate": 2.7035065748278022e-05, "loss": 0.7418, "step": 9110, "task_loss": 0.9981102347373962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.432712197303772, "epoch": 7.7, "learning_rate": 2.703193487789606e-05, "loss": 0.6643, "step": 9111, "task_loss": 0.8321713209152222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7189615964889526, "epoch": 7.7, "learning_rate": 2.7028804007514092e-05, "loss": 0.5345, "step": 9112, "task_loss": 0.5252857208251953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39486071467399597, "epoch": 7.7, "learning_rate": 2.7025673137132124e-05, "loss": 0.4118, "step": 9113, "task_loss": 0.6357818841934204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49434134364128113, "epoch": 7.7, "learning_rate": 2.7022542266750156e-05, "loss": 0.6688, "step": 9114, "task_loss": 0.7787076234817505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4368627667427063, "epoch": 7.7, "learning_rate": 2.7019411396368194e-05, "loss": 0.6303, "step": 9115, "task_loss": 1.077149510383606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46482616662979126, "epoch": 7.71, "learning_rate": 2.7016280525986226e-05, "loss": 0.6088, "step": 9116, "task_loss": 0.31154772639274597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8699373602867126, "epoch": 7.71, "learning_rate": 2.7013149655604258e-05, "loss": 0.6991, "step": 9117, "task_loss": 1.074050784111023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4269401431083679, "epoch": 7.71, "learning_rate": 2.701001878522229e-05, "loss": 0.624, "step": 9118, "task_loss": 0.9184966087341309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4728907346725464, "epoch": 7.71, "learning_rate": 2.700688791484033e-05, "loss": 0.6453, "step": 9119, "task_loss": 0.11698748916387558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2241312265396118, "epoch": 7.71, "learning_rate": 2.700375704445836e-05, "loss": 0.8109, "step": 9120, "task_loss": 0.9915633797645569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6558796167373657, "epoch": 7.71, "learning_rate": 2.7000626174076392e-05, "loss": 0.6308, "step": 9121, "task_loss": 1.7308349609375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6180750131607056, "epoch": 7.71, "learning_rate": 2.6997495303694427e-05, "loss": 0.5803, "step": 9122, "task_loss": 0.8774369359016418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5993571281433105, "epoch": 7.71, "learning_rate": 2.6994364433312462e-05, "loss": 0.6737, "step": 9123, "task_loss": 0.9909130334854126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.794301450252533, "epoch": 7.71, "learning_rate": 2.6991233562930494e-05, "loss": 0.666, "step": 9124, "task_loss": 1.6639550924301147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2360008955001831, "epoch": 7.71, "learning_rate": 2.698810269254853e-05, "loss": 0.5302, "step": 9125, "task_loss": 0.024724964052438736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5811573266983032, "epoch": 7.71, "learning_rate": 2.6984971822166564e-05, "loss": 0.5812, "step": 9126, "task_loss": 0.5035795569419861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8593453168869019, "epoch": 7.71, "learning_rate": 2.69818409517846e-05, "loss": 0.6639, "step": 9127, "task_loss": 0.829358696937561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7667109966278076, "epoch": 7.72, "learning_rate": 2.697871008140263e-05, "loss": 0.7946, "step": 9128, "task_loss": 1.2692923545837402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34404122829437256, "epoch": 7.72, "learning_rate": 2.6975579211020663e-05, "loss": 0.5794, "step": 9129, "task_loss": 1.3430746793746948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.517704427242279, "epoch": 7.72, "learning_rate": 2.6972448340638702e-05, "loss": 0.6318, "step": 9130, "task_loss": 0.7864483594894409 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37772148847579956, "epoch": 7.72, "learning_rate": 2.6969317470256733e-05, "loss": 0.7919, "step": 9131, "task_loss": 0.21389469504356384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7094185948371887, "epoch": 7.72, "learning_rate": 2.6966186599874765e-05, "loss": 0.7916, "step": 9132, "task_loss": 0.3332425057888031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18592864274978638, "epoch": 7.72, "learning_rate": 2.6963055729492797e-05, "loss": 0.5506, "step": 9133, "task_loss": 0.015073533169925213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6359863877296448, "epoch": 7.72, "learning_rate": 2.6959924859110836e-05, "loss": 0.9033, "step": 9134, "task_loss": 0.9725310802459717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43079501390457153, "epoch": 7.72, "learning_rate": 2.6956793988728867e-05, "loss": 0.4726, "step": 9135, "task_loss": 0.33227649331092834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7232146859169006, "epoch": 7.72, "learning_rate": 2.69536631183469e-05, "loss": 0.5702, "step": 9136, "task_loss": 1.7955355644226074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4633668065071106, "epoch": 7.72, "learning_rate": 2.6950532247964938e-05, "loss": 0.5391, "step": 9137, "task_loss": 0.34237217903137207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6981448531150818, "epoch": 7.72, "learning_rate": 2.694740137758297e-05, "loss": 0.6316, "step": 9138, "task_loss": 0.5763481259346008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6472916603088379, "epoch": 7.72, "learning_rate": 2.6944270507201e-05, "loss": 0.9146, "step": 9139, "task_loss": 1.19234037399292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5898531675338745, "epoch": 7.73, "learning_rate": 2.6941139636819036e-05, "loss": 0.7838, "step": 9140, "task_loss": 1.0613124370574951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5404043793678284, "epoch": 7.73, "learning_rate": 2.693800876643707e-05, "loss": 0.5564, "step": 9141, "task_loss": 0.6253383755683899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49706172943115234, "epoch": 7.73, "learning_rate": 2.6934877896055107e-05, "loss": 0.6244, "step": 9142, "task_loss": 1.1078482866287231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7234019041061401, "epoch": 7.73, "learning_rate": 2.693174702567314e-05, "loss": 0.8383, "step": 9143, "task_loss": 1.3914744853973389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41137030720710754, "epoch": 7.73, "learning_rate": 2.692861615529117e-05, "loss": 0.5326, "step": 9144, "task_loss": 0.8546251654624939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6877948045730591, "epoch": 7.73, "learning_rate": 2.692548528490921e-05, "loss": 0.5753, "step": 9145, "task_loss": 1.2407705783843994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39640626311302185, "epoch": 7.73, "learning_rate": 2.692235441452724e-05, "loss": 0.5938, "step": 9146, "task_loss": 0.21289078891277313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4559834897518158, "epoch": 7.73, "learning_rate": 2.6919223544145272e-05, "loss": 0.6036, "step": 9147, "task_loss": 0.3823857605457306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35156694054603577, "epoch": 7.73, "learning_rate": 2.691609267376331e-05, "loss": 0.5601, "step": 9148, "task_loss": 1.0141890048980713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3704158663749695, "epoch": 7.73, "learning_rate": 2.6912961803381343e-05, "loss": 0.6636, "step": 9149, "task_loss": 0.29394254088401794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6009038686752319, "epoch": 7.73, "learning_rate": 2.6909830932999375e-05, "loss": 0.7027, "step": 9150, "task_loss": 1.3807648420333862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.801001787185669, "epoch": 7.73, "learning_rate": 2.6906700062617406e-05, "loss": 0.925, "step": 9151, "task_loss": 1.3853952884674072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6331416368484497, "epoch": 7.74, "learning_rate": 2.6903569192235445e-05, "loss": 0.5551, "step": 9152, "task_loss": 0.7791748046875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.559745192527771, "epoch": 7.74, "learning_rate": 2.6900438321853477e-05, "loss": 0.6603, "step": 9153, "task_loss": 0.44008177518844604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42059290409088135, "epoch": 7.74, "learning_rate": 2.689730745147151e-05, "loss": 0.6537, "step": 9154, "task_loss": 0.5964198708534241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4557032585144043, "epoch": 7.74, "learning_rate": 2.689417658108954e-05, "loss": 0.8449, "step": 9155, "task_loss": 0.42850014567375183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5439982414245605, "epoch": 7.74, "learning_rate": 2.689104571070758e-05, "loss": 0.6045, "step": 9156, "task_loss": 0.8238292932510376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6037514805793762, "epoch": 7.74, "learning_rate": 2.688791484032561e-05, "loss": 0.4933, "step": 9157, "task_loss": 1.0810534954071045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.76021409034729, "epoch": 7.74, "learning_rate": 2.6884783969943646e-05, "loss": 0.6209, "step": 9158, "task_loss": 1.5325013399124146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8941582441329956, "epoch": 7.74, "learning_rate": 2.688165309956168e-05, "loss": 0.6726, "step": 9159, "task_loss": 0.9334650635719299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3945716321468353, "epoch": 7.74, "learning_rate": 2.6878522229179716e-05, "loss": 0.4972, "step": 9160, "task_loss": 0.3486729562282562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6275527477264404, "epoch": 7.74, "learning_rate": 2.6875391358797748e-05, "loss": 0.7046, "step": 9161, "task_loss": 1.0338784456253052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0383257865905762, "epoch": 7.74, "learning_rate": 2.687226048841578e-05, "loss": 0.8262, "step": 9162, "task_loss": 2.142730951309204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5521240830421448, "epoch": 7.75, "learning_rate": 2.6869129618033818e-05, "loss": 0.5289, "step": 9163, "task_loss": 0.6250215768814087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6050926446914673, "epoch": 7.75, "learning_rate": 2.686599874765185e-05, "loss": 0.5939, "step": 9164, "task_loss": 1.1956504583358765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5803313255310059, "epoch": 7.75, "learning_rate": 2.6862867877269882e-05, "loss": 0.6491, "step": 9165, "task_loss": 0.5469316244125366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5977094769477844, "epoch": 7.75, "learning_rate": 2.6859737006887914e-05, "loss": 0.5497, "step": 9166, "task_loss": 0.6004010438919067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9422328472137451, "epoch": 7.75, "learning_rate": 2.6856606136505952e-05, "loss": 0.7198, "step": 9167, "task_loss": 1.7190295457839966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7305265665054321, "epoch": 7.75, "learning_rate": 2.6853475266123984e-05, "loss": 0.7151, "step": 9168, "task_loss": 0.7560840249061584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43146929144859314, "epoch": 7.75, "learning_rate": 2.6850344395742016e-05, "loss": 0.5999, "step": 9169, "task_loss": 0.5697654485702515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.675452709197998, "epoch": 7.75, "learning_rate": 2.6847213525360047e-05, "loss": 0.5613, "step": 9170, "task_loss": 0.718334436416626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.552760124206543, "epoch": 7.75, "learning_rate": 2.6844082654978086e-05, "loss": 0.7887, "step": 9171, "task_loss": 0.9052979946136475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.635266900062561, "epoch": 7.75, "learning_rate": 2.6840951784596118e-05, "loss": 0.7017, "step": 9172, "task_loss": 0.48447006940841675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8776285648345947, "epoch": 7.75, "learning_rate": 2.683782091421415e-05, "loss": 0.7075, "step": 9173, "task_loss": 0.20791159570217133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7506695985794067, "epoch": 7.75, "learning_rate": 2.6834690043832188e-05, "loss": 0.6559, "step": 9174, "task_loss": 0.8048954606056213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8917046189308167, "epoch": 7.76, "learning_rate": 2.683155917345022e-05, "loss": 0.7959, "step": 9175, "task_loss": 0.9829083681106567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7374757528305054, "epoch": 7.76, "learning_rate": 2.6828428303068255e-05, "loss": 0.6194, "step": 9176, "task_loss": 0.29473158717155457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7131641507148743, "epoch": 7.76, "learning_rate": 2.6825297432686287e-05, "loss": 0.5953, "step": 9177, "task_loss": 1.4133472442626953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4728832542896271, "epoch": 7.76, "learning_rate": 2.6822166562304325e-05, "loss": 0.6507, "step": 9178, "task_loss": 0.716389000415802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5142751932144165, "epoch": 7.76, "learning_rate": 2.6819035691922357e-05, "loss": 0.6877, "step": 9179, "task_loss": 0.34924206137657166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8977372646331787, "epoch": 7.76, "learning_rate": 2.681590482154039e-05, "loss": 0.7971, "step": 9180, "task_loss": 1.061044692993164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5276252031326294, "epoch": 7.76, "learning_rate": 2.681277395115842e-05, "loss": 0.7205, "step": 9181, "task_loss": 1.2351791858673096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4838705062866211, "epoch": 7.76, "learning_rate": 2.680964308077646e-05, "loss": 0.584, "step": 9182, "task_loss": 0.08281552046537399 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6336753368377686, "epoch": 7.76, "learning_rate": 2.680651221039449e-05, "loss": 0.8068, "step": 9183, "task_loss": 0.36389410495758057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49715378880500793, "epoch": 7.76, "learning_rate": 2.6803381340012523e-05, "loss": 0.6144, "step": 9184, "task_loss": 0.7244604825973511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5417205095291138, "epoch": 7.76, "learning_rate": 2.680025046963056e-05, "loss": 0.6869, "step": 9185, "task_loss": 0.32977426052093506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.589698314666748, "epoch": 7.76, "learning_rate": 2.6797119599248593e-05, "loss": 0.6187, "step": 9186, "task_loss": 1.5766924619674683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44580602645874023, "epoch": 7.77, "learning_rate": 2.6793988728866625e-05, "loss": 0.5151, "step": 9187, "task_loss": 0.22158858180046082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34135499596595764, "epoch": 7.77, "learning_rate": 2.6790857858484657e-05, "loss": 0.4322, "step": 9188, "task_loss": 0.2685298025608063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8604387640953064, "epoch": 7.77, "learning_rate": 2.6787726988102695e-05, "loss": 0.6284, "step": 9189, "task_loss": 1.2235145568847656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5808174014091492, "epoch": 7.77, "learning_rate": 2.6784596117720727e-05, "loss": 0.6226, "step": 9190, "task_loss": 0.5682020783424377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0652730464935303, "epoch": 7.77, "learning_rate": 2.678146524733876e-05, "loss": 0.6106, "step": 9191, "task_loss": 1.274983286857605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3817666471004486, "epoch": 7.77, "learning_rate": 2.6778334376956794e-05, "loss": 0.6728, "step": 9192, "task_loss": 0.801400899887085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8180332183837891, "epoch": 7.77, "learning_rate": 2.677520350657483e-05, "loss": 0.6594, "step": 9193, "task_loss": 1.271713376045227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7267806529998779, "epoch": 7.77, "learning_rate": 2.6772072636192864e-05, "loss": 0.7888, "step": 9194, "task_loss": 1.5304819345474243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.52070552110672, "epoch": 7.77, "learning_rate": 2.6768941765810896e-05, "loss": 0.7617, "step": 9195, "task_loss": 0.1508764624595642 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.512209415435791, "epoch": 7.77, "learning_rate": 2.6765810895428935e-05, "loss": 0.718, "step": 9196, "task_loss": 0.3649105429649353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7875078916549683, "epoch": 7.77, "learning_rate": 2.6762680025046967e-05, "loss": 0.6812, "step": 9197, "task_loss": 1.2211252450942993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40517425537109375, "epoch": 7.77, "learning_rate": 2.6759549154665e-05, "loss": 0.5786, "step": 9198, "task_loss": 0.3261634409427643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5715659260749817, "epoch": 7.78, "learning_rate": 2.675641828428303e-05, "loss": 0.523, "step": 9199, "task_loss": 0.8337152600288391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8269885778427124, "epoch": 7.78, "learning_rate": 2.675328741390107e-05, "loss": 0.6575, "step": 9200, "task_loss": 1.3502451181411743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7457078695297241, "epoch": 7.78, "learning_rate": 2.67501565435191e-05, "loss": 0.8504, "step": 9201, "task_loss": 1.1310083866119385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8947073221206665, "epoch": 7.78, "learning_rate": 2.6747025673137132e-05, "loss": 0.6861, "step": 9202, "task_loss": 0.9751021862030029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5358555912971497, "epoch": 7.78, "learning_rate": 2.6743894802755164e-05, "loss": 0.6358, "step": 9203, "task_loss": 1.5834178924560547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7293906211853027, "epoch": 7.78, "learning_rate": 2.6740763932373203e-05, "loss": 0.696, "step": 9204, "task_loss": 1.1178241968154907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6302853226661682, "epoch": 7.78, "learning_rate": 2.6737633061991234e-05, "loss": 0.6301, "step": 9205, "task_loss": 0.5495689511299133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8045634031295776, "epoch": 7.78, "learning_rate": 2.6734502191609266e-05, "loss": 0.6677, "step": 9206, "task_loss": 0.5293760895729065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4343588054180145, "epoch": 7.78, "learning_rate": 2.67313713212273e-05, "loss": 0.6327, "step": 9207, "task_loss": 0.5476627349853516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5578230023384094, "epoch": 7.78, "learning_rate": 2.6728240450845336e-05, "loss": 0.7349, "step": 9208, "task_loss": 0.5010431408882141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4956943392753601, "epoch": 7.78, "learning_rate": 2.672510958046337e-05, "loss": 0.6185, "step": 9209, "task_loss": 0.3393818438053131 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5098996758460999, "epoch": 7.78, "learning_rate": 2.6721978710081403e-05, "loss": 0.7275, "step": 9210, "task_loss": 1.2103461027145386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6454741954803467, "epoch": 7.79, "learning_rate": 2.671884783969944e-05, "loss": 0.7429, "step": 9211, "task_loss": 0.4730529189109802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6113182306289673, "epoch": 7.79, "learning_rate": 2.6715716969317474e-05, "loss": 0.8538, "step": 9212, "task_loss": 1.428511142730713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45839357376098633, "epoch": 7.79, "learning_rate": 2.6712586098935506e-05, "loss": 0.5296, "step": 9213, "task_loss": 0.578133761882782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38401466608047485, "epoch": 7.79, "learning_rate": 2.6709455228553537e-05, "loss": 0.6647, "step": 9214, "task_loss": 0.22821903228759766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4478796720504761, "epoch": 7.79, "learning_rate": 2.6706324358171576e-05, "loss": 0.5029, "step": 9215, "task_loss": 0.591279923915863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5743421316146851, "epoch": 7.79, "learning_rate": 2.6703193487789608e-05, "loss": 0.7756, "step": 9216, "task_loss": 0.5161247849464417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4855937063694, "epoch": 7.79, "learning_rate": 2.670006261740764e-05, "loss": 0.6962, "step": 9217, "task_loss": 0.9424035549163818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8942943811416626, "epoch": 7.79, "learning_rate": 2.669693174702567e-05, "loss": 0.6711, "step": 9218, "task_loss": 1.6510564088821411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5233988165855408, "epoch": 7.79, "learning_rate": 2.669380087664371e-05, "loss": 0.7457, "step": 9219, "task_loss": 0.5425822138786316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43656718730926514, "epoch": 7.79, "learning_rate": 2.669067000626174e-05, "loss": 0.5057, "step": 9220, "task_loss": 0.8772517442703247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7960867881774902, "epoch": 7.79, "learning_rate": 2.6687539135879773e-05, "loss": 0.6597, "step": 9221, "task_loss": 0.38597068190574646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5789438486099243, "epoch": 7.79, "learning_rate": 2.6684408265497812e-05, "loss": 0.7179, "step": 9222, "task_loss": 0.6204952597618103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8370921015739441, "epoch": 7.8, "learning_rate": 2.6681277395115844e-05, "loss": 0.7058, "step": 9223, "task_loss": 1.0985074043273926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46336376667022705, "epoch": 7.8, "learning_rate": 2.6678146524733875e-05, "loss": 0.5609, "step": 9224, "task_loss": 0.5612239241600037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4742278456687927, "epoch": 7.8, "learning_rate": 2.667501565435191e-05, "loss": 0.6148, "step": 9225, "task_loss": 0.6976776123046875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7475371956825256, "epoch": 7.8, "learning_rate": 2.6671884783969946e-05, "loss": 0.9523, "step": 9226, "task_loss": 1.2583519220352173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7526002526283264, "epoch": 7.8, "learning_rate": 2.666875391358798e-05, "loss": 0.6903, "step": 9227, "task_loss": 1.0588525533676147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47655099630355835, "epoch": 7.8, "learning_rate": 2.6665623043206013e-05, "loss": 0.6173, "step": 9228, "task_loss": 0.18094955384731293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0975885391235352, "epoch": 7.8, "learning_rate": 2.6662492172824045e-05, "loss": 0.6989, "step": 9229, "task_loss": 0.9413714408874512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4896625280380249, "epoch": 7.8, "learning_rate": 2.6659361302442083e-05, "loss": 0.7129, "step": 9230, "task_loss": 0.8934268951416016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.587326169013977, "epoch": 7.8, "learning_rate": 2.6656230432060115e-05, "loss": 0.526, "step": 9231, "task_loss": 0.4609638750553131 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7431254982948303, "epoch": 7.8, "learning_rate": 2.6653099561678147e-05, "loss": 0.5417, "step": 9232, "task_loss": 0.7307233214378357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7022340297698975, "epoch": 7.8, "learning_rate": 2.6649968691296185e-05, "loss": 0.7131, "step": 9233, "task_loss": 0.9116626381874084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9205322861671448, "epoch": 7.81, "learning_rate": 2.6646837820914217e-05, "loss": 0.6594, "step": 9234, "task_loss": 0.5586472749710083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5777369141578674, "epoch": 7.81, "learning_rate": 2.664370695053225e-05, "loss": 0.6747, "step": 9235, "task_loss": 0.3533806800842285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7960163950920105, "epoch": 7.81, "learning_rate": 2.664057608015028e-05, "loss": 0.6405, "step": 9236, "task_loss": 0.934609055519104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7135106325149536, "epoch": 7.81, "learning_rate": 2.663744520976832e-05, "loss": 0.537, "step": 9237, "task_loss": 1.109183430671692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8131632208824158, "epoch": 7.81, "learning_rate": 2.663431433938635e-05, "loss": 0.6348, "step": 9238, "task_loss": 0.5855390429496765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2039202451705933, "epoch": 7.81, "learning_rate": 2.6631183469004383e-05, "loss": 0.6565, "step": 9239, "task_loss": 0.7804291844367981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.634016215801239, "epoch": 7.81, "learning_rate": 2.6628052598622414e-05, "loss": 0.684, "step": 9240, "task_loss": 0.20485365390777588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5084894299507141, "epoch": 7.81, "learning_rate": 2.6624921728240453e-05, "loss": 0.5499, "step": 9241, "task_loss": 0.3662354350090027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5975803136825562, "epoch": 7.81, "learning_rate": 2.6621790857858485e-05, "loss": 0.619, "step": 9242, "task_loss": 0.7465779185295105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.667792558670044, "epoch": 7.81, "learning_rate": 2.661865998747652e-05, "loss": 0.6468, "step": 9243, "task_loss": 0.5328837037086487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3583890199661255, "epoch": 7.81, "learning_rate": 2.6615529117094552e-05, "loss": 0.4138, "step": 9244, "task_loss": 1.0013641119003296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5177762508392334, "epoch": 7.81, "learning_rate": 2.661239824671259e-05, "loss": 0.7418, "step": 9245, "task_loss": 1.5827080011367798 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5794281959533691, "epoch": 7.82, "learning_rate": 2.6609267376330622e-05, "loss": 0.7053, "step": 9246, "task_loss": 0.5535288453102112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5955565571784973, "epoch": 7.82, "learning_rate": 2.6606136505948654e-05, "loss": 0.609, "step": 9247, "task_loss": 0.991658091545105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8643224239349365, "epoch": 7.82, "learning_rate": 2.6603005635566692e-05, "loss": 0.7522, "step": 9248, "task_loss": 1.574628233909607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6795052289962769, "epoch": 7.82, "learning_rate": 2.6599874765184724e-05, "loss": 0.594, "step": 9249, "task_loss": 0.8551183342933655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6969977617263794, "epoch": 7.82, "learning_rate": 2.6596743894802756e-05, "loss": 0.5752, "step": 9250, "task_loss": 0.8130627274513245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7574191689491272, "epoch": 7.82, "learning_rate": 2.6593613024420788e-05, "loss": 0.7109, "step": 9251, "task_loss": 1.4849987030029297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40362757444381714, "epoch": 7.82, "learning_rate": 2.6590482154038826e-05, "loss": 0.5228, "step": 9252, "task_loss": 1.2725516557693481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5007966756820679, "epoch": 7.82, "learning_rate": 2.6587351283656858e-05, "loss": 0.7343, "step": 9253, "task_loss": 0.5518678426742554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5868730545043945, "epoch": 7.82, "learning_rate": 2.658422041327489e-05, "loss": 0.77, "step": 9254, "task_loss": 0.20289430022239685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9212419986724854, "epoch": 7.82, "learning_rate": 2.658108954289292e-05, "loss": 0.7759, "step": 9255, "task_loss": 1.2214471101760864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5847617387771606, "epoch": 7.82, "learning_rate": 2.657795867251096e-05, "loss": 0.6034, "step": 9256, "task_loss": 0.7956830263137817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5877705216407776, "epoch": 7.82, "learning_rate": 2.6574827802128992e-05, "loss": 0.6475, "step": 9257, "task_loss": 0.9570707082748413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.628792405128479, "epoch": 7.83, "learning_rate": 2.6571696931747024e-05, "loss": 0.507, "step": 9258, "task_loss": 0.4242516756057739 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9908457398414612, "epoch": 7.83, "learning_rate": 2.6568566061365062e-05, "loss": 0.7435, "step": 9259, "task_loss": 0.825901985168457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7686820030212402, "epoch": 7.83, "learning_rate": 2.6565435190983094e-05, "loss": 0.6236, "step": 9260, "task_loss": 1.0070436000823975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.663603663444519, "epoch": 7.83, "learning_rate": 2.656230432060113e-05, "loss": 0.5836, "step": 9261, "task_loss": 1.1074930429458618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6836071014404297, "epoch": 7.83, "learning_rate": 2.655917345021916e-05, "loss": 0.7068, "step": 9262, "task_loss": 0.20346979796886444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.712429404258728, "epoch": 7.83, "learning_rate": 2.65560425798372e-05, "loss": 0.4568, "step": 9263, "task_loss": 0.8220543265342712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9652332067489624, "epoch": 7.83, "learning_rate": 2.655291170945523e-05, "loss": 0.5744, "step": 9264, "task_loss": 1.0026421546936035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4644193649291992, "epoch": 7.83, "learning_rate": 2.6549780839073263e-05, "loss": 0.5888, "step": 9265, "task_loss": 0.24481558799743652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4696943759918213, "epoch": 7.83, "learning_rate": 2.6546649968691295e-05, "loss": 0.778, "step": 9266, "task_loss": 0.07167177647352219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0001804828643799, "epoch": 7.83, "learning_rate": 2.6543519098309334e-05, "loss": 0.7251, "step": 9267, "task_loss": 1.4573928117752075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46024081110954285, "epoch": 7.83, "learning_rate": 2.6540388227927365e-05, "loss": 0.698, "step": 9268, "task_loss": 1.1027545928955078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4472285807132721, "epoch": 7.83, "learning_rate": 2.6537257357545397e-05, "loss": 0.6512, "step": 9269, "task_loss": 0.3202151954174042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6706910133361816, "epoch": 7.84, "learning_rate": 2.6534126487163436e-05, "loss": 0.6593, "step": 9270, "task_loss": 0.1355791687965393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7362587451934814, "epoch": 7.84, "learning_rate": 2.6530995616781467e-05, "loss": 0.7963, "step": 9271, "task_loss": 0.6590666174888611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7618979215621948, "epoch": 7.84, "learning_rate": 2.65278647463995e-05, "loss": 0.7391, "step": 9272, "task_loss": 1.2034478187561035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43591901659965515, "epoch": 7.84, "learning_rate": 2.652473387601753e-05, "loss": 0.6104, "step": 9273, "task_loss": 0.988061249256134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41245830059051514, "epoch": 7.84, "learning_rate": 2.652160300563557e-05, "loss": 0.4372, "step": 9274, "task_loss": 1.045927882194519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9924241304397583, "epoch": 7.84, "learning_rate": 2.65184721352536e-05, "loss": 0.7929, "step": 9275, "task_loss": 1.805729866027832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.477022260427475, "epoch": 7.84, "learning_rate": 2.6515341264871633e-05, "loss": 0.6115, "step": 9276, "task_loss": 0.45830243825912476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4934313893318176, "epoch": 7.84, "learning_rate": 2.651221039448967e-05, "loss": 0.5061, "step": 9277, "task_loss": 0.8648338913917542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5868679285049438, "epoch": 7.84, "learning_rate": 2.6509079524107703e-05, "loss": 0.6865, "step": 9278, "task_loss": 0.6288072466850281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43425801396369934, "epoch": 7.84, "learning_rate": 2.650594865372574e-05, "loss": 0.5412, "step": 9279, "task_loss": 0.22652484476566315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4973319470882416, "epoch": 7.84, "learning_rate": 2.650281778334377e-05, "loss": 0.6162, "step": 9280, "task_loss": 0.2765994966030121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6030449867248535, "epoch": 7.84, "learning_rate": 2.6499686912961802e-05, "loss": 0.5551, "step": 9281, "task_loss": 0.7147455811500549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4962291717529297, "epoch": 7.85, "learning_rate": 2.649655604257984e-05, "loss": 0.6892, "step": 9282, "task_loss": 0.466047078371048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40010204911231995, "epoch": 7.85, "learning_rate": 2.6493425172197873e-05, "loss": 0.4289, "step": 9283, "task_loss": 1.2293390035629272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8097793459892273, "epoch": 7.85, "learning_rate": 2.6490294301815904e-05, "loss": 0.589, "step": 9284, "task_loss": 0.608290433883667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.613182783126831, "epoch": 7.85, "learning_rate": 2.6487163431433943e-05, "loss": 0.6836, "step": 9285, "task_loss": 0.8006587028503418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7904977202415466, "epoch": 7.85, "learning_rate": 2.6484032561051975e-05, "loss": 0.7538, "step": 9286, "task_loss": 0.6184447407722473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3588023781776428, "epoch": 7.85, "learning_rate": 2.6480901690670006e-05, "loss": 0.5509, "step": 9287, "task_loss": 0.5030600428581238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.541572093963623, "epoch": 7.85, "learning_rate": 2.6477770820288038e-05, "loss": 0.6406, "step": 9288, "task_loss": 1.3499858379364014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5426823496818542, "epoch": 7.85, "learning_rate": 2.6474639949906077e-05, "loss": 0.6152, "step": 9289, "task_loss": 0.5274702906608582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42585599422454834, "epoch": 7.85, "learning_rate": 2.647150907952411e-05, "loss": 0.5311, "step": 9290, "task_loss": 0.857465386390686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4738660454750061, "epoch": 7.85, "learning_rate": 2.646837820914214e-05, "loss": 0.5171, "step": 9291, "task_loss": 0.21938294172286987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.796672523021698, "epoch": 7.85, "learning_rate": 2.6465247338760176e-05, "loss": 0.6093, "step": 9292, "task_loss": 1.0962607860565186 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5272184610366821, "epoch": 7.85, "learning_rate": 2.646211646837821e-05, "loss": 0.5784, "step": 9293, "task_loss": 0.34344375133514404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4836212992668152, "epoch": 7.86, "learning_rate": 2.6458985597996246e-05, "loss": 0.5105, "step": 9294, "task_loss": 0.7175370454788208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40601760149002075, "epoch": 7.86, "learning_rate": 2.6455854727614278e-05, "loss": 0.4977, "step": 9295, "task_loss": 0.6791109442710876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4554453492164612, "epoch": 7.86, "learning_rate": 2.6452723857232313e-05, "loss": 0.5515, "step": 9296, "task_loss": 0.4111195504665375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8485012054443359, "epoch": 7.86, "learning_rate": 2.6449592986850348e-05, "loss": 0.7548, "step": 9297, "task_loss": 0.6032215356826782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.06626558303833, "epoch": 7.86, "learning_rate": 2.644646211646838e-05, "loss": 0.6507, "step": 9298, "task_loss": 0.7646085619926453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9918057918548584, "epoch": 7.86, "learning_rate": 2.644333124608641e-05, "loss": 0.7591, "step": 9299, "task_loss": 1.350124716758728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7197192311286926, "epoch": 7.86, "learning_rate": 2.644020037570445e-05, "loss": 0.6251, "step": 9300, "task_loss": 1.5529576539993286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5720992684364319, "epoch": 7.86, "learning_rate": 2.6437069505322482e-05, "loss": 0.5405, "step": 9301, "task_loss": 0.7065452337265015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4943564534187317, "epoch": 7.86, "learning_rate": 2.6433938634940514e-05, "loss": 0.6399, "step": 9302, "task_loss": 1.18608820438385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.564371645450592, "epoch": 7.86, "learning_rate": 2.6430807764558545e-05, "loss": 0.5333, "step": 9303, "task_loss": 0.4874054193496704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5280548334121704, "epoch": 7.86, "learning_rate": 2.6427676894176584e-05, "loss": 0.6466, "step": 9304, "task_loss": 1.0137883424758911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.704071581363678, "epoch": 7.87, "learning_rate": 2.6424546023794616e-05, "loss": 0.7072, "step": 9305, "task_loss": 0.8883162140846252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7928359508514404, "epoch": 7.87, "learning_rate": 2.6421415153412648e-05, "loss": 0.6501, "step": 9306, "task_loss": 0.9003003835678101 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8449357748031616, "epoch": 7.87, "learning_rate": 2.6418284283030686e-05, "loss": 0.6602, "step": 9307, "task_loss": 1.4054615497589111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7677913308143616, "epoch": 7.87, "learning_rate": 2.6415153412648718e-05, "loss": 0.5254, "step": 9308, "task_loss": 0.5920577049255371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1572167873382568, "epoch": 7.87, "learning_rate": 2.641202254226675e-05, "loss": 0.8988, "step": 9309, "task_loss": 0.9718706011772156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7753601670265198, "epoch": 7.87, "learning_rate": 2.6408891671884785e-05, "loss": 0.71, "step": 9310, "task_loss": 1.0363447666168213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7899887561798096, "epoch": 7.87, "learning_rate": 2.640576080150282e-05, "loss": 0.8088, "step": 9311, "task_loss": 0.9770612716674805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6340683698654175, "epoch": 7.87, "learning_rate": 2.6402629931120855e-05, "loss": 0.9231, "step": 9312, "task_loss": 0.9529533386230469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5003923773765564, "epoch": 7.87, "learning_rate": 2.6399499060738887e-05, "loss": 0.6233, "step": 9313, "task_loss": 0.8169487714767456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44133347272872925, "epoch": 7.87, "learning_rate": 2.639636819035692e-05, "loss": 0.7322, "step": 9314, "task_loss": 0.5067716240882874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32436782121658325, "epoch": 7.87, "learning_rate": 2.6393237319974957e-05, "loss": 0.644, "step": 9315, "task_loss": 1.0155659914016724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5007916688919067, "epoch": 7.87, "learning_rate": 2.639010644959299e-05, "loss": 0.5865, "step": 9316, "task_loss": 0.5943099856376648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6835108995437622, "epoch": 7.88, "learning_rate": 2.638697557921102e-05, "loss": 0.68, "step": 9317, "task_loss": 1.291317105293274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5380743741989136, "epoch": 7.88, "learning_rate": 2.6383844708829053e-05, "loss": 0.6953, "step": 9318, "task_loss": 0.7644554376602173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7863419651985168, "epoch": 7.88, "learning_rate": 2.638071383844709e-05, "loss": 0.7293, "step": 9319, "task_loss": 0.461905300617218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.428444504737854, "epoch": 7.88, "learning_rate": 2.6377582968065123e-05, "loss": 0.572, "step": 9320, "task_loss": 0.5470868945121765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6691455245018005, "epoch": 7.88, "learning_rate": 2.6374452097683155e-05, "loss": 0.6684, "step": 9321, "task_loss": 0.811172366142273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8944703340530396, "epoch": 7.88, "learning_rate": 2.6371321227301193e-05, "loss": 0.6915, "step": 9322, "task_loss": 0.4624226987361908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4571475684642792, "epoch": 7.88, "learning_rate": 2.6368190356919225e-05, "loss": 0.585, "step": 9323, "task_loss": 0.5568109154701233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6200003027915955, "epoch": 7.88, "learning_rate": 2.6365059486537257e-05, "loss": 0.7811, "step": 9324, "task_loss": 1.2458839416503906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2970258593559265, "epoch": 7.88, "learning_rate": 2.636192861615529e-05, "loss": 0.6289, "step": 9325, "task_loss": 0.9157501459121704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9564588069915771, "epoch": 7.88, "learning_rate": 2.6358797745773327e-05, "loss": 0.9148, "step": 9326, "task_loss": 1.2216918468475342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5756273865699768, "epoch": 7.88, "learning_rate": 2.635566687539136e-05, "loss": 0.7962, "step": 9327, "task_loss": 0.5107131600379944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7616927027702332, "epoch": 7.88, "learning_rate": 2.6352536005009394e-05, "loss": 0.6417, "step": 9328, "task_loss": 1.378266453742981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4852607846260071, "epoch": 7.89, "learning_rate": 2.6349405134627426e-05, "loss": 0.6567, "step": 9329, "task_loss": 1.1085591316223145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40316298604011536, "epoch": 7.89, "learning_rate": 2.6346274264245465e-05, "loss": 0.5463, "step": 9330, "task_loss": 0.4773106575012207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0580435991287231, "epoch": 7.89, "learning_rate": 2.6343143393863496e-05, "loss": 0.6375, "step": 9331, "task_loss": 0.8668399453163147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5490967035293579, "epoch": 7.89, "learning_rate": 2.6340012523481528e-05, "loss": 0.749, "step": 9332, "task_loss": 0.7925817966461182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5817394256591797, "epoch": 7.89, "learning_rate": 2.6336881653099567e-05, "loss": 0.5528, "step": 9333, "task_loss": 1.2291899919509888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5919026732444763, "epoch": 7.89, "learning_rate": 2.63337507827176e-05, "loss": 0.657, "step": 9334, "task_loss": 0.5199276208877563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3289984166622162, "epoch": 7.89, "learning_rate": 2.633061991233563e-05, "loss": 0.5227, "step": 9335, "task_loss": 0.6141798496246338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1123275756835938, "epoch": 7.89, "learning_rate": 2.6327489041953662e-05, "loss": 0.8578, "step": 9336, "task_loss": 1.1024904251098633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3896951377391815, "epoch": 7.89, "learning_rate": 2.63243581715717e-05, "loss": 0.5881, "step": 9337, "task_loss": 0.3597799241542816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5285160541534424, "epoch": 7.89, "learning_rate": 2.6321227301189732e-05, "loss": 0.7202, "step": 9338, "task_loss": 1.1613010168075562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7277591228485107, "epoch": 7.89, "learning_rate": 2.6318096430807764e-05, "loss": 0.6768, "step": 9339, "task_loss": 1.2189006805419922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5512917637825012, "epoch": 7.89, "learning_rate": 2.6314965560425796e-05, "loss": 0.5125, "step": 9340, "task_loss": 0.5204025506973267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3299836218357086, "epoch": 7.9, "learning_rate": 2.6311834690043834e-05, "loss": 0.4672, "step": 9341, "task_loss": 0.08439314365386963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3428913652896881, "epoch": 7.9, "learning_rate": 2.6308703819661866e-05, "loss": 0.6465, "step": 9342, "task_loss": 0.21145455539226532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5187602639198303, "epoch": 7.9, "learning_rate": 2.6305572949279898e-05, "loss": 0.6903, "step": 9343, "task_loss": 0.887890100479126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8929547071456909, "epoch": 7.9, "learning_rate": 2.6302442078897937e-05, "loss": 0.7915, "step": 9344, "task_loss": 0.9381954669952393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42203137278556824, "epoch": 7.9, "learning_rate": 2.629931120851597e-05, "loss": 0.4985, "step": 9345, "task_loss": 0.1595657914876938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7182854413986206, "epoch": 7.9, "learning_rate": 2.6296180338134004e-05, "loss": 0.6369, "step": 9346, "task_loss": 1.2117211818695068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5900564193725586, "epoch": 7.9, "learning_rate": 2.6293049467752035e-05, "loss": 0.5886, "step": 9347, "task_loss": 0.46759557723999023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5240297317504883, "epoch": 7.9, "learning_rate": 2.6289918597370074e-05, "loss": 0.5978, "step": 9348, "task_loss": 0.5658738017082214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.841759979724884, "epoch": 7.9, "learning_rate": 2.6286787726988106e-05, "loss": 0.7367, "step": 9349, "task_loss": 0.846861720085144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2734088897705078, "epoch": 7.9, "learning_rate": 2.6283656856606137e-05, "loss": 0.5741, "step": 9350, "task_loss": 0.6389260292053223 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3418160080909729, "epoch": 7.9, "learning_rate": 2.628052598622417e-05, "loss": 0.5191, "step": 9351, "task_loss": 0.03168213367462158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6644775867462158, "epoch": 7.9, "learning_rate": 2.6277395115842208e-05, "loss": 0.629, "step": 9352, "task_loss": 0.5027097463607788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.085130214691162, "epoch": 7.91, "learning_rate": 2.627426424546024e-05, "loss": 0.7823, "step": 9353, "task_loss": 1.167847752571106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5684057474136353, "epoch": 7.91, "learning_rate": 2.627113337507827e-05, "loss": 0.6604, "step": 9354, "task_loss": 0.9169731140136719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5032474398612976, "epoch": 7.91, "learning_rate": 2.6268002504696303e-05, "loss": 0.4908, "step": 9355, "task_loss": 0.4583607614040375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7406501770019531, "epoch": 7.91, "learning_rate": 2.626487163431434e-05, "loss": 0.8252, "step": 9356, "task_loss": 0.9404286742210388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5131732225418091, "epoch": 7.91, "learning_rate": 2.6261740763932373e-05, "loss": 0.5304, "step": 9357, "task_loss": 0.8052062392234802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6987851858139038, "epoch": 7.91, "learning_rate": 2.6258609893550405e-05, "loss": 0.6785, "step": 9358, "task_loss": 0.5692974925041199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6434668302536011, "epoch": 7.91, "learning_rate": 2.6255479023168444e-05, "loss": 0.6611, "step": 9359, "task_loss": 0.10130046308040619 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.562297523021698, "epoch": 7.91, "learning_rate": 2.6252348152786476e-05, "loss": 0.5461, "step": 9360, "task_loss": 0.7579794526100159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8335858583450317, "epoch": 7.91, "learning_rate": 2.624921728240451e-05, "loss": 0.6642, "step": 9361, "task_loss": 1.6068545579910278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8229688405990601, "epoch": 7.91, "learning_rate": 2.6246086412022543e-05, "loss": 0.6961, "step": 9362, "task_loss": 0.5232741832733154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4362673759460449, "epoch": 7.91, "learning_rate": 2.6242955541640578e-05, "loss": 0.5478, "step": 9363, "task_loss": 0.43976709246635437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5081331133842468, "epoch": 7.91, "learning_rate": 2.6239824671258613e-05, "loss": 0.615, "step": 9364, "task_loss": 0.2826498746871948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1123449802398682, "epoch": 7.92, "learning_rate": 2.6236693800876645e-05, "loss": 0.6374, "step": 9365, "task_loss": 0.6675984859466553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9791793823242188, "epoch": 7.92, "learning_rate": 2.6233562930494676e-05, "loss": 0.5453, "step": 9366, "task_loss": 0.7215614914894104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8511583805084229, "epoch": 7.92, "learning_rate": 2.6230432060112715e-05, "loss": 0.6961, "step": 9367, "task_loss": 0.6191051602363586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37676307559013367, "epoch": 7.92, "learning_rate": 2.6227301189730747e-05, "loss": 0.7975, "step": 9368, "task_loss": 0.3966169059276581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6009849309921265, "epoch": 7.92, "learning_rate": 2.622417031934878e-05, "loss": 0.6546, "step": 9369, "task_loss": 1.25436532497406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4462031424045563, "epoch": 7.92, "learning_rate": 2.6221039448966817e-05, "loss": 0.7007, "step": 9370, "task_loss": 0.6118512749671936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.462983638048172, "epoch": 7.92, "learning_rate": 2.621790857858485e-05, "loss": 0.6377, "step": 9371, "task_loss": 1.085371732711792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32398712635040283, "epoch": 7.92, "learning_rate": 2.621477770820288e-05, "loss": 0.5868, "step": 9372, "task_loss": 0.5320574641227722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47642120718955994, "epoch": 7.92, "learning_rate": 2.6211646837820912e-05, "loss": 0.6429, "step": 9373, "task_loss": 1.2664083242416382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8522099852561951, "epoch": 7.92, "learning_rate": 2.620851596743895e-05, "loss": 0.6298, "step": 9374, "task_loss": 0.9169889092445374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4387223720550537, "epoch": 7.92, "learning_rate": 2.6205385097056983e-05, "loss": 0.7189, "step": 9375, "task_loss": 0.8078044652938843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5148763656616211, "epoch": 7.93, "learning_rate": 2.6202254226675015e-05, "loss": 0.6839, "step": 9376, "task_loss": 1.0820128917694092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5897771120071411, "epoch": 7.93, "learning_rate": 2.619912335629305e-05, "loss": 0.9341, "step": 9377, "task_loss": 0.3495389521121979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.928942084312439, "epoch": 7.93, "learning_rate": 2.6195992485911085e-05, "loss": 0.7479, "step": 9378, "task_loss": 1.052635669708252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5640789270401001, "epoch": 7.93, "learning_rate": 2.619286161552912e-05, "loss": 0.655, "step": 9379, "task_loss": 1.1689515113830566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7048307657241821, "epoch": 7.93, "learning_rate": 2.6189730745147152e-05, "loss": 0.7856, "step": 9380, "task_loss": 1.123335599899292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.544011116027832, "epoch": 7.93, "learning_rate": 2.6186599874765187e-05, "loss": 0.5522, "step": 9381, "task_loss": 0.13520658016204834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26411330699920654, "epoch": 7.93, "learning_rate": 2.6183469004383222e-05, "loss": 0.3892, "step": 9382, "task_loss": 0.250189870595932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0258244276046753, "epoch": 7.93, "learning_rate": 2.6180338134001254e-05, "loss": 0.7696, "step": 9383, "task_loss": 1.0003072023391724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49534711241722107, "epoch": 7.93, "learning_rate": 2.6177207263619286e-05, "loss": 0.7138, "step": 9384, "task_loss": 0.7372139692306519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.410172700881958, "epoch": 7.93, "learning_rate": 2.6174076393237324e-05, "loss": 0.5329, "step": 9385, "task_loss": 0.9143943190574646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6105179786682129, "epoch": 7.93, "learning_rate": 2.6170945522855356e-05, "loss": 0.6036, "step": 9386, "task_loss": 1.2219114303588867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5780544877052307, "epoch": 7.93, "learning_rate": 2.6167814652473388e-05, "loss": 0.6325, "step": 9387, "task_loss": 0.346829354763031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5846349000930786, "epoch": 7.94, "learning_rate": 2.616468378209142e-05, "loss": 0.658, "step": 9388, "task_loss": 1.2064036130905151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46985170245170593, "epoch": 7.94, "learning_rate": 2.6161552911709458e-05, "loss": 0.4141, "step": 9389, "task_loss": 0.40717899799346924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0399644374847412, "epoch": 7.94, "learning_rate": 2.615842204132749e-05, "loss": 0.589, "step": 9390, "task_loss": 0.6739776730537415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7163761854171753, "epoch": 7.94, "learning_rate": 2.6155291170945522e-05, "loss": 0.6142, "step": 9391, "task_loss": 1.5287554264068604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4507385492324829, "epoch": 7.94, "learning_rate": 2.6152160300563554e-05, "loss": 0.7075, "step": 9392, "task_loss": 0.3624308109283447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5673612356185913, "epoch": 7.94, "learning_rate": 2.6149029430181592e-05, "loss": 0.5238, "step": 9393, "task_loss": 0.7195814847946167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2776908576488495, "epoch": 7.94, "learning_rate": 2.6145898559799624e-05, "loss": 0.6082, "step": 9394, "task_loss": 0.009602994658052921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1521052122116089, "epoch": 7.94, "learning_rate": 2.614276768941766e-05, "loss": 0.8695, "step": 9395, "task_loss": 1.7038264274597168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5769229531288147, "epoch": 7.94, "learning_rate": 2.6139636819035694e-05, "loss": 0.6069, "step": 9396, "task_loss": 1.1659736633300781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.741196870803833, "epoch": 7.94, "learning_rate": 2.613650594865373e-05, "loss": 0.5766, "step": 9397, "task_loss": 0.7894127368927002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7883368134498596, "epoch": 7.94, "learning_rate": 2.613337507827176e-05, "loss": 0.696, "step": 9398, "task_loss": 2.7792344093322754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6554974913597107, "epoch": 7.94, "learning_rate": 2.6130244207889793e-05, "loss": 0.6543, "step": 9399, "task_loss": 0.5316829681396484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.515598714351654, "epoch": 7.95, "learning_rate": 2.612711333750783e-05, "loss": 0.6316, "step": 9400, "task_loss": 1.1673476696014404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36086446046829224, "epoch": 7.95, "learning_rate": 2.6123982467125863e-05, "loss": 0.5346, "step": 9401, "task_loss": 0.09114090353250504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42424049973487854, "epoch": 7.95, "learning_rate": 2.6120851596743895e-05, "loss": 0.6503, "step": 9402, "task_loss": 1.5616929531097412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5498998165130615, "epoch": 7.95, "learning_rate": 2.6117720726361927e-05, "loss": 0.7337, "step": 9403, "task_loss": 0.47391027212142944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6231445670127869, "epoch": 7.95, "learning_rate": 2.6114589855979965e-05, "loss": 0.7413, "step": 9404, "task_loss": 0.7035879492759705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37753087282180786, "epoch": 7.95, "learning_rate": 2.6111458985597997e-05, "loss": 0.7168, "step": 9405, "task_loss": 0.20330378413200378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7276468276977539, "epoch": 7.95, "learning_rate": 2.610832811521603e-05, "loss": 0.6702, "step": 9406, "task_loss": 0.4897690415382385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3416674733161926, "epoch": 7.95, "learning_rate": 2.6105197244834068e-05, "loss": 0.5369, "step": 9407, "task_loss": 0.30747827887535095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5749080777168274, "epoch": 7.95, "learning_rate": 2.61020663744521e-05, "loss": 0.6022, "step": 9408, "task_loss": 0.38317736983299255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5247406363487244, "epoch": 7.95, "learning_rate": 2.609893550407013e-05, "loss": 0.6314, "step": 9409, "task_loss": 0.48993104696273804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6031020879745483, "epoch": 7.95, "learning_rate": 2.6095804633688163e-05, "loss": 0.6664, "step": 9410, "task_loss": 0.5743944048881531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4987487196922302, "epoch": 7.95, "learning_rate": 2.60926737633062e-05, "loss": 0.4467, "step": 9411, "task_loss": 0.14106020331382751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6303673982620239, "epoch": 7.96, "learning_rate": 2.6089542892924233e-05, "loss": 0.684, "step": 9412, "task_loss": 0.6038556098937988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3771728277206421, "epoch": 7.96, "learning_rate": 2.608641202254227e-05, "loss": 0.4651, "step": 9413, "task_loss": 0.4619561731815338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7802491188049316, "epoch": 7.96, "learning_rate": 2.60832811521603e-05, "loss": 0.7372, "step": 9414, "task_loss": 0.35579031705856323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48948347568511963, "epoch": 7.96, "learning_rate": 2.608015028177834e-05, "loss": 0.6069, "step": 9415, "task_loss": 1.4515137672424316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7085574269294739, "epoch": 7.96, "learning_rate": 2.607701941139637e-05, "loss": 0.5628, "step": 9416, "task_loss": 0.7701581120491028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4865577816963196, "epoch": 7.96, "learning_rate": 2.6073888541014402e-05, "loss": 0.5074, "step": 9417, "task_loss": 0.5167707800865173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5602061748504639, "epoch": 7.96, "learning_rate": 2.607075767063244e-05, "loss": 0.7044, "step": 9418, "task_loss": 0.8063457608222961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6498192548751831, "epoch": 7.96, "learning_rate": 2.6067626800250473e-05, "loss": 0.7601, "step": 9419, "task_loss": 0.9463391900062561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5575873851776123, "epoch": 7.96, "learning_rate": 2.6064495929868504e-05, "loss": 0.5556, "step": 9420, "task_loss": 0.7187119722366333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7753435373306274, "epoch": 7.96, "learning_rate": 2.6061365059486536e-05, "loss": 0.5645, "step": 9421, "task_loss": 0.5484713912010193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.502394437789917, "epoch": 7.96, "learning_rate": 2.6058234189104575e-05, "loss": 0.9315, "step": 9422, "task_loss": 1.021091341972351 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8097443580627441, "epoch": 7.96, "learning_rate": 2.6055103318722607e-05, "loss": 0.598, "step": 9423, "task_loss": 1.0966875553131104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5386875867843628, "epoch": 7.97, "learning_rate": 2.6051972448340638e-05, "loss": 0.5672, "step": 9424, "task_loss": 0.37331894040107727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6949682235717773, "epoch": 7.97, "learning_rate": 2.604884157795867e-05, "loss": 0.6849, "step": 9425, "task_loss": 1.0116904973983765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3641091287136078, "epoch": 7.97, "learning_rate": 2.604571070757671e-05, "loss": 0.5954, "step": 9426, "task_loss": 0.06732720136642456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6232872605323792, "epoch": 7.97, "learning_rate": 2.604257983719474e-05, "loss": 0.5796, "step": 9427, "task_loss": 0.8872344493865967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.048365831375122, "epoch": 7.97, "learning_rate": 2.6039448966812776e-05, "loss": 0.7284, "step": 9428, "task_loss": 0.9888719320297241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34838905930519104, "epoch": 7.97, "learning_rate": 2.603631809643081e-05, "loss": 0.6091, "step": 9429, "task_loss": 0.42238640785217285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6488430500030518, "epoch": 7.97, "learning_rate": 2.6033187226048843e-05, "loss": 0.5601, "step": 9430, "task_loss": 0.967505156993866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44708067178726196, "epoch": 7.97, "learning_rate": 2.6030056355666878e-05, "loss": 0.7107, "step": 9431, "task_loss": 0.058820124715566635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8617878556251526, "epoch": 7.97, "learning_rate": 2.602692548528491e-05, "loss": 0.6604, "step": 9432, "task_loss": 0.2620382010936737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9152582883834839, "epoch": 7.97, "learning_rate": 2.6023794614902948e-05, "loss": 0.694, "step": 9433, "task_loss": 1.3064684867858887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5423201322555542, "epoch": 7.97, "learning_rate": 2.602066374452098e-05, "loss": 0.5231, "step": 9434, "task_loss": 0.8099385499954224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5063750147819519, "epoch": 7.97, "learning_rate": 2.601753287413901e-05, "loss": 0.7752, "step": 9435, "task_loss": 0.13737323880195618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6110212206840515, "epoch": 7.98, "learning_rate": 2.6014402003757043e-05, "loss": 0.8173, "step": 9436, "task_loss": 0.8579599261283875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8041408061981201, "epoch": 7.98, "learning_rate": 2.6011271133375082e-05, "loss": 0.7977, "step": 9437, "task_loss": 0.9621957540512085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42480766773223877, "epoch": 7.98, "learning_rate": 2.6008140262993114e-05, "loss": 0.584, "step": 9438, "task_loss": 0.6029318571090698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47893577814102173, "epoch": 7.98, "learning_rate": 2.6005009392611146e-05, "loss": 0.6231, "step": 9439, "task_loss": 0.7587630748748779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5024349093437195, "epoch": 7.98, "learning_rate": 2.6001878522229177e-05, "loss": 0.6414, "step": 9440, "task_loss": 0.735466718673706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5382870435714722, "epoch": 7.98, "learning_rate": 2.5998747651847216e-05, "loss": 0.5021, "step": 9441, "task_loss": 0.4793981611728668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5290486812591553, "epoch": 7.98, "learning_rate": 2.5995616781465248e-05, "loss": 0.692, "step": 9442, "task_loss": 0.9503871202468872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5836115479469299, "epoch": 7.98, "learning_rate": 2.599248591108328e-05, "loss": 0.696, "step": 9443, "task_loss": 1.128892421722412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47441017627716064, "epoch": 7.98, "learning_rate": 2.5989355040701318e-05, "loss": 0.5504, "step": 9444, "task_loss": 0.3044808506965637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26005038619041443, "epoch": 7.98, "learning_rate": 2.598622417031935e-05, "loss": 0.6115, "step": 9445, "task_loss": 0.7386947870254517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4791017770767212, "epoch": 7.98, "learning_rate": 2.5983093299937385e-05, "loss": 0.4328, "step": 9446, "task_loss": 1.3789352178573608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.210294246673584, "epoch": 7.99, "learning_rate": 2.5979962429555417e-05, "loss": 0.6871, "step": 9447, "task_loss": 0.7353560924530029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5890647768974304, "epoch": 7.99, "learning_rate": 2.5976831559173452e-05, "loss": 0.8074, "step": 9448, "task_loss": 0.6747661232948303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7244791388511658, "epoch": 7.99, "learning_rate": 2.5973700688791487e-05, "loss": 0.7915, "step": 9449, "task_loss": 1.2220650911331177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4652203321456909, "epoch": 7.99, "learning_rate": 2.597056981840952e-05, "loss": 0.6277, "step": 9450, "task_loss": 0.31935709714889526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4284331798553467, "epoch": 7.99, "learning_rate": 2.596743894802755e-05, "loss": 0.5825, "step": 9451, "task_loss": 1.1488032341003418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6096150279045105, "epoch": 7.99, "learning_rate": 2.596430807764559e-05, "loss": 0.6413, "step": 9452, "task_loss": 0.9601779580116272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4133976697921753, "epoch": 7.99, "learning_rate": 2.596117720726362e-05, "loss": 0.5555, "step": 9453, "task_loss": 0.7433341145515442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44403427839279175, "epoch": 7.99, "learning_rate": 2.5958046336881653e-05, "loss": 0.6117, "step": 9454, "task_loss": 0.7075508236885071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5410842895507812, "epoch": 7.99, "learning_rate": 2.595491546649969e-05, "loss": 0.5694, "step": 9455, "task_loss": 1.1951857805252075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5486364364624023, "epoch": 7.99, "learning_rate": 2.5951784596117723e-05, "loss": 0.6251, "step": 9456, "task_loss": 0.2176392376422882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6250971555709839, "epoch": 7.99, "learning_rate": 2.5948653725735755e-05, "loss": 0.5612, "step": 9457, "task_loss": 0.7093429565429688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.612447202205658, "epoch": 7.99, "learning_rate": 2.5945522855353787e-05, "loss": 0.6523, "step": 9458, "task_loss": 0.3964165449142456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9440664052963257, "epoch": 8.0, "learning_rate": 2.5942391984971825e-05, "loss": 0.6894, "step": 9459, "task_loss": 0.5693852305412292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6208261251449585, "epoch": 8.0, "learning_rate": 2.5939261114589857e-05, "loss": 0.6969, "step": 9460, "task_loss": 0.7996400594711304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0801990032196045, "epoch": 8.0, "learning_rate": 2.593613024420789e-05, "loss": 0.7062, "step": 9461, "task_loss": 1.0872877836227417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8959672451019287, "epoch": 8.0, "learning_rate": 2.5932999373825924e-05, "loss": 0.7018, "step": 9462, "task_loss": 1.5299794673919678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8990058302879333, "epoch": 8.0, "learning_rate": 2.592986850344396e-05, "loss": 0.8051, "step": 9463, "task_loss": 1.4390032291412354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7602263689041138, "epoch": 8.0, "learning_rate": 2.5926737633061994e-05, "loss": 0.4792, "step": 9464, "task_loss": 0.531136691570282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40493297576904297, "epoch": 8.0, "learning_rate": 2.5923606762680026e-05, "loss": 1.3156, "step": 9465, "task_loss": 0.20418496429920197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4805058240890503, "epoch": 8.0, "learning_rate": 2.592047589229806e-05, "loss": 0.5107, "step": 9466, "task_loss": 0.22030767798423767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3272056579589844, "epoch": 8.0, "learning_rate": 2.5917345021916096e-05, "loss": 0.7194, "step": 9467, "task_loss": 0.9305848479270935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7627222537994385, "epoch": 8.0, "learning_rate": 2.5914214151534128e-05, "loss": 0.6304, "step": 9468, "task_loss": 1.5871202945709229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6400711536407471, "epoch": 8.0, "learning_rate": 2.591108328115216e-05, "loss": 0.7058, "step": 9469, "task_loss": 0.38720619678497314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5041638612747192, "epoch": 8.01, "learning_rate": 2.59079524107702e-05, "loss": 0.5197, "step": 9470, "task_loss": 0.7368938326835632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7635452747344971, "epoch": 8.01, "learning_rate": 2.590482154038823e-05, "loss": 0.6637, "step": 9471, "task_loss": 0.7752496004104614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9043078422546387, "epoch": 8.01, "learning_rate": 2.5901690670006262e-05, "loss": 0.8578, "step": 9472, "task_loss": 0.6429308652877808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5908634662628174, "epoch": 8.01, "learning_rate": 2.5898559799624294e-05, "loss": 0.593, "step": 9473, "task_loss": 1.114374041557312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5358446836471558, "epoch": 8.01, "learning_rate": 2.5895428929242332e-05, "loss": 0.754, "step": 9474, "task_loss": 0.5739922523498535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4911760985851288, "epoch": 8.01, "learning_rate": 2.5892298058860364e-05, "loss": 0.5911, "step": 9475, "task_loss": 0.9473483562469482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6584561467170715, "epoch": 8.01, "learning_rate": 2.5889167188478396e-05, "loss": 0.6539, "step": 9476, "task_loss": 1.2181719541549683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5431656837463379, "epoch": 8.01, "learning_rate": 2.5886036318096428e-05, "loss": 0.7327, "step": 9477, "task_loss": 0.286673367023468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0166093111038208, "epoch": 8.01, "learning_rate": 2.5882905447714466e-05, "loss": 0.7161, "step": 9478, "task_loss": 0.5947268605232239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8766754865646362, "epoch": 8.01, "learning_rate": 2.5879774577332498e-05, "loss": 0.6402, "step": 9479, "task_loss": 0.6928406357765198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.554659366607666, "epoch": 8.01, "learning_rate": 2.5876643706950533e-05, "loss": 0.6538, "step": 9480, "task_loss": 0.898950457572937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9658907055854797, "epoch": 8.01, "learning_rate": 2.587351283656857e-05, "loss": 0.6563, "step": 9481, "task_loss": 0.6451122164726257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4442984163761139, "epoch": 8.02, "learning_rate": 2.5870381966186604e-05, "loss": 0.6976, "step": 9482, "task_loss": 0.7318933606147766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0225954055786133, "epoch": 8.02, "learning_rate": 2.5867251095804635e-05, "loss": 0.7959, "step": 9483, "task_loss": 0.299864262342453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.392708420753479, "epoch": 8.02, "learning_rate": 2.5864120225422667e-05, "loss": 0.5378, "step": 9484, "task_loss": 0.3352820873260498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5694526433944702, "epoch": 8.02, "learning_rate": 2.5860989355040706e-05, "loss": 0.5262, "step": 9485, "task_loss": 0.2699090242385864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5054669380187988, "epoch": 8.02, "learning_rate": 2.5857858484658737e-05, "loss": 0.534, "step": 9486, "task_loss": 1.2129404544830322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196722745895386, "epoch": 8.02, "learning_rate": 2.585472761427677e-05, "loss": 0.6249, "step": 9487, "task_loss": 0.42092031240463257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8780670166015625, "epoch": 8.02, "learning_rate": 2.58515967438948e-05, "loss": 0.5763, "step": 9488, "task_loss": 1.023160696029663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0897823572158813, "epoch": 8.02, "learning_rate": 2.584846587351284e-05, "loss": 0.6728, "step": 9489, "task_loss": 2.1009795665740967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.684663712978363, "epoch": 8.02, "learning_rate": 2.584533500313087e-05, "loss": 0.7628, "step": 9490, "task_loss": 1.059114933013916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7650578618049622, "epoch": 8.02, "learning_rate": 2.5842204132748903e-05, "loss": 0.715, "step": 9491, "task_loss": 1.9908535480499268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6482884287834167, "epoch": 8.02, "learning_rate": 2.5839073262366942e-05, "loss": 0.6214, "step": 9492, "task_loss": 1.0943528413772583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.624677300453186, "epoch": 8.02, "learning_rate": 2.5835942391984974e-05, "loss": 0.8581, "step": 9493, "task_loss": 0.8004274964332581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7109591364860535, "epoch": 8.03, "learning_rate": 2.5832811521603005e-05, "loss": 0.6102, "step": 9494, "task_loss": 0.30904605984687805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6413161158561707, "epoch": 8.03, "learning_rate": 2.5829680651221037e-05, "loss": 0.6003, "step": 9495, "task_loss": 0.6954140663146973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6340187191963196, "epoch": 8.03, "learning_rate": 2.5826549780839076e-05, "loss": 0.5339, "step": 9496, "task_loss": 0.5692787170410156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7334662079811096, "epoch": 8.03, "learning_rate": 2.5823418910457107e-05, "loss": 0.6078, "step": 9497, "task_loss": 0.47687798738479614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7268984317779541, "epoch": 8.03, "learning_rate": 2.5820288040075143e-05, "loss": 0.531, "step": 9498, "task_loss": 0.618720293045044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.51483553647995, "epoch": 8.03, "learning_rate": 2.5817157169693174e-05, "loss": 0.4996, "step": 9499, "task_loss": 0.2620408535003662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.535693883895874, "epoch": 8.03, "learning_rate": 2.5814026299311213e-05, "loss": 0.6021, "step": 9500, "task_loss": 1.0126409530639648 }, { "epoch": 8.03, "eval_accuracy": 0.8979009900990099, "eval_loss": 0.4202496111392975, "eval_runtime": 207.1964, "eval_samples_per_second": 121.865, "eval_steps_per_second": 0.956, "step": 9500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48358774185180664, "epoch": 8.03, "learning_rate": 2.5810895428929245e-05, "loss": 0.5775, "step": 9501, "task_loss": 0.57383793592453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6048062443733215, "epoch": 8.03, "learning_rate": 2.5807764558547276e-05, "loss": 0.6382, "step": 9502, "task_loss": 1.2340830564498901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6741732358932495, "epoch": 8.03, "learning_rate": 2.5804633688165315e-05, "loss": 0.5898, "step": 9503, "task_loss": 0.8179440498352051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9456860423088074, "epoch": 8.03, "learning_rate": 2.5801502817783347e-05, "loss": 0.6483, "step": 9504, "task_loss": 1.3663380146026611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5779914259910583, "epoch": 8.03, "learning_rate": 2.579837194740138e-05, "loss": 0.6303, "step": 9505, "task_loss": 0.7185970544815063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7610511779785156, "epoch": 8.04, "learning_rate": 2.579524107701941e-05, "loss": 0.6949, "step": 9506, "task_loss": 0.5917132496833801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6034567356109619, "epoch": 8.04, "learning_rate": 2.579211020663745e-05, "loss": 0.6168, "step": 9507, "task_loss": 0.7493675947189331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8272644281387329, "epoch": 8.04, "learning_rate": 2.578897933625548e-05, "loss": 0.6557, "step": 9508, "task_loss": 1.0381020307540894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5597089529037476, "epoch": 8.04, "learning_rate": 2.5785848465873513e-05, "loss": 0.5871, "step": 9509, "task_loss": 0.19758839905261993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1473991870880127, "epoch": 8.04, "learning_rate": 2.5782717595491544e-05, "loss": 0.6547, "step": 9510, "task_loss": 1.5422964096069336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29744890332221985, "epoch": 8.04, "learning_rate": 2.5779586725109583e-05, "loss": 0.5276, "step": 9511, "task_loss": 0.7286332845687866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27245599031448364, "epoch": 8.04, "learning_rate": 2.5776455854727615e-05, "loss": 0.3935, "step": 9512, "task_loss": 0.45734575390815735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5940107107162476, "epoch": 8.04, "learning_rate": 2.577332498434565e-05, "loss": 0.4837, "step": 9513, "task_loss": 1.3543310165405273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4970851242542267, "epoch": 8.04, "learning_rate": 2.577019411396368e-05, "loss": 0.7043, "step": 9514, "task_loss": 0.6898879408836365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6582046747207642, "epoch": 8.04, "learning_rate": 2.5767063243581717e-05, "loss": 0.6196, "step": 9515, "task_loss": 0.9002025127410889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42652738094329834, "epoch": 8.04, "learning_rate": 2.5763932373199752e-05, "loss": 0.532, "step": 9516, "task_loss": 0.49857544898986816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40560728311538696, "epoch": 8.04, "learning_rate": 2.5760801502817784e-05, "loss": 0.5408, "step": 9517, "task_loss": 0.3382502496242523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9616913795471191, "epoch": 8.05, "learning_rate": 2.5757670632435822e-05, "loss": 0.6746, "step": 9518, "task_loss": 1.2382404804229736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.611966609954834, "epoch": 8.05, "learning_rate": 2.5754539762053854e-05, "loss": 0.6559, "step": 9519, "task_loss": 0.9553315043449402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6029883623123169, "epoch": 8.05, "learning_rate": 2.5751408891671886e-05, "loss": 0.5583, "step": 9520, "task_loss": 0.14849939942359924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8744843006134033, "epoch": 8.05, "learning_rate": 2.5748278021289918e-05, "loss": 0.7241, "step": 9521, "task_loss": 0.43535900115966797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.153090476989746, "epoch": 8.05, "learning_rate": 2.5745147150907956e-05, "loss": 0.9187, "step": 9522, "task_loss": 0.8058032393455505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5794163942337036, "epoch": 8.05, "learning_rate": 2.5742016280525988e-05, "loss": 0.7354, "step": 9523, "task_loss": 0.503695547580719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4858200252056122, "epoch": 8.05, "learning_rate": 2.573888541014402e-05, "loss": 0.8327, "step": 9524, "task_loss": 0.4191473722457886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3775385618209839, "epoch": 8.05, "learning_rate": 2.573575453976205e-05, "loss": 0.5329, "step": 9525, "task_loss": 0.6246687173843384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.712108314037323, "epoch": 8.05, "learning_rate": 2.573262366938009e-05, "loss": 0.6411, "step": 9526, "task_loss": 0.11870142817497253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8974450826644897, "epoch": 8.05, "learning_rate": 2.5729492798998122e-05, "loss": 0.6187, "step": 9527, "task_loss": 1.290686011314392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030809044837952, "epoch": 8.05, "learning_rate": 2.5726361928616154e-05, "loss": 0.4542, "step": 9528, "task_loss": 0.7721290588378906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3780890107154846, "epoch": 8.05, "learning_rate": 2.5723231058234192e-05, "loss": 0.6497, "step": 9529, "task_loss": 0.8706468939781189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5306352376937866, "epoch": 8.06, "learning_rate": 2.5720100187852224e-05, "loss": 0.6317, "step": 9530, "task_loss": 0.4210657775402069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6945241093635559, "epoch": 8.06, "learning_rate": 2.571696931747026e-05, "loss": 0.6628, "step": 9531, "task_loss": 0.4057821035385132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8563560247421265, "epoch": 8.06, "learning_rate": 2.571383844708829e-05, "loss": 0.4995, "step": 9532, "task_loss": 0.3340384364128113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7213001251220703, "epoch": 8.06, "learning_rate": 2.5710707576706326e-05, "loss": 0.6187, "step": 9533, "task_loss": 0.991081953048706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5799235105514526, "epoch": 8.06, "learning_rate": 2.570757670632436e-05, "loss": 0.6332, "step": 9534, "task_loss": 1.1296018362045288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7479032278060913, "epoch": 8.06, "learning_rate": 2.5704445835942393e-05, "loss": 0.5621, "step": 9535, "task_loss": 0.3671419024467468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5578415989875793, "epoch": 8.06, "learning_rate": 2.5701314965560425e-05, "loss": 0.4745, "step": 9536, "task_loss": 0.3837338984012604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5913974642753601, "epoch": 8.06, "learning_rate": 2.5698184095178463e-05, "loss": 0.4401, "step": 9537, "task_loss": 0.7473679184913635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6994251012802124, "epoch": 8.06, "learning_rate": 2.5695053224796495e-05, "loss": 0.4842, "step": 9538, "task_loss": 0.7453098297119141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7720438241958618, "epoch": 8.06, "learning_rate": 2.5691922354414527e-05, "loss": 0.5414, "step": 9539, "task_loss": 1.5022172927856445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3230380117893219, "epoch": 8.06, "learning_rate": 2.5688791484032565e-05, "loss": 0.628, "step": 9540, "task_loss": 0.11658510565757751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7793655395507812, "epoch": 8.07, "learning_rate": 2.5685660613650597e-05, "loss": 0.7186, "step": 9541, "task_loss": 0.6225934028625488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5107183456420898, "epoch": 8.07, "learning_rate": 2.568252974326863e-05, "loss": 0.7468, "step": 9542, "task_loss": 1.1425209045410156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7070357203483582, "epoch": 8.07, "learning_rate": 2.567939887288666e-05, "loss": 0.7377, "step": 9543, "task_loss": 0.5094785094261169 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6440021991729736, "epoch": 8.07, "learning_rate": 2.56762680025047e-05, "loss": 0.6431, "step": 9544, "task_loss": 0.5144655108451843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6095285415649414, "epoch": 8.07, "learning_rate": 2.567313713212273e-05, "loss": 0.5927, "step": 9545, "task_loss": 0.550056517124176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9433820247650146, "epoch": 8.07, "learning_rate": 2.5670006261740763e-05, "loss": 0.7019, "step": 9546, "task_loss": 1.3092015981674194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5735055804252625, "epoch": 8.07, "learning_rate": 2.5666875391358798e-05, "loss": 0.7936, "step": 9547, "task_loss": 0.6648797988891602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6498998999595642, "epoch": 8.07, "learning_rate": 2.5663744520976833e-05, "loss": 0.7473, "step": 9548, "task_loss": 0.8122998476028442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5301042199134827, "epoch": 8.07, "learning_rate": 2.566061365059487e-05, "loss": 0.6938, "step": 9549, "task_loss": 1.1722551584243774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6366837620735168, "epoch": 8.07, "learning_rate": 2.56574827802129e-05, "loss": 0.6718, "step": 9550, "task_loss": 0.13064587116241455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6808183789253235, "epoch": 8.07, "learning_rate": 2.5654351909830932e-05, "loss": 0.6107, "step": 9551, "task_loss": 0.6895596981048584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4425044655799866, "epoch": 8.07, "learning_rate": 2.565122103944897e-05, "loss": 0.365, "step": 9552, "task_loss": 0.7876421213150024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6202681064605713, "epoch": 8.08, "learning_rate": 2.5648090169067002e-05, "loss": 0.6366, "step": 9553, "task_loss": 0.3348346948623657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7779496908187866, "epoch": 8.08, "learning_rate": 2.5644959298685034e-05, "loss": 0.7076, "step": 9554, "task_loss": 0.7305543422698975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8965659141540527, "epoch": 8.08, "learning_rate": 2.5641828428303073e-05, "loss": 0.7051, "step": 9555, "task_loss": 0.9974831342697144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7835591435432434, "epoch": 8.08, "learning_rate": 2.5638697557921104e-05, "loss": 0.5683, "step": 9556, "task_loss": 0.5769684314727783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5753788948059082, "epoch": 8.08, "learning_rate": 2.5635566687539136e-05, "loss": 0.5816, "step": 9557, "task_loss": 0.4654757082462311 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35605379939079285, "epoch": 8.08, "learning_rate": 2.5632435817157168e-05, "loss": 0.5144, "step": 9558, "task_loss": 0.31107622385025024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5308482050895691, "epoch": 8.08, "learning_rate": 2.5629304946775207e-05, "loss": 0.6899, "step": 9559, "task_loss": 0.44707223773002625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5910089612007141, "epoch": 8.08, "learning_rate": 2.562617407639324e-05, "loss": 0.6585, "step": 9560, "task_loss": 0.7067992687225342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7970684170722961, "epoch": 8.08, "learning_rate": 2.562304320601127e-05, "loss": 0.5841, "step": 9561, "task_loss": 0.9572331309318542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6173934936523438, "epoch": 8.08, "learning_rate": 2.5619912335629302e-05, "loss": 0.6534, "step": 9562, "task_loss": 0.29756224155426025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4951367974281311, "epoch": 8.08, "learning_rate": 2.561678146524734e-05, "loss": 0.7064, "step": 9563, "task_loss": 1.0754907131195068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6889551877975464, "epoch": 8.08, "learning_rate": 2.5613650594865372e-05, "loss": 0.6638, "step": 9564, "task_loss": 0.9315205216407776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6017736792564392, "epoch": 8.09, "learning_rate": 2.5610519724483407e-05, "loss": 0.7311, "step": 9565, "task_loss": 0.6515302658081055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5555147528648376, "epoch": 8.09, "learning_rate": 2.5607388854101443e-05, "loss": 0.5795, "step": 9566, "task_loss": 0.5221213102340698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8576714992523193, "epoch": 8.09, "learning_rate": 2.5604257983719478e-05, "loss": 0.7232, "step": 9567, "task_loss": 1.3987363576889038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0119340419769287, "epoch": 8.09, "learning_rate": 2.560112711333751e-05, "loss": 0.6977, "step": 9568, "task_loss": 1.4994897842407227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1301989555358887, "epoch": 8.09, "learning_rate": 2.559799624295554e-05, "loss": 0.8764, "step": 9569, "task_loss": 1.321608543395996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6715595126152039, "epoch": 8.09, "learning_rate": 2.559486537257358e-05, "loss": 0.7286, "step": 9570, "task_loss": 1.4764540195465088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9168463945388794, "epoch": 8.09, "learning_rate": 2.559173450219161e-05, "loss": 0.6218, "step": 9571, "task_loss": 1.6275674104690552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5858972668647766, "epoch": 8.09, "learning_rate": 2.5588603631809643e-05, "loss": 0.6429, "step": 9572, "task_loss": 0.6500300765037537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8399564027786255, "epoch": 8.09, "learning_rate": 2.5585472761427675e-05, "loss": 0.4252, "step": 9573, "task_loss": 1.1648294925689697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5321627855300903, "epoch": 8.09, "learning_rate": 2.5582341891045714e-05, "loss": 0.6886, "step": 9574, "task_loss": 0.36192697286605835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0696601867675781, "epoch": 8.09, "learning_rate": 2.5579211020663746e-05, "loss": 0.8208, "step": 9575, "task_loss": 0.7199418544769287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7306951284408569, "epoch": 8.09, "learning_rate": 2.5576080150281777e-05, "loss": 0.6862, "step": 9576, "task_loss": 0.7682673931121826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.75617915391922, "epoch": 8.1, "learning_rate": 2.5572949279899816e-05, "loss": 0.5953, "step": 9577, "task_loss": 0.838303804397583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7792779207229614, "epoch": 8.1, "learning_rate": 2.5569818409517848e-05, "loss": 0.5689, "step": 9578, "task_loss": 1.0271424055099487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8277554512023926, "epoch": 8.1, "learning_rate": 2.556668753913588e-05, "loss": 0.758, "step": 9579, "task_loss": 1.0490671396255493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9833515286445618, "epoch": 8.1, "learning_rate": 2.5563556668753915e-05, "loss": 0.6554, "step": 9580, "task_loss": 0.6816461682319641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7996233105659485, "epoch": 8.1, "learning_rate": 2.556042579837195e-05, "loss": 0.5705, "step": 9581, "task_loss": 1.0755360126495361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31369146704673767, "epoch": 8.1, "learning_rate": 2.555729492798998e-05, "loss": 0.5092, "step": 9582, "task_loss": 0.6770860552787781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.578947901725769, "epoch": 8.1, "learning_rate": 2.5554164057608017e-05, "loss": 0.5972, "step": 9583, "task_loss": 0.9392202496528625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5867350101470947, "epoch": 8.1, "learning_rate": 2.555103318722605e-05, "loss": 0.6754, "step": 9584, "task_loss": 0.4179498553276062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.245731920003891, "epoch": 8.1, "learning_rate": 2.5547902316844087e-05, "loss": 0.5912, "step": 9585, "task_loss": 0.4553441107273102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46888723969459534, "epoch": 8.1, "learning_rate": 2.554477144646212e-05, "loss": 0.6222, "step": 9586, "task_loss": 1.2910476922988892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.668941855430603, "epoch": 8.1, "learning_rate": 2.554164057608015e-05, "loss": 0.554, "step": 9587, "task_loss": 1.0155181884765625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4167858362197876, "epoch": 8.1, "learning_rate": 2.5538509705698182e-05, "loss": 0.8064, "step": 9588, "task_loss": 1.1315107345581055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45227572321891785, "epoch": 8.11, "learning_rate": 2.553537883531622e-05, "loss": 0.6359, "step": 9589, "task_loss": 0.16220325231552124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7347128391265869, "epoch": 8.11, "learning_rate": 2.5532247964934253e-05, "loss": 0.5774, "step": 9590, "task_loss": 0.6392708420753479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5246169567108154, "epoch": 8.11, "learning_rate": 2.5529117094552285e-05, "loss": 0.7192, "step": 9591, "task_loss": 0.657183825969696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8130521774291992, "epoch": 8.11, "learning_rate": 2.5525986224170323e-05, "loss": 0.6726, "step": 9592, "task_loss": 1.514998435974121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8547704219818115, "epoch": 8.11, "learning_rate": 2.5522855353788355e-05, "loss": 0.5982, "step": 9593, "task_loss": 0.8979197144508362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46386539936065674, "epoch": 8.11, "learning_rate": 2.5519724483406387e-05, "loss": 0.6414, "step": 9594, "task_loss": 0.4809092581272125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8833751678466797, "epoch": 8.11, "learning_rate": 2.551659361302442e-05, "loss": 0.7629, "step": 9595, "task_loss": 1.017622470855713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7868577241897583, "epoch": 8.11, "learning_rate": 2.5513462742642457e-05, "loss": 0.6774, "step": 9596, "task_loss": 0.35974374413490295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4300953149795532, "epoch": 8.11, "learning_rate": 2.551033187226049e-05, "loss": 0.4592, "step": 9597, "task_loss": 0.4454633295536041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45883291959762573, "epoch": 8.11, "learning_rate": 2.5507201001878524e-05, "loss": 0.5351, "step": 9598, "task_loss": 0.23750022053718567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9552971124649048, "epoch": 8.11, "learning_rate": 2.5504070131496556e-05, "loss": 0.7307, "step": 9599, "task_loss": 0.2768580913543701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29829153418540955, "epoch": 8.11, "learning_rate": 2.550093926111459e-05, "loss": 0.4871, "step": 9600, "task_loss": 0.8613899946212769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6438475847244263, "epoch": 8.12, "learning_rate": 2.5497808390732626e-05, "loss": 0.6919, "step": 9601, "task_loss": 1.2424225807189941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5164478421211243, "epoch": 8.12, "learning_rate": 2.5494677520350658e-05, "loss": 0.5019, "step": 9602, "task_loss": 0.7268494367599487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7925105094909668, "epoch": 8.12, "learning_rate": 2.5491546649968696e-05, "loss": 0.9615, "step": 9603, "task_loss": 1.6412252187728882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5080054998397827, "epoch": 8.12, "learning_rate": 2.5488415779586728e-05, "loss": 0.4558, "step": 9604, "task_loss": 0.19533167779445648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7837927341461182, "epoch": 8.12, "learning_rate": 2.548528490920476e-05, "loss": 0.6181, "step": 9605, "task_loss": 0.8471038341522217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.571692705154419, "epoch": 8.12, "learning_rate": 2.5482154038822792e-05, "loss": 0.5855, "step": 9606, "task_loss": 0.39977994561195374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4399479627609253, "epoch": 8.12, "learning_rate": 2.547902316844083e-05, "loss": 0.563, "step": 9607, "task_loss": 0.6484033465385437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8446363210678101, "epoch": 8.12, "learning_rate": 2.5475892298058862e-05, "loss": 0.6362, "step": 9608, "task_loss": 0.8657239675521851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4532686471939087, "epoch": 8.12, "learning_rate": 2.5472761427676894e-05, "loss": 0.5774, "step": 9609, "task_loss": 0.11240512877702713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7023826241493225, "epoch": 8.12, "learning_rate": 2.5469630557294926e-05, "loss": 0.5309, "step": 9610, "task_loss": 1.5416871309280396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7240844964981079, "epoch": 8.12, "learning_rate": 2.5466499686912964e-05, "loss": 0.5956, "step": 9611, "task_loss": 0.6635152101516724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5511894822120667, "epoch": 8.13, "learning_rate": 2.5463368816530996e-05, "loss": 0.5923, "step": 9612, "task_loss": 0.16926248371601105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7315818071365356, "epoch": 8.13, "learning_rate": 2.5460237946149028e-05, "loss": 0.7396, "step": 9613, "task_loss": 0.8151178956031799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4139784872531891, "epoch": 8.13, "learning_rate": 2.5457107075767066e-05, "loss": 0.5352, "step": 9614, "task_loss": 0.0864553451538086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9371407628059387, "epoch": 8.13, "learning_rate": 2.5453976205385098e-05, "loss": 0.7291, "step": 9615, "task_loss": 0.6554353833198547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49250197410583496, "epoch": 8.13, "learning_rate": 2.5450845335003133e-05, "loss": 0.4657, "step": 9616, "task_loss": 1.225512981414795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8915331363677979, "epoch": 8.13, "learning_rate": 2.5447714464621165e-05, "loss": 0.7989, "step": 9617, "task_loss": 0.7773279547691345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4731777012348175, "epoch": 8.13, "learning_rate": 2.54445835942392e-05, "loss": 0.5119, "step": 9618, "task_loss": 0.1723177433013916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6641817092895508, "epoch": 8.13, "learning_rate": 2.5441452723857235e-05, "loss": 0.7718, "step": 9619, "task_loss": 1.086281180381775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4389575719833374, "epoch": 8.13, "learning_rate": 2.5438321853475267e-05, "loss": 0.4604, "step": 9620, "task_loss": 0.9568436145782471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46343883872032166, "epoch": 8.13, "learning_rate": 2.54351909830933e-05, "loss": 0.6942, "step": 9621, "task_loss": 1.0595747232437134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4939122796058655, "epoch": 8.13, "learning_rate": 2.5432060112711338e-05, "loss": 0.6796, "step": 9622, "task_loss": 0.7629585862159729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3883133828639984, "epoch": 8.13, "learning_rate": 2.542892924232937e-05, "loss": 0.5526, "step": 9623, "task_loss": 0.41494420170783997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7220919728279114, "epoch": 8.14, "learning_rate": 2.54257983719474e-05, "loss": 0.5812, "step": 9624, "task_loss": 1.2785509824752808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7011469006538391, "epoch": 8.14, "learning_rate": 2.5422667501565433e-05, "loss": 0.6938, "step": 9625, "task_loss": 0.27214160561561584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5268486738204956, "epoch": 8.14, "learning_rate": 2.541953663118347e-05, "loss": 0.488, "step": 9626, "task_loss": 0.09654208272695541 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7060502171516418, "epoch": 8.14, "learning_rate": 2.5416405760801503e-05, "loss": 0.718, "step": 9627, "task_loss": 1.0160988569259644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24678410589694977, "epoch": 8.14, "learning_rate": 2.5413274890419535e-05, "loss": 0.6448, "step": 9628, "task_loss": 0.032727278769016266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0448825359344482, "epoch": 8.14, "learning_rate": 2.5410144020037574e-05, "loss": 0.7907, "step": 9629, "task_loss": 1.302258014678955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4778451919555664, "epoch": 8.14, "learning_rate": 2.5407013149655605e-05, "loss": 0.5581, "step": 9630, "task_loss": 0.6130968928337097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41670066118240356, "epoch": 8.14, "learning_rate": 2.5403882279273637e-05, "loss": 0.4744, "step": 9631, "task_loss": 0.5596149563789368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5824335813522339, "epoch": 8.14, "learning_rate": 2.5400751408891672e-05, "loss": 0.6759, "step": 9632, "task_loss": 0.46603837609291077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2820415794849396, "epoch": 8.14, "learning_rate": 2.5397620538509707e-05, "loss": 0.6071, "step": 9633, "task_loss": 0.24913251399993896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5396164059638977, "epoch": 8.14, "learning_rate": 2.5394489668127743e-05, "loss": 0.8013, "step": 9634, "task_loss": 0.9894593358039856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.719986617565155, "epoch": 8.14, "learning_rate": 2.5391358797745774e-05, "loss": 0.6673, "step": 9635, "task_loss": 1.042793869972229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8575578927993774, "epoch": 8.15, "learning_rate": 2.5388227927363806e-05, "loss": 0.7707, "step": 9636, "task_loss": 0.5016798973083496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9153276681900024, "epoch": 8.15, "learning_rate": 2.5385097056981845e-05, "loss": 0.8139, "step": 9637, "task_loss": 1.3980708122253418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6774334907531738, "epoch": 8.15, "learning_rate": 2.5381966186599877e-05, "loss": 0.6619, "step": 9638, "task_loss": 0.6898733377456665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3844747841358185, "epoch": 8.15, "learning_rate": 2.537883531621791e-05, "loss": 0.5771, "step": 9639, "task_loss": 1.3701329231262207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6570324897766113, "epoch": 8.15, "learning_rate": 2.5375704445835947e-05, "loss": 0.8157, "step": 9640, "task_loss": 0.5638008713722229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4445846974849701, "epoch": 8.15, "learning_rate": 2.537257357545398e-05, "loss": 0.6955, "step": 9641, "task_loss": 0.7954586148262024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4245174527168274, "epoch": 8.15, "learning_rate": 2.536944270507201e-05, "loss": 0.5999, "step": 9642, "task_loss": 0.5617968440055847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46280866861343384, "epoch": 8.15, "learning_rate": 2.5366311834690042e-05, "loss": 0.5405, "step": 9643, "task_loss": 1.4387035369873047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6392872333526611, "epoch": 8.15, "learning_rate": 2.536318096430808e-05, "loss": 0.4903, "step": 9644, "task_loss": 0.7769944071769714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5416843891143799, "epoch": 8.15, "learning_rate": 2.5360050093926113e-05, "loss": 0.5815, "step": 9645, "task_loss": 0.3541070520877838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5261127948760986, "epoch": 8.15, "learning_rate": 2.5356919223544144e-05, "loss": 0.5105, "step": 9646, "task_loss": 0.2480679303407669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.589089035987854, "epoch": 8.15, "learning_rate": 2.5353788353162176e-05, "loss": 0.7432, "step": 9647, "task_loss": 0.594060480594635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4982929527759552, "epoch": 8.16, "learning_rate": 2.5350657482780215e-05, "loss": 0.5632, "step": 9648, "task_loss": 0.9999788999557495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5776656866073608, "epoch": 8.16, "learning_rate": 2.5347526612398246e-05, "loss": 0.6447, "step": 9649, "task_loss": 0.06691469997167587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4507962167263031, "epoch": 8.16, "learning_rate": 2.534439574201628e-05, "loss": 0.4398, "step": 9650, "task_loss": 0.43898165225982666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.336259663105011, "epoch": 8.16, "learning_rate": 2.5341264871634317e-05, "loss": 0.5993, "step": 9651, "task_loss": 1.0443663597106934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9344462752342224, "epoch": 8.16, "learning_rate": 2.5338134001252352e-05, "loss": 0.7764, "step": 9652, "task_loss": 0.7063491940498352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8438107967376709, "epoch": 8.16, "learning_rate": 2.5335003130870384e-05, "loss": 0.6762, "step": 9653, "task_loss": 1.0690546035766602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9633786678314209, "epoch": 8.16, "learning_rate": 2.5331872260488416e-05, "loss": 0.7436, "step": 9654, "task_loss": 1.4770820140838623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5607155561447144, "epoch": 8.16, "learning_rate": 2.5328741390106454e-05, "loss": 0.7628, "step": 9655, "task_loss": 0.7228819131851196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0833505392074585, "epoch": 8.16, "learning_rate": 2.5325610519724486e-05, "loss": 0.767, "step": 9656, "task_loss": 0.8650422692298889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3220139443874359, "epoch": 8.16, "learning_rate": 2.5322479649342518e-05, "loss": 0.4939, "step": 9657, "task_loss": 0.6746513247489929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4000473916530609, "epoch": 8.16, "learning_rate": 2.531934877896055e-05, "loss": 0.5985, "step": 9658, "task_loss": 0.13283102214336395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5129899978637695, "epoch": 8.16, "learning_rate": 2.5316217908578588e-05, "loss": 0.7289, "step": 9659, "task_loss": 0.5478000044822693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5531086325645447, "epoch": 8.17, "learning_rate": 2.531308703819662e-05, "loss": 0.6924, "step": 9660, "task_loss": 1.520575761795044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7721791863441467, "epoch": 8.17, "learning_rate": 2.530995616781465e-05, "loss": 0.7424, "step": 9661, "task_loss": 1.375849723815918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6192275285720825, "epoch": 8.17, "learning_rate": 2.530682529743269e-05, "loss": 0.5168, "step": 9662, "task_loss": 0.7607097029685974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2975720167160034, "epoch": 8.17, "learning_rate": 2.5303694427050722e-05, "loss": 0.459, "step": 9663, "task_loss": 0.06372404098510742 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5819917917251587, "epoch": 8.17, "learning_rate": 2.5300563556668754e-05, "loss": 0.619, "step": 9664, "task_loss": 1.4437073469161987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6770885586738586, "epoch": 8.17, "learning_rate": 2.529743268628679e-05, "loss": 0.7613, "step": 9665, "task_loss": 1.2764095067977905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5193214416503906, "epoch": 8.17, "learning_rate": 2.5294301815904824e-05, "loss": 0.6322, "step": 9666, "task_loss": 0.9418734312057495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7502713203430176, "epoch": 8.17, "learning_rate": 2.5291170945522856e-05, "loss": 0.5744, "step": 9667, "task_loss": 0.8527737259864807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35305190086364746, "epoch": 8.17, "learning_rate": 2.528804007514089e-05, "loss": 0.711, "step": 9668, "task_loss": 0.2415962666273117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6614627838134766, "epoch": 8.17, "learning_rate": 2.5284909204758923e-05, "loss": 0.9664, "step": 9669, "task_loss": 0.5813974142074585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38990408182144165, "epoch": 8.17, "learning_rate": 2.528177833437696e-05, "loss": 0.5806, "step": 9670, "task_loss": 0.9137741923332214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5822473168373108, "epoch": 8.17, "learning_rate": 2.5278647463994993e-05, "loss": 0.7463, "step": 9671, "task_loss": 0.5298535227775574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.034416913986206, "epoch": 8.18, "learning_rate": 2.5275516593613025e-05, "loss": 0.7411, "step": 9672, "task_loss": 0.9809331893920898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5049123764038086, "epoch": 8.18, "learning_rate": 2.5272385723231057e-05, "loss": 0.485, "step": 9673, "task_loss": 0.5540680289268494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8048176169395447, "epoch": 8.18, "learning_rate": 2.5269254852849095e-05, "loss": 0.5597, "step": 9674, "task_loss": 0.5085134506225586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030750036239624, "epoch": 8.18, "learning_rate": 2.5266123982467127e-05, "loss": 0.4319, "step": 9675, "task_loss": 0.9866622090339661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7414937615394592, "epoch": 8.18, "learning_rate": 2.526299311208516e-05, "loss": 0.7626, "step": 9676, "task_loss": 1.0730187892913818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6891570091247559, "epoch": 8.18, "learning_rate": 2.5259862241703197e-05, "loss": 0.5075, "step": 9677, "task_loss": 0.6283197402954102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4803940951824188, "epoch": 8.18, "learning_rate": 2.525673137132123e-05, "loss": 0.6871, "step": 9678, "task_loss": 0.7552738785743713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3502376079559326, "epoch": 8.18, "learning_rate": 2.525360050093926e-05, "loss": 0.5182, "step": 9679, "task_loss": 0.23687584698200226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9172979593276978, "epoch": 8.18, "learning_rate": 2.5250469630557293e-05, "loss": 0.5891, "step": 9680, "task_loss": 0.2064061462879181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0963844060897827, "epoch": 8.18, "learning_rate": 2.524733876017533e-05, "loss": 0.7393, "step": 9681, "task_loss": 0.9356691837310791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42726731300354004, "epoch": 8.18, "learning_rate": 2.5244207889793363e-05, "loss": 0.724, "step": 9682, "task_loss": 0.2098228931427002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5561780333518982, "epoch": 8.19, "learning_rate": 2.5241077019411398e-05, "loss": 0.6201, "step": 9683, "task_loss": 0.41062262654304504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5559496283531189, "epoch": 8.19, "learning_rate": 2.523794614902943e-05, "loss": 0.5537, "step": 9684, "task_loss": 0.47007325291633606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.624455988407135, "epoch": 8.19, "learning_rate": 2.5234815278647465e-05, "loss": 0.4911, "step": 9685, "task_loss": 1.0104092359542847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3517225384712219, "epoch": 8.19, "learning_rate": 2.52316844082655e-05, "loss": 0.4943, "step": 9686, "task_loss": 0.7324825525283813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8056408166885376, "epoch": 8.19, "learning_rate": 2.5228553537883532e-05, "loss": 0.6706, "step": 9687, "task_loss": 1.4773486852645874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38224703073501587, "epoch": 8.19, "learning_rate": 2.522542266750157e-05, "loss": 0.7189, "step": 9688, "task_loss": 0.8729695081710815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7384762167930603, "epoch": 8.19, "learning_rate": 2.5222291797119602e-05, "loss": 0.8112, "step": 9689, "task_loss": 0.633073627948761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4297798275947571, "epoch": 8.19, "learning_rate": 2.5219160926737634e-05, "loss": 0.6124, "step": 9690, "task_loss": 0.4007762372493744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9344835877418518, "epoch": 8.19, "learning_rate": 2.5216030056355666e-05, "loss": 0.7457, "step": 9691, "task_loss": 1.1234118938446045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7709020376205444, "epoch": 8.19, "learning_rate": 2.5212899185973705e-05, "loss": 0.8252, "step": 9692, "task_loss": 0.8488457798957825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30341804027557373, "epoch": 8.19, "learning_rate": 2.5209768315591736e-05, "loss": 0.5516, "step": 9693, "task_loss": 0.17449010908603668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6301251649856567, "epoch": 8.19, "learning_rate": 2.5206637445209768e-05, "loss": 0.4833, "step": 9694, "task_loss": 0.4440884590148926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6665301322937012, "epoch": 8.2, "learning_rate": 2.52035065748278e-05, "loss": 0.5638, "step": 9695, "task_loss": 1.7943780422210693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.831689715385437, "epoch": 8.2, "learning_rate": 2.520037570444584e-05, "loss": 0.7506, "step": 9696, "task_loss": 0.907779335975647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3005545735359192, "epoch": 8.2, "learning_rate": 2.519724483406387e-05, "loss": 0.5586, "step": 9697, "task_loss": 0.2898761034011841 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3387804627418518, "epoch": 8.2, "learning_rate": 2.5194113963681902e-05, "loss": 0.6652, "step": 9698, "task_loss": 0.5504265427589417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3477489650249481, "epoch": 8.2, "learning_rate": 2.519098309329994e-05, "loss": 0.5171, "step": 9699, "task_loss": 0.6322521567344666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.654631495475769, "epoch": 8.2, "learning_rate": 2.5187852222917972e-05, "loss": 0.504, "step": 9700, "task_loss": 0.5685202479362488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5869540572166443, "epoch": 8.2, "learning_rate": 2.5184721352536008e-05, "loss": 0.6, "step": 9701, "task_loss": 0.5295785665512085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5934743285179138, "epoch": 8.2, "learning_rate": 2.518159048215404e-05, "loss": 0.5002, "step": 9702, "task_loss": 0.6681454181671143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6447452306747437, "epoch": 8.2, "learning_rate": 2.5178459611772074e-05, "loss": 0.5964, "step": 9703, "task_loss": 0.9834814071655273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6782992482185364, "epoch": 8.2, "learning_rate": 2.517532874139011e-05, "loss": 0.532, "step": 9704, "task_loss": 0.9407495260238647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5968315601348877, "epoch": 8.2, "learning_rate": 2.517219787100814e-05, "loss": 0.7341, "step": 9705, "task_loss": 0.9390581846237183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3887748122215271, "epoch": 8.2, "learning_rate": 2.5169067000626173e-05, "loss": 0.6567, "step": 9706, "task_loss": 1.1213641166687012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2977091372013092, "epoch": 8.21, "learning_rate": 2.5165936130244212e-05, "loss": 0.431, "step": 9707, "task_loss": 0.48944926261901855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5384232401847839, "epoch": 8.21, "learning_rate": 2.5162805259862244e-05, "loss": 0.5817, "step": 9708, "task_loss": 0.2208660989999771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5890361070632935, "epoch": 8.21, "learning_rate": 2.5159674389480275e-05, "loss": 0.704, "step": 9709, "task_loss": 0.5923968553543091 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21685370802879333, "epoch": 8.21, "learning_rate": 2.5156543519098307e-05, "loss": 0.5748, "step": 9710, "task_loss": 0.10080135613679886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9523053169250488, "epoch": 8.21, "learning_rate": 2.5153412648716346e-05, "loss": 0.7474, "step": 9711, "task_loss": 0.8837370276451111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4278120994567871, "epoch": 8.21, "learning_rate": 2.5150281778334377e-05, "loss": 0.4596, "step": 9712, "task_loss": 0.459662526845932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.551475465297699, "epoch": 8.21, "learning_rate": 2.514715090795241e-05, "loss": 0.6275, "step": 9713, "task_loss": 0.5588355660438538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3735628128051758, "epoch": 8.21, "learning_rate": 2.5144020037570448e-05, "loss": 0.5977, "step": 9714, "task_loss": 0.45972955226898193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.525063693523407, "epoch": 8.21, "learning_rate": 2.514088916718848e-05, "loss": 0.5396, "step": 9715, "task_loss": 0.59718918800354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25382721424102783, "epoch": 8.21, "learning_rate": 2.513775829680651e-05, "loss": 0.4196, "step": 9716, "task_loss": 0.298596054315567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6730455756187439, "epoch": 8.21, "learning_rate": 2.5134627426424547e-05, "loss": 0.6617, "step": 9717, "task_loss": 0.8860340714454651 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6133168935775757, "epoch": 8.21, "learning_rate": 2.513149655604258e-05, "loss": 0.6371, "step": 9718, "task_loss": 0.95582515001297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3021832704544067, "epoch": 8.22, "learning_rate": 2.5128365685660617e-05, "loss": 0.7068, "step": 9719, "task_loss": 1.3813908100128174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6871944665908813, "epoch": 8.22, "learning_rate": 2.512523481527865e-05, "loss": 0.5303, "step": 9720, "task_loss": 0.411578506231308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5037197470664978, "epoch": 8.22, "learning_rate": 2.512210394489668e-05, "loss": 0.5508, "step": 9721, "task_loss": 1.258310317993164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5444134473800659, "epoch": 8.22, "learning_rate": 2.511897307451472e-05, "loss": 0.4807, "step": 9722, "task_loss": 0.6924225091934204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9285165071487427, "epoch": 8.22, "learning_rate": 2.511584220413275e-05, "loss": 0.7159, "step": 9723, "task_loss": 1.530746579170227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7767833471298218, "epoch": 8.22, "learning_rate": 2.5112711333750783e-05, "loss": 0.6384, "step": 9724, "task_loss": 0.6033667325973511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.780463695526123, "epoch": 8.22, "learning_rate": 2.510958046336882e-05, "loss": 0.6029, "step": 9725, "task_loss": 0.4881612956523895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6265745162963867, "epoch": 8.22, "learning_rate": 2.5106449592986853e-05, "loss": 0.4402, "step": 9726, "task_loss": 0.24621140956878662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6851193904876709, "epoch": 8.22, "learning_rate": 2.5103318722604885e-05, "loss": 0.4908, "step": 9727, "task_loss": 1.1177592277526855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.695264458656311, "epoch": 8.22, "learning_rate": 2.5100187852222916e-05, "loss": 0.7591, "step": 9728, "task_loss": 1.2605290412902832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8009285926818848, "epoch": 8.22, "learning_rate": 2.5097056981840955e-05, "loss": 0.6929, "step": 9729, "task_loss": 0.5387163162231445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2753615081310272, "epoch": 8.22, "learning_rate": 2.5093926111458987e-05, "loss": 0.3889, "step": 9730, "task_loss": 0.661942720413208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37279006838798523, "epoch": 8.23, "learning_rate": 2.509079524107702e-05, "loss": 0.5228, "step": 9731, "task_loss": 0.3500799238681793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36901283264160156, "epoch": 8.23, "learning_rate": 2.5087664370695054e-05, "loss": 0.5869, "step": 9732, "task_loss": 0.5313740372657776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6581944227218628, "epoch": 8.23, "learning_rate": 2.508453350031309e-05, "loss": 0.6964, "step": 9733, "task_loss": 0.6531825661659241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7060216069221497, "epoch": 8.23, "learning_rate": 2.508140262993112e-05, "loss": 0.6405, "step": 9734, "task_loss": 0.8174428343772888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41080838441848755, "epoch": 8.23, "learning_rate": 2.5078271759549156e-05, "loss": 0.5928, "step": 9735, "task_loss": 0.1935158520936966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6998468041419983, "epoch": 8.23, "learning_rate": 2.507514088916719e-05, "loss": 0.8209, "step": 9736, "task_loss": 1.4009943008422852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4285593032836914, "epoch": 8.23, "learning_rate": 2.5072010018785226e-05, "loss": 0.6817, "step": 9737, "task_loss": 0.4511098265647888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40663355588912964, "epoch": 8.23, "learning_rate": 2.5068879148403258e-05, "loss": 0.4785, "step": 9738, "task_loss": 1.1511794328689575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5158841609954834, "epoch": 8.23, "learning_rate": 2.506574827802129e-05, "loss": 0.4908, "step": 9739, "task_loss": 1.5321252346038818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9355796575546265, "epoch": 8.23, "learning_rate": 2.5062617407639328e-05, "loss": 0.7805, "step": 9740, "task_loss": 1.3391231298446655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5939784646034241, "epoch": 8.23, "learning_rate": 2.505948653725736e-05, "loss": 0.6694, "step": 9741, "task_loss": 0.6121826767921448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6689040660858154, "epoch": 8.23, "learning_rate": 2.5056355666875392e-05, "loss": 0.5078, "step": 9742, "task_loss": 0.9044114351272583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7162946462631226, "epoch": 8.24, "learning_rate": 2.5053224796493424e-05, "loss": 0.4995, "step": 9743, "task_loss": 0.9830955266952515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5199261903762817, "epoch": 8.24, "learning_rate": 2.5050093926111462e-05, "loss": 0.6114, "step": 9744, "task_loss": 0.41753652691841125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45577213168144226, "epoch": 8.24, "learning_rate": 2.5046963055729494e-05, "loss": 0.4151, "step": 9745, "task_loss": 0.9322614669799805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5690054297447205, "epoch": 8.24, "learning_rate": 2.5043832185347526e-05, "loss": 0.6482, "step": 9746, "task_loss": 0.6894667744636536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8027087450027466, "epoch": 8.24, "learning_rate": 2.5040701314965558e-05, "loss": 0.6203, "step": 9747, "task_loss": 1.0763559341430664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5643907785415649, "epoch": 8.24, "learning_rate": 2.5037570444583596e-05, "loss": 0.542, "step": 9748, "task_loss": 1.3799687623977661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39440640807151794, "epoch": 8.24, "learning_rate": 2.5034439574201628e-05, "loss": 0.5557, "step": 9749, "task_loss": 0.6879622936248779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5068648457527161, "epoch": 8.24, "learning_rate": 2.5031308703819663e-05, "loss": 0.4259, "step": 9750, "task_loss": 0.5996882915496826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8313378095626831, "epoch": 8.24, "learning_rate": 2.5028177833437698e-05, "loss": 0.7052, "step": 9751, "task_loss": 1.4271879196166992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47955000400543213, "epoch": 8.24, "learning_rate": 2.502504696305573e-05, "loss": 0.6307, "step": 9752, "task_loss": 0.6445948481559753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8037792444229126, "epoch": 8.24, "learning_rate": 2.5021916092673765e-05, "loss": 0.6165, "step": 9753, "task_loss": 0.5567881464958191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3897264003753662, "epoch": 8.24, "learning_rate": 2.5018785222291797e-05, "loss": 0.4543, "step": 9754, "task_loss": 0.35491886734962463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3952568769454956, "epoch": 8.25, "learning_rate": 2.5015654351909836e-05, "loss": 0.4875, "step": 9755, "task_loss": 1.1545878648757935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5054922103881836, "epoch": 8.25, "learning_rate": 2.5012523481527867e-05, "loss": 0.515, "step": 9756, "task_loss": 0.5881970524787903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5330572724342346, "epoch": 8.25, "learning_rate": 2.50093926111459e-05, "loss": 0.5433, "step": 9757, "task_loss": 0.9034220576286316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7116758823394775, "epoch": 8.25, "learning_rate": 2.500626174076393e-05, "loss": 0.5214, "step": 9758, "task_loss": 0.6816636323928833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42546501755714417, "epoch": 8.25, "learning_rate": 2.500313087038197e-05, "loss": 0.4186, "step": 9759, "task_loss": 0.5598639249801636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27842020988464355, "epoch": 8.25, "learning_rate": 2.5e-05, "loss": 0.4466, "step": 9760, "task_loss": 0.3172523081302643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8322896957397461, "epoch": 8.25, "learning_rate": 2.4996869129618036e-05, "loss": 0.6107, "step": 9761, "task_loss": 1.0653760433197021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9027613401412964, "epoch": 8.25, "learning_rate": 2.4993738259236068e-05, "loss": 0.713, "step": 9762, "task_loss": 0.6902487874031067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4165083169937134, "epoch": 8.25, "learning_rate": 2.4990607388854103e-05, "loss": 0.5594, "step": 9763, "task_loss": 0.5994189977645874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5048587918281555, "epoch": 8.25, "learning_rate": 2.4987476518472135e-05, "loss": 0.7399, "step": 9764, "task_loss": 0.47846612334251404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4988597333431244, "epoch": 8.25, "learning_rate": 2.498434564809017e-05, "loss": 0.4251, "step": 9765, "task_loss": 0.8319604396820068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.617461621761322, "epoch": 8.26, "learning_rate": 2.4981214777708202e-05, "loss": 0.7457, "step": 9766, "task_loss": 0.7916346192359924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.446189284324646, "epoch": 8.26, "learning_rate": 2.4978083907326237e-05, "loss": 0.6142, "step": 9767, "task_loss": 0.0719948410987854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7715965509414673, "epoch": 8.26, "learning_rate": 2.4974953036944272e-05, "loss": 0.6097, "step": 9768, "task_loss": 0.5898035764694214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9232056140899658, "epoch": 8.26, "learning_rate": 2.4971822166562308e-05, "loss": 0.7637, "step": 9769, "task_loss": 0.8287847638130188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7511048316955566, "epoch": 8.26, "learning_rate": 2.496869129618034e-05, "loss": 0.7581, "step": 9770, "task_loss": 0.875433087348938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9275923371315002, "epoch": 8.26, "learning_rate": 2.4965560425798375e-05, "loss": 0.6748, "step": 9771, "task_loss": 1.1024384498596191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8835299611091614, "epoch": 8.26, "learning_rate": 2.4962429555416406e-05, "loss": 0.6811, "step": 9772, "task_loss": 0.6389249563217163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3322923481464386, "epoch": 8.26, "learning_rate": 2.495929868503444e-05, "loss": 0.523, "step": 9773, "task_loss": 0.10774056613445282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6005110740661621, "epoch": 8.26, "learning_rate": 2.4956167814652477e-05, "loss": 0.7037, "step": 9774, "task_loss": 0.6089823246002197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.289539635181427, "epoch": 8.26, "learning_rate": 2.495303694427051e-05, "loss": 0.5155, "step": 9775, "task_loss": 0.6378463506698608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8447316884994507, "epoch": 8.26, "learning_rate": 2.4949906073888544e-05, "loss": 0.5385, "step": 9776, "task_loss": 1.0898407697677612 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3522007167339325, "epoch": 8.26, "learning_rate": 2.4946775203506575e-05, "loss": 0.6297, "step": 9777, "task_loss": 0.34557488560676575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45411217212677, "epoch": 8.27, "learning_rate": 2.494364433312461e-05, "loss": 0.6143, "step": 9778, "task_loss": 0.5791823267936707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5594055652618408, "epoch": 8.27, "learning_rate": 2.4940513462742642e-05, "loss": 0.6921, "step": 9779, "task_loss": 0.7267779111862183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.370950847864151, "epoch": 8.27, "learning_rate": 2.4937382592360677e-05, "loss": 0.5133, "step": 9780, "task_loss": 0.062184982001781464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.586974024772644, "epoch": 8.27, "learning_rate": 2.493425172197871e-05, "loss": 0.6772, "step": 9781, "task_loss": 0.4926864802837372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4595065116882324, "epoch": 8.27, "learning_rate": 2.4931120851596744e-05, "loss": 0.5703, "step": 9782, "task_loss": 0.5684661865234375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.522114098072052, "epoch": 8.27, "learning_rate": 2.4927989981214776e-05, "loss": 0.5526, "step": 9783, "task_loss": 1.1456571817398071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7290828824043274, "epoch": 8.27, "learning_rate": 2.492485911083281e-05, "loss": 0.4901, "step": 9784, "task_loss": 0.9151787757873535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47360944747924805, "epoch": 8.27, "learning_rate": 2.4921728240450847e-05, "loss": 0.6254, "step": 9785, "task_loss": 0.9127264618873596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9442384839057922, "epoch": 8.27, "learning_rate": 2.4918597370068882e-05, "loss": 0.7879, "step": 9786, "task_loss": 0.6726751923561096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7030582427978516, "epoch": 8.27, "learning_rate": 2.4915466499686917e-05, "loss": 0.6473, "step": 9787, "task_loss": 0.9800394773483276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6978806853294373, "epoch": 8.27, "learning_rate": 2.491233562930495e-05, "loss": 0.7229, "step": 9788, "task_loss": 0.621996283531189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6398156881332397, "epoch": 8.27, "learning_rate": 2.4909204758922984e-05, "loss": 0.6498, "step": 9789, "task_loss": 0.8311284184455872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6060096025466919, "epoch": 8.28, "learning_rate": 2.4906073888541016e-05, "loss": 0.7353, "step": 9790, "task_loss": 1.0580732822418213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5791722536087036, "epoch": 8.28, "learning_rate": 2.490294301815905e-05, "loss": 0.7514, "step": 9791, "task_loss": 1.3705590963363647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8838489651679993, "epoch": 8.28, "learning_rate": 2.4899812147777083e-05, "loss": 0.6007, "step": 9792, "task_loss": 2.0057320594787598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5771565437316895, "epoch": 8.28, "learning_rate": 2.4896681277395118e-05, "loss": 0.5672, "step": 9793, "task_loss": 0.6778990030288696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4481932520866394, "epoch": 8.28, "learning_rate": 2.489355040701315e-05, "loss": 0.4802, "step": 9794, "task_loss": 0.5558140873908997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6865863800048828, "epoch": 8.28, "learning_rate": 2.4890419536631185e-05, "loss": 0.5807, "step": 9795, "task_loss": 1.336366891860962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6308372020721436, "epoch": 8.28, "learning_rate": 2.4887288666249216e-05, "loss": 0.7172, "step": 9796, "task_loss": 0.8156384229660034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7888388633728027, "epoch": 8.28, "learning_rate": 2.488415779586725e-05, "loss": 0.6356, "step": 9797, "task_loss": 0.6191509366035461 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4926626980304718, "epoch": 8.28, "learning_rate": 2.4881026925485287e-05, "loss": 0.4349, "step": 9798, "task_loss": 0.03461359068751335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6761987209320068, "epoch": 8.28, "learning_rate": 2.487789605510332e-05, "loss": 0.6501, "step": 9799, "task_loss": 0.5202962756156921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0001249313354492, "epoch": 8.28, "learning_rate": 2.4874765184721354e-05, "loss": 0.5901, "step": 9800, "task_loss": 1.032617449760437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4301905930042267, "epoch": 8.28, "learning_rate": 2.4871634314339386e-05, "loss": 0.5714, "step": 9801, "task_loss": 0.8728812336921692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36374250054359436, "epoch": 8.29, "learning_rate": 2.486850344395742e-05, "loss": 0.5214, "step": 9802, "task_loss": 0.7105289101600647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5905493497848511, "epoch": 8.29, "learning_rate": 2.4865372573575456e-05, "loss": 0.5964, "step": 9803, "task_loss": 0.2358376532793045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40099385380744934, "epoch": 8.29, "learning_rate": 2.486224170319349e-05, "loss": 0.4617, "step": 9804, "task_loss": 0.5151045918464661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.560271680355072, "epoch": 8.29, "learning_rate": 2.4859110832811523e-05, "loss": 0.6567, "step": 9805, "task_loss": 0.45720091462135315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6772330403327942, "epoch": 8.29, "learning_rate": 2.4855979962429558e-05, "loss": 0.5291, "step": 9806, "task_loss": 1.3350905179977417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24696660041809082, "epoch": 8.29, "learning_rate": 2.485284909204759e-05, "loss": 0.6288, "step": 9807, "task_loss": 0.6429446339607239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196793675422668, "epoch": 8.29, "learning_rate": 2.4849718221665625e-05, "loss": 0.4752, "step": 9808, "task_loss": 0.43215757608413696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.646846354007721, "epoch": 8.29, "learning_rate": 2.4846587351283657e-05, "loss": 0.8215, "step": 9809, "task_loss": 1.4600716829299927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.803666353225708, "epoch": 8.29, "learning_rate": 2.4843456480901692e-05, "loss": 0.5307, "step": 9810, "task_loss": 0.903242290019989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43563300371170044, "epoch": 8.29, "learning_rate": 2.4840325610519727e-05, "loss": 0.4953, "step": 9811, "task_loss": 0.43490755558013916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.59868985414505, "epoch": 8.29, "learning_rate": 2.483719474013776e-05, "loss": 0.6484, "step": 9812, "task_loss": 0.6087547540664673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31849607825279236, "epoch": 8.29, "learning_rate": 2.4834063869755794e-05, "loss": 0.4534, "step": 9813, "task_loss": 0.26377278566360474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.578727126121521, "epoch": 8.3, "learning_rate": 2.4830932999373826e-05, "loss": 0.6984, "step": 9814, "task_loss": 0.6595275402069092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8201743364334106, "epoch": 8.3, "learning_rate": 2.482780212899186e-05, "loss": 0.6974, "step": 9815, "task_loss": 0.4420449137687683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7716559171676636, "epoch": 8.3, "learning_rate": 2.4824671258609893e-05, "loss": 0.5462, "step": 9816, "task_loss": 0.9571405649185181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48683983087539673, "epoch": 8.3, "learning_rate": 2.4821540388227928e-05, "loss": 0.6399, "step": 9817, "task_loss": 0.37899547815322876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25826671719551086, "epoch": 8.3, "learning_rate": 2.481840951784596e-05, "loss": 0.5003, "step": 9818, "task_loss": 1.1129982471466064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42819541692733765, "epoch": 8.3, "learning_rate": 2.4815278647463995e-05, "loss": 0.7256, "step": 9819, "task_loss": 0.5075312852859497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35246962308883667, "epoch": 8.3, "learning_rate": 2.481214777708203e-05, "loss": 0.4882, "step": 9820, "task_loss": 1.0756168365478516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9138228893280029, "epoch": 8.3, "learning_rate": 2.4809016906700065e-05, "loss": 0.6041, "step": 9821, "task_loss": 0.9418540000915527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8453227877616882, "epoch": 8.3, "learning_rate": 2.48058860363181e-05, "loss": 0.7089, "step": 9822, "task_loss": 1.5711437463760376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.846814751625061, "epoch": 8.3, "learning_rate": 2.4802755165936132e-05, "loss": 0.5759, "step": 9823, "task_loss": 0.717609167098999 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7205470204353333, "epoch": 8.3, "learning_rate": 2.4799624295554167e-05, "loss": 0.5016, "step": 9824, "task_loss": 0.7568073272705078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3810436725616455, "epoch": 8.3, "learning_rate": 2.47964934251722e-05, "loss": 0.5267, "step": 9825, "task_loss": 0.6432167291641235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5395567417144775, "epoch": 8.31, "learning_rate": 2.4793362554790234e-05, "loss": 0.572, "step": 9826, "task_loss": 0.4770897626876831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5763996839523315, "epoch": 8.31, "learning_rate": 2.4790231684408266e-05, "loss": 0.5513, "step": 9827, "task_loss": 0.18301555514335632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5889967679977417, "epoch": 8.31, "learning_rate": 2.47871008140263e-05, "loss": 0.7693, "step": 9828, "task_loss": 0.8509719371795654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4895634353160858, "epoch": 8.31, "learning_rate": 2.4783969943644333e-05, "loss": 0.4969, "step": 9829, "task_loss": 0.5889846086502075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6162539720535278, "epoch": 8.31, "learning_rate": 2.4780839073262368e-05, "loss": 0.5706, "step": 9830, "task_loss": 0.6609732508659363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23041072487831116, "epoch": 8.31, "learning_rate": 2.47777082028804e-05, "loss": 0.6049, "step": 9831, "task_loss": 0.4625551104545593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.514205276966095, "epoch": 8.31, "learning_rate": 2.4774577332498435e-05, "loss": 0.4153, "step": 9832, "task_loss": 0.6173587441444397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6452626585960388, "epoch": 8.31, "learning_rate": 2.4771446462116467e-05, "loss": 0.6064, "step": 9833, "task_loss": 0.8224924802780151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7317449450492859, "epoch": 8.31, "learning_rate": 2.4768315591734502e-05, "loss": 0.7362, "step": 9834, "task_loss": 0.3640364110469818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4431402385234833, "epoch": 8.31, "learning_rate": 2.4765184721352537e-05, "loss": 0.5029, "step": 9835, "task_loss": 0.4230149984359741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7061294317245483, "epoch": 8.31, "learning_rate": 2.4762053850970572e-05, "loss": 0.8085, "step": 9836, "task_loss": 0.685654878616333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6785064339637756, "epoch": 8.32, "learning_rate": 2.4758922980588604e-05, "loss": 0.6537, "step": 9837, "task_loss": 0.7557619214057922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5097963213920593, "epoch": 8.32, "learning_rate": 2.475579211020664e-05, "loss": 0.7413, "step": 9838, "task_loss": 0.8827743530273438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5368660092353821, "epoch": 8.32, "learning_rate": 2.4752661239824675e-05, "loss": 0.6517, "step": 9839, "task_loss": 0.47442033886909485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5106333494186401, "epoch": 8.32, "learning_rate": 2.4749530369442706e-05, "loss": 0.6059, "step": 9840, "task_loss": 0.6703458428382874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5872725248336792, "epoch": 8.32, "learning_rate": 2.474639949906074e-05, "loss": 0.5007, "step": 9841, "task_loss": 0.8153627514839172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2969198226928711, "epoch": 8.32, "learning_rate": 2.4743268628678773e-05, "loss": 0.4252, "step": 9842, "task_loss": 0.3649481534957886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6012288331985474, "epoch": 8.32, "learning_rate": 2.474013775829681e-05, "loss": 0.7112, "step": 9843, "task_loss": 0.7785567045211792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7293040752410889, "epoch": 8.32, "learning_rate": 2.473700688791484e-05, "loss": 0.5862, "step": 9844, "task_loss": 0.583050012588501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4419546127319336, "epoch": 8.32, "learning_rate": 2.4733876017532875e-05, "loss": 0.7271, "step": 9845, "task_loss": 0.2402992695569992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6005607843399048, "epoch": 8.32, "learning_rate": 2.473074514715091e-05, "loss": 0.6238, "step": 9846, "task_loss": 1.16754150390625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.789362907409668, "epoch": 8.32, "learning_rate": 2.4727614276768942e-05, "loss": 0.7373, "step": 9847, "task_loss": 0.4799686372280121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3846607804298401, "epoch": 8.32, "learning_rate": 2.4724483406386978e-05, "loss": 0.4133, "step": 9848, "task_loss": 0.47650423645973206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9437605142593384, "epoch": 8.33, "learning_rate": 2.472135253600501e-05, "loss": 0.5426, "step": 9849, "task_loss": 0.9442684650421143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9144519567489624, "epoch": 8.33, "learning_rate": 2.4718221665623044e-05, "loss": 0.6273, "step": 9850, "task_loss": 1.1353188753128052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6332501173019409, "epoch": 8.33, "learning_rate": 2.4715090795241076e-05, "loss": 0.5556, "step": 9851, "task_loss": 0.8133047223091125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5231428146362305, "epoch": 8.33, "learning_rate": 2.471195992485911e-05, "loss": 0.5109, "step": 9852, "task_loss": 0.5231909155845642 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8956202864646912, "epoch": 8.33, "learning_rate": 2.4708829054477147e-05, "loss": 0.7053, "step": 9853, "task_loss": 0.675145149230957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1399009227752686, "epoch": 8.33, "learning_rate": 2.4705698184095182e-05, "loss": 0.6794, "step": 9854, "task_loss": 1.0677565336227417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5294022560119629, "epoch": 8.33, "learning_rate": 2.4702567313713214e-05, "loss": 0.6131, "step": 9855, "task_loss": 0.6205459833145142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4171682596206665, "epoch": 8.33, "learning_rate": 2.469943644333125e-05, "loss": 0.5023, "step": 9856, "task_loss": 0.037686292082071304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8402076959609985, "epoch": 8.33, "learning_rate": 2.469630557294928e-05, "loss": 0.6469, "step": 9857, "task_loss": 0.5891963839530945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5141419172286987, "epoch": 8.33, "learning_rate": 2.4693174702567316e-05, "loss": 0.5572, "step": 9858, "task_loss": 0.7667086720466614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6158074736595154, "epoch": 8.33, "learning_rate": 2.469004383218535e-05, "loss": 0.5794, "step": 9859, "task_loss": 0.3401864469051361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4195372760295868, "epoch": 8.33, "learning_rate": 2.4686912961803383e-05, "loss": 0.6774, "step": 9860, "task_loss": 1.0835553407669067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4078543782234192, "epoch": 8.34, "learning_rate": 2.4683782091421418e-05, "loss": 0.6458, "step": 9861, "task_loss": 0.2916333079338074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34048277139663696, "epoch": 8.34, "learning_rate": 2.468065122103945e-05, "loss": 0.7472, "step": 9862, "task_loss": 0.7828197479248047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6293551325798035, "epoch": 8.34, "learning_rate": 2.4677520350657485e-05, "loss": 0.539, "step": 9863, "task_loss": 0.7441173195838928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5367777347564697, "epoch": 8.34, "learning_rate": 2.4674389480275517e-05, "loss": 0.58, "step": 9864, "task_loss": 0.3959091007709503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35975193977355957, "epoch": 8.34, "learning_rate": 2.467125860989355e-05, "loss": 0.5073, "step": 9865, "task_loss": 0.23721276223659515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7828965783119202, "epoch": 8.34, "learning_rate": 2.4668127739511583e-05, "loss": 0.6859, "step": 9866, "task_loss": 0.9274908304214478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40985533595085144, "epoch": 8.34, "learning_rate": 2.466499686912962e-05, "loss": 0.5682, "step": 9867, "task_loss": 0.18314972519874573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5380086898803711, "epoch": 8.34, "learning_rate": 2.466186599874765e-05, "loss": 0.5556, "step": 9868, "task_loss": 0.6917690634727478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5463570356369019, "epoch": 8.34, "learning_rate": 2.4658735128365686e-05, "loss": 0.5163, "step": 9869, "task_loss": 0.16490790247917175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6575272083282471, "epoch": 8.34, "learning_rate": 2.465560425798372e-05, "loss": 0.5537, "step": 9870, "task_loss": 0.8246152997016907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33709871768951416, "epoch": 8.34, "learning_rate": 2.4652473387601756e-05, "loss": 0.3581, "step": 9871, "task_loss": 0.7172170281410217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6032953262329102, "epoch": 8.34, "learning_rate": 2.464934251721979e-05, "loss": 0.6673, "step": 9872, "task_loss": 0.42482560873031616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4876018762588501, "epoch": 8.35, "learning_rate": 2.4646211646837823e-05, "loss": 0.5915, "step": 9873, "task_loss": 0.7916715145111084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3742116093635559, "epoch": 8.35, "learning_rate": 2.4643080776455858e-05, "loss": 0.7431, "step": 9874, "task_loss": 0.6383658051490784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6789668798446655, "epoch": 8.35, "learning_rate": 2.463994990607389e-05, "loss": 0.6481, "step": 9875, "task_loss": 1.7623684406280518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4184691309928894, "epoch": 8.35, "learning_rate": 2.4636819035691925e-05, "loss": 0.4883, "step": 9876, "task_loss": 0.06622372567653656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21534386277198792, "epoch": 8.35, "learning_rate": 2.4633688165309957e-05, "loss": 0.4421, "step": 9877, "task_loss": 0.4368375837802887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6860195398330688, "epoch": 8.35, "learning_rate": 2.4630557294927992e-05, "loss": 0.5637, "step": 9878, "task_loss": 0.8221977353096008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29188886284828186, "epoch": 8.35, "learning_rate": 2.4627426424546024e-05, "loss": 0.3283, "step": 9879, "task_loss": 0.0759565532207489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21478693187236786, "epoch": 8.35, "learning_rate": 2.462429555416406e-05, "loss": 0.5437, "step": 9880, "task_loss": 0.4772525131702423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6138417720794678, "epoch": 8.35, "learning_rate": 2.462116468378209e-05, "loss": 0.7148, "step": 9881, "task_loss": 0.34902307391166687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7106969952583313, "epoch": 8.35, "learning_rate": 2.4618033813400126e-05, "loss": 0.6, "step": 9882, "task_loss": 0.7208659648895264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7422816753387451, "epoch": 8.35, "learning_rate": 2.461490294301816e-05, "loss": 0.5683, "step": 9883, "task_loss": 0.897875964641571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5673907995223999, "epoch": 8.35, "learning_rate": 2.4611772072636193e-05, "loss": 0.7471, "step": 9884, "task_loss": 0.5297415256500244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4081304967403412, "epoch": 8.36, "learning_rate": 2.4608641202254228e-05, "loss": 0.4634, "step": 9885, "task_loss": 0.6357421875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.434589684009552, "epoch": 8.36, "learning_rate": 2.460551033187226e-05, "loss": 0.3888, "step": 9886, "task_loss": 0.47533607482910156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6000601053237915, "epoch": 8.36, "learning_rate": 2.4602379461490295e-05, "loss": 0.4618, "step": 9887, "task_loss": 1.30251145362854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5206533074378967, "epoch": 8.36, "learning_rate": 2.459924859110833e-05, "loss": 0.6049, "step": 9888, "task_loss": 0.8804683685302734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7371069192886353, "epoch": 8.36, "learning_rate": 2.4596117720726365e-05, "loss": 0.7538, "step": 9889, "task_loss": 0.4413435757160187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41529735922813416, "epoch": 8.36, "learning_rate": 2.4592986850344397e-05, "loss": 0.6814, "step": 9890, "task_loss": 0.46244531869888306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3115867078304291, "epoch": 8.36, "learning_rate": 2.4589855979962432e-05, "loss": 0.4341, "step": 9891, "task_loss": 0.30138614773750305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8029758930206299, "epoch": 8.36, "learning_rate": 2.4586725109580464e-05, "loss": 0.7064, "step": 9892, "task_loss": 0.8548686504364014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4826154112815857, "epoch": 8.36, "learning_rate": 2.45835942391985e-05, "loss": 0.5168, "step": 9893, "task_loss": 0.7385037541389465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46660467982292175, "epoch": 8.36, "learning_rate": 2.458046336881653e-05, "loss": 0.4911, "step": 9894, "task_loss": 0.4055531322956085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48844194412231445, "epoch": 8.36, "learning_rate": 2.4577332498434566e-05, "loss": 0.4976, "step": 9895, "task_loss": 0.4661655128002167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25871533155441284, "epoch": 8.36, "learning_rate": 2.45742016280526e-05, "loss": 0.493, "step": 9896, "task_loss": 0.6176729798316956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32276540994644165, "epoch": 8.37, "learning_rate": 2.4571070757670633e-05, "loss": 0.4519, "step": 9897, "task_loss": 0.7840133905410767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46233344078063965, "epoch": 8.37, "learning_rate": 2.4567939887288668e-05, "loss": 0.6279, "step": 9898, "task_loss": 0.633537769317627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1750200986862183, "epoch": 8.37, "learning_rate": 2.45648090169067e-05, "loss": 0.7871, "step": 9899, "task_loss": 0.5443633794784546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7930914759635925, "epoch": 8.37, "learning_rate": 2.4561678146524735e-05, "loss": 0.836, "step": 9900, "task_loss": 1.156526803970337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9664674997329712, "epoch": 8.37, "learning_rate": 2.4558547276142767e-05, "loss": 0.7363, "step": 9901, "task_loss": 1.19019615650177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7851515412330627, "epoch": 8.37, "learning_rate": 2.4555416405760802e-05, "loss": 0.7644, "step": 9902, "task_loss": 0.8157622218132019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34464430809020996, "epoch": 8.37, "learning_rate": 2.4552285535378834e-05, "loss": 0.5382, "step": 9903, "task_loss": 0.1706211417913437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.657658576965332, "epoch": 8.37, "learning_rate": 2.454915466499687e-05, "loss": 0.6409, "step": 9904, "task_loss": 0.30190494656562805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6913620233535767, "epoch": 8.37, "learning_rate": 2.4546023794614904e-05, "loss": 0.5071, "step": 9905, "task_loss": 0.30070996284484863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6792372465133667, "epoch": 8.37, "learning_rate": 2.454289292423294e-05, "loss": 0.7744, "step": 9906, "task_loss": 1.4710584878921509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45893043279647827, "epoch": 8.37, "learning_rate": 2.453976205385097e-05, "loss": 0.5484, "step": 9907, "task_loss": 0.44546574354171753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6444333791732788, "epoch": 8.38, "learning_rate": 2.4536631183469006e-05, "loss": 0.5832, "step": 9908, "task_loss": 0.909501314163208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4099351167678833, "epoch": 8.38, "learning_rate": 2.453350031308704e-05, "loss": 0.6337, "step": 9909, "task_loss": 1.043605923652649 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5520780086517334, "epoch": 8.38, "learning_rate": 2.4530369442705073e-05, "loss": 0.4879, "step": 9910, "task_loss": 0.47671180963516235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1342377662658691, "epoch": 8.38, "learning_rate": 2.452723857232311e-05, "loss": 0.6362, "step": 9911, "task_loss": 1.4097089767456055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5928246378898621, "epoch": 8.38, "learning_rate": 2.452410770194114e-05, "loss": 0.4965, "step": 9912, "task_loss": 0.3586793541908264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3315298557281494, "epoch": 8.38, "learning_rate": 2.4520976831559175e-05, "loss": 0.5426, "step": 9913, "task_loss": 1.1511166095733643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5311892628669739, "epoch": 8.38, "learning_rate": 2.4517845961177207e-05, "loss": 0.5549, "step": 9914, "task_loss": 0.5487390160560608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27792179584503174, "epoch": 8.38, "learning_rate": 2.4514715090795242e-05, "loss": 0.5044, "step": 9915, "task_loss": 0.07966852188110352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6477609276771545, "epoch": 8.38, "learning_rate": 2.4511584220413274e-05, "loss": 0.673, "step": 9916, "task_loss": 0.44931381940841675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3145618140697479, "epoch": 8.38, "learning_rate": 2.450845335003131e-05, "loss": 0.784, "step": 9917, "task_loss": 0.25357958674430847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5165849924087524, "epoch": 8.38, "learning_rate": 2.450532247964934e-05, "loss": 0.6877, "step": 9918, "task_loss": 0.36811959743499756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6639126539230347, "epoch": 8.38, "learning_rate": 2.4502191609267376e-05, "loss": 0.4586, "step": 9919, "task_loss": 1.0195274353027344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6723572015762329, "epoch": 8.39, "learning_rate": 2.449906073888541e-05, "loss": 0.628, "step": 9920, "task_loss": 1.3522298336029053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3856327533721924, "epoch": 8.39, "learning_rate": 2.4495929868503447e-05, "loss": 0.4572, "step": 9921, "task_loss": 0.2986103892326355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2274285852909088, "epoch": 8.39, "learning_rate": 2.449279899812148e-05, "loss": 0.5942, "step": 9922, "task_loss": 0.42663300037384033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3421671986579895, "epoch": 8.39, "learning_rate": 2.4489668127739514e-05, "loss": 0.5683, "step": 9923, "task_loss": 0.21007700264453888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3910793960094452, "epoch": 8.39, "learning_rate": 2.448653725735755e-05, "loss": 0.5779, "step": 9924, "task_loss": 0.5518826842308044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7828851938247681, "epoch": 8.39, "learning_rate": 2.448340638697558e-05, "loss": 0.6664, "step": 9925, "task_loss": 0.7176758050918579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5850119590759277, "epoch": 8.39, "learning_rate": 2.4480275516593616e-05, "loss": 0.5331, "step": 9926, "task_loss": 0.9597911238670349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6099032163619995, "epoch": 8.39, "learning_rate": 2.4477144646211647e-05, "loss": 0.7603, "step": 9927, "task_loss": 0.6709460616111755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6014007925987244, "epoch": 8.39, "learning_rate": 2.4474013775829683e-05, "loss": 0.589, "step": 9928, "task_loss": 0.8610801696777344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49778586626052856, "epoch": 8.39, "learning_rate": 2.4470882905447714e-05, "loss": 0.6271, "step": 9929, "task_loss": 0.12230429798364639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5932390689849854, "epoch": 8.39, "learning_rate": 2.446775203506575e-05, "loss": 0.5, "step": 9930, "task_loss": 0.603852391242981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7614178657531738, "epoch": 8.39, "learning_rate": 2.446462116468378e-05, "loss": 0.5277, "step": 9931, "task_loss": 0.5778088569641113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6240192651748657, "epoch": 8.4, "learning_rate": 2.4461490294301817e-05, "loss": 0.8097, "step": 9932, "task_loss": 0.6008275151252747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6886321902275085, "epoch": 8.4, "learning_rate": 2.4458359423919852e-05, "loss": 0.5711, "step": 9933, "task_loss": 0.4352661669254303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.448507159948349, "epoch": 8.4, "learning_rate": 2.4455228553537883e-05, "loss": 0.4347, "step": 9934, "task_loss": 0.5392898321151733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5317127108573914, "epoch": 8.4, "learning_rate": 2.445209768315592e-05, "loss": 0.4773, "step": 9935, "task_loss": 0.7487993836402893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3034961223602295, "epoch": 8.4, "learning_rate": 2.444896681277395e-05, "loss": 0.5321, "step": 9936, "task_loss": 0.08846379816532135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41622430086135864, "epoch": 8.4, "learning_rate": 2.4445835942391986e-05, "loss": 0.5276, "step": 9937, "task_loss": 0.5625048875808716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4710637331008911, "epoch": 8.4, "learning_rate": 2.444270507201002e-05, "loss": 0.5442, "step": 9938, "task_loss": 0.34839844703674316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44591912627220154, "epoch": 8.4, "learning_rate": 2.4439574201628056e-05, "loss": 0.5019, "step": 9939, "task_loss": 0.3428493142127991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7319766283035278, "epoch": 8.4, "learning_rate": 2.4436443331246088e-05, "loss": 0.5354, "step": 9940, "task_loss": 0.5464314818382263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5178499221801758, "epoch": 8.4, "learning_rate": 2.4433312460864123e-05, "loss": 0.6034, "step": 9941, "task_loss": 0.5528638958930969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6277734041213989, "epoch": 8.4, "learning_rate": 2.4430181590482155e-05, "loss": 0.6536, "step": 9942, "task_loss": 0.6893191337585449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3855842053890228, "epoch": 8.4, "learning_rate": 2.442705072010019e-05, "loss": 0.493, "step": 9943, "task_loss": 0.2892525792121887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3788519501686096, "epoch": 8.41, "learning_rate": 2.4423919849718225e-05, "loss": 0.5061, "step": 9944, "task_loss": 0.18143777549266815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.455813467502594, "epoch": 8.41, "learning_rate": 2.4420788979336257e-05, "loss": 0.5828, "step": 9945, "task_loss": 0.3734906017780304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5464104413986206, "epoch": 8.41, "learning_rate": 2.4417658108954292e-05, "loss": 0.8566, "step": 9946, "task_loss": 0.6741085052490234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43964749574661255, "epoch": 8.41, "learning_rate": 2.4414527238572324e-05, "loss": 0.547, "step": 9947, "task_loss": 0.9466725587844849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5698896050453186, "epoch": 8.41, "learning_rate": 2.441139636819036e-05, "loss": 0.7493, "step": 9948, "task_loss": 0.5391853451728821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7117801308631897, "epoch": 8.41, "learning_rate": 2.440826549780839e-05, "loss": 0.7075, "step": 9949, "task_loss": 1.6436822414398193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3445143699645996, "epoch": 8.41, "learning_rate": 2.4405134627426426e-05, "loss": 0.443, "step": 9950, "task_loss": 0.219416543841362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.358405590057373, "epoch": 8.41, "learning_rate": 2.4402003757044458e-05, "loss": 0.7483, "step": 9951, "task_loss": 1.4144647121429443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7176732420921326, "epoch": 8.41, "learning_rate": 2.4398872886662493e-05, "loss": 0.6294, "step": 9952, "task_loss": 0.5734506845474243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3244240880012512, "epoch": 8.41, "learning_rate": 2.4395742016280525e-05, "loss": 0.4581, "step": 9953, "task_loss": 0.618253231048584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6196889281272888, "epoch": 8.41, "learning_rate": 2.439261114589856e-05, "loss": 0.6069, "step": 9954, "task_loss": 1.1679439544677734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3547516167163849, "epoch": 8.41, "learning_rate": 2.4389480275516595e-05, "loss": 0.5014, "step": 9955, "task_loss": 1.1187589168548584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6418308615684509, "epoch": 8.42, "learning_rate": 2.438634940513463e-05, "loss": 0.6135, "step": 9956, "task_loss": 1.3975107669830322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44960618019104004, "epoch": 8.42, "learning_rate": 2.4383218534752665e-05, "loss": 0.4314, "step": 9957, "task_loss": 0.46712368726730347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3678209185600281, "epoch": 8.42, "learning_rate": 2.4380087664370697e-05, "loss": 0.499, "step": 9958, "task_loss": 0.7347737550735474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6811479330062866, "epoch": 8.42, "learning_rate": 2.4376956793988732e-05, "loss": 0.6027, "step": 9959, "task_loss": 0.9821649193763733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3231372833251953, "epoch": 8.42, "learning_rate": 2.4373825923606764e-05, "loss": 0.5054, "step": 9960, "task_loss": 0.25403353571891785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8462283611297607, "epoch": 8.42, "learning_rate": 2.43706950532248e-05, "loss": 0.6002, "step": 9961, "task_loss": 0.34565621614456177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9598999619483948, "epoch": 8.42, "learning_rate": 2.436756418284283e-05, "loss": 0.6388, "step": 9962, "task_loss": 1.0221996307373047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7037984132766724, "epoch": 8.42, "learning_rate": 2.4364433312460866e-05, "loss": 0.5015, "step": 9963, "task_loss": 1.7381186485290527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6254231929779053, "epoch": 8.42, "learning_rate": 2.4361302442078898e-05, "loss": 0.8332, "step": 9964, "task_loss": 1.2084017992019653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3993983268737793, "epoch": 8.42, "learning_rate": 2.4358171571696933e-05, "loss": 0.5726, "step": 9965, "task_loss": 1.1140997409820557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7023029327392578, "epoch": 8.42, "learning_rate": 2.4355040701314965e-05, "loss": 0.6658, "step": 9966, "task_loss": 0.3759397864341736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.910719633102417, "epoch": 8.42, "learning_rate": 2.4351909830933e-05, "loss": 0.6888, "step": 9967, "task_loss": 0.7321150898933411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7067651748657227, "epoch": 8.43, "learning_rate": 2.4348778960551032e-05, "loss": 0.6746, "step": 9968, "task_loss": 1.856488585472107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7458220720291138, "epoch": 8.43, "learning_rate": 2.4345648090169067e-05, "loss": 0.4836, "step": 9969, "task_loss": 0.49604514241218567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3404237627983093, "epoch": 8.43, "learning_rate": 2.4342517219787102e-05, "loss": 0.5289, "step": 9970, "task_loss": 1.1282235383987427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38179469108581543, "epoch": 8.43, "learning_rate": 2.4339386349405134e-05, "loss": 0.4583, "step": 9971, "task_loss": 0.23799386620521545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.404116153717041, "epoch": 8.43, "learning_rate": 2.433625547902317e-05, "loss": 0.5364, "step": 9972, "task_loss": 0.47067582607269287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6628414392471313, "epoch": 8.43, "learning_rate": 2.4333124608641204e-05, "loss": 0.6524, "step": 9973, "task_loss": 1.064932107925415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5949008464813232, "epoch": 8.43, "learning_rate": 2.432999373825924e-05, "loss": 0.5537, "step": 9974, "task_loss": 1.1405278444290161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5190251469612122, "epoch": 8.43, "learning_rate": 2.432686286787727e-05, "loss": 0.6075, "step": 9975, "task_loss": 0.5566295385360718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7141621112823486, "epoch": 8.43, "learning_rate": 2.4323731997495306e-05, "loss": 0.6823, "step": 9976, "task_loss": 0.811552882194519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7980354428291321, "epoch": 8.43, "learning_rate": 2.4320601127113338e-05, "loss": 0.7002, "step": 9977, "task_loss": 0.8878332376480103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5444315671920776, "epoch": 8.43, "learning_rate": 2.4317470256731373e-05, "loss": 0.5103, "step": 9978, "task_loss": 0.8267861008644104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5498738288879395, "epoch": 8.44, "learning_rate": 2.4314339386349405e-05, "loss": 0.6799, "step": 9979, "task_loss": 2.00424861907959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.672055721282959, "epoch": 8.44, "learning_rate": 2.431120851596744e-05, "loss": 0.756, "step": 9980, "task_loss": 1.1604928970336914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5658978223800659, "epoch": 8.44, "learning_rate": 2.4308077645585475e-05, "loss": 0.5219, "step": 9981, "task_loss": 0.3493707478046417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9688014984130859, "epoch": 8.44, "learning_rate": 2.4304946775203507e-05, "loss": 0.7177, "step": 9982, "task_loss": 1.7470383644104004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23485460877418518, "epoch": 8.44, "learning_rate": 2.4301815904821542e-05, "loss": 0.488, "step": 9983, "task_loss": 0.3482387065887451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5366823673248291, "epoch": 8.44, "learning_rate": 2.4298685034439574e-05, "loss": 0.5734, "step": 9984, "task_loss": 1.0642200708389282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4551658630371094, "epoch": 8.44, "learning_rate": 2.429555416405761e-05, "loss": 0.655, "step": 9985, "task_loss": 0.16337250173091888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40326690673828125, "epoch": 8.44, "learning_rate": 2.429242329367564e-05, "loss": 0.535, "step": 9986, "task_loss": 0.5087176561355591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9036825895309448, "epoch": 8.44, "learning_rate": 2.4289292423293676e-05, "loss": 0.6947, "step": 9987, "task_loss": 1.3097330331802368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3586169481277466, "epoch": 8.44, "learning_rate": 2.428616155291171e-05, "loss": 0.6005, "step": 9988, "task_loss": 0.7905902862548828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5670782327651978, "epoch": 8.44, "learning_rate": 2.4283030682529743e-05, "loss": 0.5376, "step": 9989, "task_loss": 0.6293444633483887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.606636643409729, "epoch": 8.44, "learning_rate": 2.427989981214778e-05, "loss": 0.5831, "step": 9990, "task_loss": 0.763329803943634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8373957872390747, "epoch": 8.45, "learning_rate": 2.4276768941765814e-05, "loss": 0.7246, "step": 9991, "task_loss": 0.5222405791282654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0980116128921509, "epoch": 8.45, "learning_rate": 2.4273638071383845e-05, "loss": 0.7213, "step": 9992, "task_loss": 1.1848018169403076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40121960639953613, "epoch": 8.45, "learning_rate": 2.427050720100188e-05, "loss": 0.7836, "step": 9993, "task_loss": 0.12637750804424286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7144135236740112, "epoch": 8.45, "learning_rate": 2.4267376330619916e-05, "loss": 0.6367, "step": 9994, "task_loss": 0.5235933661460876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.666379451751709, "epoch": 8.45, "learning_rate": 2.4264245460237948e-05, "loss": 0.5273, "step": 9995, "task_loss": 0.6461248993873596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.342473566532135, "epoch": 8.45, "learning_rate": 2.4261114589855983e-05, "loss": 0.6032, "step": 9996, "task_loss": 0.3875517249107361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8045641183853149, "epoch": 8.45, "learning_rate": 2.4257983719474014e-05, "loss": 0.7739, "step": 9997, "task_loss": 1.0643842220306396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5942565202713013, "epoch": 8.45, "learning_rate": 2.425485284909205e-05, "loss": 0.5301, "step": 9998, "task_loss": 0.6240968108177185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0759449005126953, "epoch": 8.45, "learning_rate": 2.425172197871008e-05, "loss": 0.6748, "step": 9999, "task_loss": 1.446075439453125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7185943722724915, "epoch": 8.45, "learning_rate": 2.4248591108328117e-05, "loss": 0.5617, "step": 10000, "task_loss": 0.48977795243263245 }, { "epoch": 8.45, "eval_accuracy": 0.9002772277227723, "eval_loss": 0.3940442204475403, "eval_runtime": 208.9848, "eval_samples_per_second": 120.822, "eval_steps_per_second": 0.947, "step": 10000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4695115387439728, "epoch": 8.45, "learning_rate": 2.424546023794615e-05, "loss": 0.4896, "step": 10001, "task_loss": 0.8285620808601379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6521762609481812, "epoch": 8.45, "learning_rate": 2.4242329367564184e-05, "loss": 0.6095, "step": 10002, "task_loss": 0.7066267728805542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43545934557914734, "epoch": 8.46, "learning_rate": 2.4239198497182215e-05, "loss": 0.4137, "step": 10003, "task_loss": 0.6516613960266113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6728017330169678, "epoch": 8.46, "learning_rate": 2.423606762680025e-05, "loss": 0.6438, "step": 10004, "task_loss": 0.43445441126823425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4340176582336426, "epoch": 8.46, "learning_rate": 2.4232936756418286e-05, "loss": 0.7331, "step": 10005, "task_loss": 0.36168721318244934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4232768714427948, "epoch": 8.46, "learning_rate": 2.422980588603632e-05, "loss": 0.4833, "step": 10006, "task_loss": 0.7519941329956055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8345622420310974, "epoch": 8.46, "learning_rate": 2.4226675015654353e-05, "loss": 0.7483, "step": 10007, "task_loss": 0.23146821558475494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32085973024368286, "epoch": 8.46, "learning_rate": 2.4223544145272388e-05, "loss": 0.5392, "step": 10008, "task_loss": 0.4156527519226074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6315290331840515, "epoch": 8.46, "learning_rate": 2.4220413274890423e-05, "loss": 0.4606, "step": 10009, "task_loss": 0.3406989574432373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30340293049812317, "epoch": 8.46, "learning_rate": 2.4217282404508455e-05, "loss": 0.5055, "step": 10010, "task_loss": 0.5106039643287659 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7309026718139648, "epoch": 8.46, "learning_rate": 2.421415153412649e-05, "loss": 0.6509, "step": 10011, "task_loss": 0.65108722448349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8221062421798706, "epoch": 8.46, "learning_rate": 2.421102066374452e-05, "loss": 0.758, "step": 10012, "task_loss": 0.9883490204811096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5622241497039795, "epoch": 8.46, "learning_rate": 2.4207889793362557e-05, "loss": 0.6013, "step": 10013, "task_loss": 0.8154910802841187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7545502781867981, "epoch": 8.46, "learning_rate": 2.420475892298059e-05, "loss": 0.6365, "step": 10014, "task_loss": 1.3718868494033813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3548926115036011, "epoch": 8.47, "learning_rate": 2.4201628052598624e-05, "loss": 0.6287, "step": 10015, "task_loss": 0.48067307472229004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6058254837989807, "epoch": 8.47, "learning_rate": 2.4198497182216656e-05, "loss": 0.7065, "step": 10016, "task_loss": 0.21324816346168518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6380953788757324, "epoch": 8.47, "learning_rate": 2.419536631183469e-05, "loss": 0.7247, "step": 10017, "task_loss": 0.42369064688682556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6375634074211121, "epoch": 8.47, "learning_rate": 2.4192235441452726e-05, "loss": 0.5986, "step": 10018, "task_loss": 0.4677978754043579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25553107261657715, "epoch": 8.47, "learning_rate": 2.4189104571070758e-05, "loss": 0.4137, "step": 10019, "task_loss": 0.20135992765426636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.669151782989502, "epoch": 8.47, "learning_rate": 2.4185973700688793e-05, "loss": 0.716, "step": 10020, "task_loss": 0.20380055904388428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.576530396938324, "epoch": 8.47, "learning_rate": 2.4182842830306825e-05, "loss": 0.4179, "step": 10021, "task_loss": 0.1882864087820053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3519124388694763, "epoch": 8.47, "learning_rate": 2.417971195992486e-05, "loss": 0.5184, "step": 10022, "task_loss": 0.585736870765686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5820249915122986, "epoch": 8.47, "learning_rate": 2.4176581089542895e-05, "loss": 0.5561, "step": 10023, "task_loss": 1.1234941482543945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5387487411499023, "epoch": 8.47, "learning_rate": 2.417345021916093e-05, "loss": 0.693, "step": 10024, "task_loss": 0.7885149121284485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6121490597724915, "epoch": 8.47, "learning_rate": 2.4170319348778962e-05, "loss": 0.5819, "step": 10025, "task_loss": 1.032874584197998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.849434494972229, "epoch": 8.47, "learning_rate": 2.4167188478396997e-05, "loss": 0.6349, "step": 10026, "task_loss": 0.49951547384262085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6523745059967041, "epoch": 8.48, "learning_rate": 2.416405760801503e-05, "loss": 0.6163, "step": 10027, "task_loss": 0.950152575969696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6120736598968506, "epoch": 8.48, "learning_rate": 2.4160926737633064e-05, "loss": 0.537, "step": 10028, "task_loss": 0.9116626977920532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6619846820831299, "epoch": 8.48, "learning_rate": 2.4157795867251096e-05, "loss": 0.7883, "step": 10029, "task_loss": 0.9125848412513733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4165514409542084, "epoch": 8.48, "learning_rate": 2.415466499686913e-05, "loss": 0.506, "step": 10030, "task_loss": 0.3042545020580292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.83123779296875, "epoch": 8.48, "learning_rate": 2.4151534126487166e-05, "loss": 0.5891, "step": 10031, "task_loss": 1.307411551475525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4516924023628235, "epoch": 8.48, "learning_rate": 2.4148403256105198e-05, "loss": 0.6436, "step": 10032, "task_loss": 0.7145549058914185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7274592518806458, "epoch": 8.48, "learning_rate": 2.4145272385723233e-05, "loss": 0.6287, "step": 10033, "task_loss": 0.33185163140296936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4643300175666809, "epoch": 8.48, "learning_rate": 2.4142141515341265e-05, "loss": 0.5584, "step": 10034, "task_loss": 0.5079402923583984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5182172060012817, "epoch": 8.48, "learning_rate": 2.41390106449593e-05, "loss": 0.662, "step": 10035, "task_loss": 0.5576354265213013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6107136607170105, "epoch": 8.48, "learning_rate": 2.4135879774577332e-05, "loss": 0.5055, "step": 10036, "task_loss": 0.6720110774040222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43989989161491394, "epoch": 8.48, "learning_rate": 2.4132748904195367e-05, "loss": 0.5675, "step": 10037, "task_loss": 0.8424568176269531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7813329100608826, "epoch": 8.48, "learning_rate": 2.41296180338134e-05, "loss": 0.6757, "step": 10038, "task_loss": 0.6108272671699524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6076241731643677, "epoch": 8.49, "learning_rate": 2.4126487163431434e-05, "loss": 0.5644, "step": 10039, "task_loss": 1.3787912130355835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5449113845825195, "epoch": 8.49, "learning_rate": 2.412335629304947e-05, "loss": 0.5371, "step": 10040, "task_loss": 0.3515976667404175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47949302196502686, "epoch": 8.49, "learning_rate": 2.4120225422667504e-05, "loss": 0.6352, "step": 10041, "task_loss": 0.34261560440063477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8516799211502075, "epoch": 8.49, "learning_rate": 2.4117094552285536e-05, "loss": 0.6789, "step": 10042, "task_loss": 1.3319565057754517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.814790666103363, "epoch": 8.49, "learning_rate": 2.411396368190357e-05, "loss": 0.6965, "step": 10043, "task_loss": 1.2482868432998657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7649241089820862, "epoch": 8.49, "learning_rate": 2.4110832811521606e-05, "loss": 0.5795, "step": 10044, "task_loss": 0.8935831785202026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6489366292953491, "epoch": 8.49, "learning_rate": 2.4107701941139638e-05, "loss": 0.6531, "step": 10045, "task_loss": 0.9061636328697205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8481380343437195, "epoch": 8.49, "learning_rate": 2.4104571070757673e-05, "loss": 0.5394, "step": 10046, "task_loss": 1.5611951351165771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7463185787200928, "epoch": 8.49, "learning_rate": 2.4101440200375705e-05, "loss": 0.5059, "step": 10047, "task_loss": 0.8823367953300476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4578933119773865, "epoch": 8.49, "learning_rate": 2.409830932999374e-05, "loss": 0.5355, "step": 10048, "task_loss": 0.6856697797775269 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3959437906742096, "epoch": 8.49, "learning_rate": 2.4095178459611772e-05, "loss": 0.6058, "step": 10049, "task_loss": 0.39225563406944275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8268585205078125, "epoch": 8.5, "learning_rate": 2.4092047589229807e-05, "loss": 0.5898, "step": 10050, "task_loss": 0.3212881088256836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37715452909469604, "epoch": 8.5, "learning_rate": 2.408891671884784e-05, "loss": 0.4795, "step": 10051, "task_loss": 0.15342839062213898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5306301116943359, "epoch": 8.5, "learning_rate": 2.4085785848465874e-05, "loss": 0.6753, "step": 10052, "task_loss": 0.48091644048690796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6271367073059082, "epoch": 8.5, "learning_rate": 2.4082654978083906e-05, "loss": 0.6638, "step": 10053, "task_loss": 1.0990879535675049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7916500568389893, "epoch": 8.5, "learning_rate": 2.407952410770194e-05, "loss": 0.6897, "step": 10054, "task_loss": 0.427519291639328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5497397184371948, "epoch": 8.5, "learning_rate": 2.4076393237319976e-05, "loss": 0.5039, "step": 10055, "task_loss": 0.3319660425186157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6097036600112915, "epoch": 8.5, "learning_rate": 2.4073262366938008e-05, "loss": 0.6168, "step": 10056, "task_loss": 0.7948583960533142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6388046741485596, "epoch": 8.5, "learning_rate": 2.4070131496556043e-05, "loss": 0.6594, "step": 10057, "task_loss": 1.332379698753357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5061922669410706, "epoch": 8.5, "learning_rate": 2.406700062617408e-05, "loss": 0.7999, "step": 10058, "task_loss": 0.6815206408500671 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6328309774398804, "epoch": 8.5, "learning_rate": 2.4063869755792114e-05, "loss": 0.5189, "step": 10059, "task_loss": 0.9016672968864441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7039703726768494, "epoch": 8.5, "learning_rate": 2.4060738885410145e-05, "loss": 0.6212, "step": 10060, "task_loss": 0.2377149760723114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6217050552368164, "epoch": 8.5, "learning_rate": 2.405760801502818e-05, "loss": 0.6962, "step": 10061, "task_loss": 0.9154659509658813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4088413715362549, "epoch": 8.51, "learning_rate": 2.4054477144646212e-05, "loss": 0.495, "step": 10062, "task_loss": 0.38036972284317017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5422283411026001, "epoch": 8.51, "learning_rate": 2.4051346274264248e-05, "loss": 0.5549, "step": 10063, "task_loss": 0.15044796466827393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36089956760406494, "epoch": 8.51, "learning_rate": 2.404821540388228e-05, "loss": 0.451, "step": 10064, "task_loss": 0.4183090329170227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8278560638427734, "epoch": 8.51, "learning_rate": 2.4045084533500314e-05, "loss": 0.7388, "step": 10065, "task_loss": 1.2189710140228271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3243861794471741, "epoch": 8.51, "learning_rate": 2.4041953663118346e-05, "loss": 0.6845, "step": 10066, "task_loss": 0.364583820104599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6037830710411072, "epoch": 8.51, "learning_rate": 2.403882279273638e-05, "loss": 0.7455, "step": 10067, "task_loss": 0.5041179060935974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40328243374824524, "epoch": 8.51, "learning_rate": 2.4035691922354417e-05, "loss": 0.7154, "step": 10068, "task_loss": 0.5689697861671448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29852166771888733, "epoch": 8.51, "learning_rate": 2.403256105197245e-05, "loss": 0.4525, "step": 10069, "task_loss": 0.556950032711029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39268195629119873, "epoch": 8.51, "learning_rate": 2.4029430181590484e-05, "loss": 0.4177, "step": 10070, "task_loss": 1.085085391998291 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4722900390625, "epoch": 8.51, "learning_rate": 2.4026299311208515e-05, "loss": 0.547, "step": 10071, "task_loss": 0.3280726969242096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5606147646903992, "epoch": 8.51, "learning_rate": 2.402316844082655e-05, "loss": 0.487, "step": 10072, "task_loss": 0.925628125667572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.703110933303833, "epoch": 8.51, "learning_rate": 2.4020037570444586e-05, "loss": 0.641, "step": 10073, "task_loss": 0.8123272061347961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6540012359619141, "epoch": 8.52, "learning_rate": 2.4016906700062617e-05, "loss": 0.711, "step": 10074, "task_loss": 1.0038928985595703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47289371490478516, "epoch": 8.52, "learning_rate": 2.4013775829680653e-05, "loss": 0.5714, "step": 10075, "task_loss": 0.5560613870620728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5598366260528564, "epoch": 8.52, "learning_rate": 2.4010644959298688e-05, "loss": 0.6244, "step": 10076, "task_loss": 0.7958265542984009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6754499077796936, "epoch": 8.52, "learning_rate": 2.400751408891672e-05, "loss": 0.7329, "step": 10077, "task_loss": 0.9756925702095032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48760074377059937, "epoch": 8.52, "learning_rate": 2.4004383218534755e-05, "loss": 0.5337, "step": 10078, "task_loss": 0.3495398461818695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7756227850914001, "epoch": 8.52, "learning_rate": 2.400125234815279e-05, "loss": 0.6639, "step": 10079, "task_loss": 0.7953657507896423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8964133858680725, "epoch": 8.52, "learning_rate": 2.3998121477770822e-05, "loss": 0.5806, "step": 10080, "task_loss": 1.6641080379486084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6283518075942993, "epoch": 8.52, "learning_rate": 2.3994990607388857e-05, "loss": 0.575, "step": 10081, "task_loss": 0.2374730110168457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4046695828437805, "epoch": 8.52, "learning_rate": 2.399185973700689e-05, "loss": 0.6469, "step": 10082, "task_loss": 1.0479753017425537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4784066081047058, "epoch": 8.52, "learning_rate": 2.3988728866624924e-05, "loss": 0.8183, "step": 10083, "task_loss": 0.8701391816139221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5335546731948853, "epoch": 8.52, "learning_rate": 2.3985597996242956e-05, "loss": 0.6281, "step": 10084, "task_loss": 0.7767938375473022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5283236503601074, "epoch": 8.52, "learning_rate": 2.398246712586099e-05, "loss": 0.613, "step": 10085, "task_loss": 0.2014520764350891 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7104955911636353, "epoch": 8.53, "learning_rate": 2.3979336255479023e-05, "loss": 0.7022, "step": 10086, "task_loss": 1.0238661766052246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45077240467071533, "epoch": 8.53, "learning_rate": 2.3976205385097058e-05, "loss": 0.6749, "step": 10087, "task_loss": 0.6980252265930176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7209736108779907, "epoch": 8.53, "learning_rate": 2.397307451471509e-05, "loss": 0.5436, "step": 10088, "task_loss": 0.7870707511901855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4215553402900696, "epoch": 8.53, "learning_rate": 2.3969943644333125e-05, "loss": 0.4465, "step": 10089, "task_loss": 1.3785171508789062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5783225297927856, "epoch": 8.53, "learning_rate": 2.396681277395116e-05, "loss": 0.6192, "step": 10090, "task_loss": 0.956068754196167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5005993843078613, "epoch": 8.53, "learning_rate": 2.3963681903569195e-05, "loss": 0.5378, "step": 10091, "task_loss": 1.0372499227523804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5231114625930786, "epoch": 8.53, "learning_rate": 2.3960551033187227e-05, "loss": 0.7009, "step": 10092, "task_loss": 0.375636488199234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7257409691810608, "epoch": 8.53, "learning_rate": 2.3957420162805262e-05, "loss": 0.5055, "step": 10093, "task_loss": 0.34078606963157654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5953303575515747, "epoch": 8.53, "learning_rate": 2.3954289292423297e-05, "loss": 0.5227, "step": 10094, "task_loss": 0.26136958599090576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4498937427997589, "epoch": 8.53, "learning_rate": 2.395115842204133e-05, "loss": 0.4785, "step": 10095, "task_loss": 0.2831118702888489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5362799167633057, "epoch": 8.53, "learning_rate": 2.3948027551659364e-05, "loss": 0.6293, "step": 10096, "task_loss": 0.52106112241745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4418151378631592, "epoch": 8.53, "learning_rate": 2.3944896681277396e-05, "loss": 0.5312, "step": 10097, "task_loss": 0.6767556667327881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7232069969177246, "epoch": 8.54, "learning_rate": 2.394176581089543e-05, "loss": 0.4913, "step": 10098, "task_loss": 1.1303107738494873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39561253786087036, "epoch": 8.54, "learning_rate": 2.3938634940513463e-05, "loss": 0.5336, "step": 10099, "task_loss": 0.39325106143951416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28984320163726807, "epoch": 8.54, "learning_rate": 2.3935504070131498e-05, "loss": 0.4767, "step": 10100, "task_loss": 0.10505656898021698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6892770528793335, "epoch": 8.54, "learning_rate": 2.393237319974953e-05, "loss": 0.5932, "step": 10101, "task_loss": 0.6771829128265381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3966529369354248, "epoch": 8.54, "learning_rate": 2.3929242329367565e-05, "loss": 0.5489, "step": 10102, "task_loss": 0.5574904680252075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3247409462928772, "epoch": 8.54, "learning_rate": 2.3926111458985597e-05, "loss": 0.696, "step": 10103, "task_loss": 0.1884775608778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5676581263542175, "epoch": 8.54, "learning_rate": 2.3922980588603632e-05, "loss": 0.5702, "step": 10104, "task_loss": 1.1835283041000366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6025272607803345, "epoch": 8.54, "learning_rate": 2.3919849718221667e-05, "loss": 0.6827, "step": 10105, "task_loss": 0.7730641961097717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22643622756004333, "epoch": 8.54, "learning_rate": 2.39167188478397e-05, "loss": 0.5675, "step": 10106, "task_loss": 0.25530001521110535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5551347732543945, "epoch": 8.54, "learning_rate": 2.3913587977457734e-05, "loss": 0.6922, "step": 10107, "task_loss": 0.49050480127334595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7737491726875305, "epoch": 8.54, "learning_rate": 2.391045710707577e-05, "loss": 0.4991, "step": 10108, "task_loss": 0.7795315980911255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5023193359375, "epoch": 8.54, "learning_rate": 2.3907326236693804e-05, "loss": 0.6547, "step": 10109, "task_loss": 0.05789393186569214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6899288296699524, "epoch": 8.55, "learning_rate": 2.3904195366311836e-05, "loss": 0.767, "step": 10110, "task_loss": 0.9030506610870361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28211578726768494, "epoch": 8.55, "learning_rate": 2.390106449592987e-05, "loss": 0.4335, "step": 10111, "task_loss": 0.5023936033248901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5140576362609863, "epoch": 8.55, "learning_rate": 2.3897933625547903e-05, "loss": 0.4648, "step": 10112, "task_loss": 0.505460798740387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40457332134246826, "epoch": 8.55, "learning_rate": 2.3894802755165938e-05, "loss": 0.4105, "step": 10113, "task_loss": 0.8458707928657532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6769025325775146, "epoch": 8.55, "learning_rate": 2.389167188478397e-05, "loss": 0.6678, "step": 10114, "task_loss": 1.7988958358764648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7296000719070435, "epoch": 8.55, "learning_rate": 2.3888541014402005e-05, "loss": 0.5592, "step": 10115, "task_loss": 1.114647626876831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4947141408920288, "epoch": 8.55, "learning_rate": 2.388541014402004e-05, "loss": 0.6461, "step": 10116, "task_loss": 1.4892443418502808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2927776873111725, "epoch": 8.55, "learning_rate": 2.3882279273638072e-05, "loss": 0.5758, "step": 10117, "task_loss": 2.2592933177948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5055782794952393, "epoch": 8.55, "learning_rate": 2.3879148403256107e-05, "loss": 0.5012, "step": 10118, "task_loss": 1.0046498775482178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7998346090316772, "epoch": 8.55, "learning_rate": 2.387601753287414e-05, "loss": 0.6496, "step": 10119, "task_loss": 0.6706708073616028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36393895745277405, "epoch": 8.55, "learning_rate": 2.3872886662492174e-05, "loss": 0.4271, "step": 10120, "task_loss": 0.6176459193229675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42988961935043335, "epoch": 8.56, "learning_rate": 2.3869755792110206e-05, "loss": 0.5664, "step": 10121, "task_loss": 0.7923341989517212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6605059504508972, "epoch": 8.56, "learning_rate": 2.386662492172824e-05, "loss": 0.5254, "step": 10122, "task_loss": 0.2701987624168396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39452454447746277, "epoch": 8.56, "learning_rate": 2.3863494051346273e-05, "loss": 0.4208, "step": 10123, "task_loss": 0.7490967512130737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5049225091934204, "epoch": 8.56, "learning_rate": 2.3860363180964308e-05, "loss": 0.5086, "step": 10124, "task_loss": 0.2576926052570343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2799944579601288, "epoch": 8.56, "learning_rate": 2.3857232310582343e-05, "loss": 0.5589, "step": 10125, "task_loss": 0.07580314576625824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5126856565475464, "epoch": 8.56, "learning_rate": 2.385410144020038e-05, "loss": 0.5531, "step": 10126, "task_loss": 0.8502635359764099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5862709879875183, "epoch": 8.56, "learning_rate": 2.385097056981841e-05, "loss": 0.5688, "step": 10127, "task_loss": 0.45631399750709534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29877597093582153, "epoch": 8.56, "learning_rate": 2.3847839699436445e-05, "loss": 0.4817, "step": 10128, "task_loss": 0.18970119953155518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3240800201892853, "epoch": 8.56, "learning_rate": 2.384470882905448e-05, "loss": 0.5062, "step": 10129, "task_loss": 0.36879992485046387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6782348155975342, "epoch": 8.56, "learning_rate": 2.3841577958672512e-05, "loss": 0.6933, "step": 10130, "task_loss": 0.5239977240562439 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4393279552459717, "epoch": 8.56, "learning_rate": 2.3838447088290548e-05, "loss": 0.6205, "step": 10131, "task_loss": 0.5055133700370789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6159020066261292, "epoch": 8.56, "learning_rate": 2.383531621790858e-05, "loss": 0.4791, "step": 10132, "task_loss": 1.5327608585357666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0358376502990723, "epoch": 8.57, "learning_rate": 2.3832185347526615e-05, "loss": 0.6785, "step": 10133, "task_loss": 0.6536751985549927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.564987063407898, "epoch": 8.57, "learning_rate": 2.3829054477144646e-05, "loss": 0.6899, "step": 10134, "task_loss": 0.6257984638214111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5802450776100159, "epoch": 8.57, "learning_rate": 2.382592360676268e-05, "loss": 0.6374, "step": 10135, "task_loss": 0.44522494077682495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7800324559211731, "epoch": 8.57, "learning_rate": 2.3822792736380713e-05, "loss": 0.6189, "step": 10136, "task_loss": 0.32860732078552246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5211220979690552, "epoch": 8.57, "learning_rate": 2.381966186599875e-05, "loss": 0.4826, "step": 10137, "task_loss": 0.2220020741224289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7728121876716614, "epoch": 8.57, "learning_rate": 2.381653099561678e-05, "loss": 0.5525, "step": 10138, "task_loss": 0.44184428453445435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.641999363899231, "epoch": 8.57, "learning_rate": 2.3813400125234815e-05, "loss": 0.6037, "step": 10139, "task_loss": 0.6626383066177368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3547142446041107, "epoch": 8.57, "learning_rate": 2.381026925485285e-05, "loss": 0.5461, "step": 10140, "task_loss": 0.6293083429336548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5259841084480286, "epoch": 8.57, "learning_rate": 2.3807138384470882e-05, "loss": 0.548, "step": 10141, "task_loss": 0.6014701128005981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4086444675922394, "epoch": 8.57, "learning_rate": 2.3804007514088918e-05, "loss": 0.3251, "step": 10142, "task_loss": 0.39229798316955566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.597890317440033, "epoch": 8.57, "learning_rate": 2.3800876643706953e-05, "loss": 0.6016, "step": 10143, "task_loss": 0.7161086797714233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5337141752243042, "epoch": 8.57, "learning_rate": 2.3797745773324988e-05, "loss": 0.4734, "step": 10144, "task_loss": 0.3432406187057495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3428012430667877, "epoch": 8.58, "learning_rate": 2.379461490294302e-05, "loss": 0.5376, "step": 10145, "task_loss": 0.3800116777420044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32495027780532837, "epoch": 8.58, "learning_rate": 2.3791484032561055e-05, "loss": 0.4864, "step": 10146, "task_loss": 0.24609117209911346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.156248688697815, "epoch": 8.58, "learning_rate": 2.3788353162179087e-05, "loss": 0.55, "step": 10147, "task_loss": 0.8708246350288391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2911372780799866, "epoch": 8.58, "learning_rate": 2.3785222291797122e-05, "loss": 0.6192, "step": 10148, "task_loss": 0.8181488513946533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34719133377075195, "epoch": 8.58, "learning_rate": 2.3782091421415154e-05, "loss": 0.5126, "step": 10149, "task_loss": 0.8037369251251221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9989386796951294, "epoch": 8.58, "learning_rate": 2.377896055103319e-05, "loss": 0.568, "step": 10150, "task_loss": 0.8667119741439819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4438078701496124, "epoch": 8.58, "learning_rate": 2.377582968065122e-05, "loss": 0.6297, "step": 10151, "task_loss": 0.38925495743751526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5794225931167603, "epoch": 8.58, "learning_rate": 2.3772698810269256e-05, "loss": 0.5005, "step": 10152, "task_loss": 0.6384279131889343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7459295988082886, "epoch": 8.58, "learning_rate": 2.376956793988729e-05, "loss": 0.7543, "step": 10153, "task_loss": 1.0766552686691284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5179533958435059, "epoch": 8.58, "learning_rate": 2.3766437069505323e-05, "loss": 0.5514, "step": 10154, "task_loss": 0.3289638161659241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6605609655380249, "epoch": 8.58, "learning_rate": 2.3763306199123358e-05, "loss": 0.702, "step": 10155, "task_loss": 0.48800450563430786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6476040482521057, "epoch": 8.58, "learning_rate": 2.376017532874139e-05, "loss": 0.6308, "step": 10156, "task_loss": 0.7089284062385559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5036085844039917, "epoch": 8.59, "learning_rate": 2.3757044458359425e-05, "loss": 0.6419, "step": 10157, "task_loss": 0.15392273664474487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5704149007797241, "epoch": 8.59, "learning_rate": 2.375391358797746e-05, "loss": 0.4624, "step": 10158, "task_loss": 0.7008436322212219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49258989095687866, "epoch": 8.59, "learning_rate": 2.375078271759549e-05, "loss": 0.5904, "step": 10159, "task_loss": 0.6355120539665222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6202499270439148, "epoch": 8.59, "learning_rate": 2.3747651847213527e-05, "loss": 0.5779, "step": 10160, "task_loss": 0.6147964000701904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46213942766189575, "epoch": 8.59, "learning_rate": 2.3744520976831562e-05, "loss": 0.6163, "step": 10161, "task_loss": 0.4554673135280609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7612292766571045, "epoch": 8.59, "learning_rate": 2.3741390106449594e-05, "loss": 0.6354, "step": 10162, "task_loss": 1.3618316650390625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5611318349838257, "epoch": 8.59, "learning_rate": 2.373825923606763e-05, "loss": 0.5422, "step": 10163, "task_loss": 0.870278000831604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4214290380477905, "epoch": 8.59, "learning_rate": 2.373512836568566e-05, "loss": 0.4682, "step": 10164, "task_loss": 0.4936971068382263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5270121097564697, "epoch": 8.59, "learning_rate": 2.3731997495303696e-05, "loss": 0.5984, "step": 10165, "task_loss": 0.9372133016586304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2579607367515564, "epoch": 8.59, "learning_rate": 2.372886662492173e-05, "loss": 0.4882, "step": 10166, "task_loss": 0.21810606122016907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44191205501556396, "epoch": 8.59, "learning_rate": 2.3725735754539763e-05, "loss": 0.4483, "step": 10167, "task_loss": 0.35128718614578247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6154458522796631, "epoch": 8.59, "learning_rate": 2.3722604884157798e-05, "loss": 0.4887, "step": 10168, "task_loss": 0.7379226684570312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6482371091842651, "epoch": 8.6, "learning_rate": 2.371947401377583e-05, "loss": 0.6923, "step": 10169, "task_loss": 0.8343385457992554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7869358062744141, "epoch": 8.6, "learning_rate": 2.3716343143393865e-05, "loss": 0.6943, "step": 10170, "task_loss": 1.2994048595428467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5930205583572388, "epoch": 8.6, "learning_rate": 2.3713212273011897e-05, "loss": 0.6406, "step": 10171, "task_loss": 0.4785839319229126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6378641724586487, "epoch": 8.6, "learning_rate": 2.3710081402629932e-05, "loss": 0.5762, "step": 10172, "task_loss": 0.7836578488349915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.61504065990448, "epoch": 8.6, "learning_rate": 2.3706950532247964e-05, "loss": 0.6354, "step": 10173, "task_loss": 0.23387162387371063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48103588819503784, "epoch": 8.6, "learning_rate": 2.3703819661866e-05, "loss": 0.6106, "step": 10174, "task_loss": 1.0581769943237305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3434467613697052, "epoch": 8.6, "learning_rate": 2.3700688791484034e-05, "loss": 0.4765, "step": 10175, "task_loss": 0.19560149312019348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7814637422561646, "epoch": 8.6, "learning_rate": 2.369755792110207e-05, "loss": 0.6161, "step": 10176, "task_loss": 0.5739794373512268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7519869804382324, "epoch": 8.6, "learning_rate": 2.36944270507201e-05, "loss": 0.792, "step": 10177, "task_loss": 0.6717224717140198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4525953531265259, "epoch": 8.6, "learning_rate": 2.3691296180338136e-05, "loss": 0.5494, "step": 10178, "task_loss": 0.7758942246437073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5517619252204895, "epoch": 8.6, "learning_rate": 2.368816530995617e-05, "loss": 0.5368, "step": 10179, "task_loss": 0.23561213910579681 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44188347458839417, "epoch": 8.6, "learning_rate": 2.3685034439574203e-05, "loss": 0.6238, "step": 10180, "task_loss": 0.07732204347848892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0012847185134888, "epoch": 8.61, "learning_rate": 2.3681903569192238e-05, "loss": 0.7495, "step": 10181, "task_loss": 0.9096921682357788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7883697748184204, "epoch": 8.61, "learning_rate": 2.367877269881027e-05, "loss": 0.761, "step": 10182, "task_loss": 0.9211120009422302 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5370867252349854, "epoch": 8.61, "learning_rate": 2.3675641828428305e-05, "loss": 0.4746, "step": 10183, "task_loss": 0.7954909801483154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6515308618545532, "epoch": 8.61, "learning_rate": 2.3672510958046337e-05, "loss": 0.6392, "step": 10184, "task_loss": 0.49216729402542114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6592167615890503, "epoch": 8.61, "learning_rate": 2.3669380087664372e-05, "loss": 0.4646, "step": 10185, "task_loss": 1.0732494592666626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48789912462234497, "epoch": 8.61, "learning_rate": 2.3666249217282404e-05, "loss": 0.6219, "step": 10186, "task_loss": 0.7316896915435791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3229483962059021, "epoch": 8.61, "learning_rate": 2.366311834690044e-05, "loss": 0.4998, "step": 10187, "task_loss": 0.8267841339111328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3462075889110565, "epoch": 8.61, "learning_rate": 2.365998747651847e-05, "loss": 0.5374, "step": 10188, "task_loss": 0.6354756355285645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.606490433216095, "epoch": 8.61, "learning_rate": 2.3656856606136506e-05, "loss": 0.6793, "step": 10189, "task_loss": 0.2299000322818756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6095920205116272, "epoch": 8.61, "learning_rate": 2.365372573575454e-05, "loss": 0.6887, "step": 10190, "task_loss": 1.418766975402832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7626671195030212, "epoch": 8.61, "learning_rate": 2.3650594865372573e-05, "loss": 0.5522, "step": 10191, "task_loss": 0.5805853009223938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5445733070373535, "epoch": 8.61, "learning_rate": 2.3647463994990608e-05, "loss": 0.5416, "step": 10192, "task_loss": 0.25713858008384705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45664137601852417, "epoch": 8.62, "learning_rate": 2.3644333124608643e-05, "loss": 0.5857, "step": 10193, "task_loss": 0.7254993915557861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5652408599853516, "epoch": 8.62, "learning_rate": 2.364120225422668e-05, "loss": 0.7031, "step": 10194, "task_loss": 0.9663349986076355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5212764739990234, "epoch": 8.62, "learning_rate": 2.363807138384471e-05, "loss": 0.602, "step": 10195, "task_loss": 1.1425092220306396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7307181358337402, "epoch": 8.62, "learning_rate": 2.3634940513462746e-05, "loss": 0.8597, "step": 10196, "task_loss": 0.764624297618866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3545001149177551, "epoch": 8.62, "learning_rate": 2.3631809643080777e-05, "loss": 0.6162, "step": 10197, "task_loss": 0.18777787685394287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.760377049446106, "epoch": 8.62, "learning_rate": 2.3628678772698812e-05, "loss": 0.5657, "step": 10198, "task_loss": 0.8651447296142578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5090301036834717, "epoch": 8.62, "learning_rate": 2.3625547902316844e-05, "loss": 0.7072, "step": 10199, "task_loss": 0.4183909296989441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.375724196434021, "epoch": 8.62, "learning_rate": 2.362241703193488e-05, "loss": 0.5832, "step": 10200, "task_loss": 0.9365507364273071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.755237340927124, "epoch": 8.62, "learning_rate": 2.361928616155291e-05, "loss": 0.6535, "step": 10201, "task_loss": 0.3650089204311371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9193294644355774, "epoch": 8.62, "learning_rate": 2.3616155291170946e-05, "loss": 0.6912, "step": 10202, "task_loss": 1.4943350553512573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5407570004463196, "epoch": 8.62, "learning_rate": 2.361302442078898e-05, "loss": 0.686, "step": 10203, "task_loss": 0.919589102268219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7966090440750122, "epoch": 8.63, "learning_rate": 2.3609893550407013e-05, "loss": 0.7019, "step": 10204, "task_loss": 1.3591121435165405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.59217369556427, "epoch": 8.63, "learning_rate": 2.360676268002505e-05, "loss": 0.7158, "step": 10205, "task_loss": 0.7182009220123291 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3928186893463135, "epoch": 8.63, "learning_rate": 2.360363180964308e-05, "loss": 0.4948, "step": 10206, "task_loss": 1.3313243389129639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5183713436126709, "epoch": 8.63, "learning_rate": 2.3600500939261115e-05, "loss": 0.7336, "step": 10207, "task_loss": 0.31375083327293396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5726734399795532, "epoch": 8.63, "learning_rate": 2.3597370068879147e-05, "loss": 0.683, "step": 10208, "task_loss": 0.90077805519104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6111127734184265, "epoch": 8.63, "learning_rate": 2.3594239198497182e-05, "loss": 0.5395, "step": 10209, "task_loss": 1.7960373163223267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49353837966918945, "epoch": 8.63, "learning_rate": 2.3591108328115218e-05, "loss": 0.5678, "step": 10210, "task_loss": 1.1405293941497803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48263102769851685, "epoch": 8.63, "learning_rate": 2.3587977457733253e-05, "loss": 0.671, "step": 10211, "task_loss": 1.5126467943191528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5896906852722168, "epoch": 8.63, "learning_rate": 2.3584846587351284e-05, "loss": 0.6967, "step": 10212, "task_loss": 0.8166379332542419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43555283546447754, "epoch": 8.63, "learning_rate": 2.358171571696932e-05, "loss": 0.5345, "step": 10213, "task_loss": 0.969637393951416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6552468538284302, "epoch": 8.63, "learning_rate": 2.3578584846587355e-05, "loss": 0.7621, "step": 10214, "task_loss": 0.32959964871406555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9020136594772339, "epoch": 8.63, "learning_rate": 2.3575453976205387e-05, "loss": 0.7742, "step": 10215, "task_loss": 1.217661738395691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8649163246154785, "epoch": 8.64, "learning_rate": 2.3572323105823422e-05, "loss": 0.6599, "step": 10216, "task_loss": 0.7724397778511047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4835333526134491, "epoch": 8.64, "learning_rate": 2.3569192235441454e-05, "loss": 0.5913, "step": 10217, "task_loss": 0.11627571284770966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7342992424964905, "epoch": 8.64, "learning_rate": 2.356606136505949e-05, "loss": 0.7337, "step": 10218, "task_loss": 1.1766886711120605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6544406414031982, "epoch": 8.64, "learning_rate": 2.356293049467752e-05, "loss": 0.6405, "step": 10219, "task_loss": 0.5235410928726196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6336920261383057, "epoch": 8.64, "learning_rate": 2.3559799624295556e-05, "loss": 0.5363, "step": 10220, "task_loss": 0.8673962354660034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5932207703590393, "epoch": 8.64, "learning_rate": 2.3556668753913587e-05, "loss": 0.5586, "step": 10221, "task_loss": 1.0572631359100342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3851572275161743, "epoch": 8.64, "learning_rate": 2.3553537883531623e-05, "loss": 0.5198, "step": 10222, "task_loss": 0.928636372089386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3696083128452301, "epoch": 8.64, "learning_rate": 2.3550407013149654e-05, "loss": 0.5065, "step": 10223, "task_loss": 0.6655989289283752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6929459571838379, "epoch": 8.64, "learning_rate": 2.354727614276769e-05, "loss": 0.5984, "step": 10224, "task_loss": 0.2993921637535095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41333869099617004, "epoch": 8.64, "learning_rate": 2.3544145272385725e-05, "loss": 0.6409, "step": 10225, "task_loss": 0.5133637189865112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40386107563972473, "epoch": 8.64, "learning_rate": 2.3541014402003757e-05, "loss": 0.6907, "step": 10226, "task_loss": 0.06004468351602554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4937637746334076, "epoch": 8.64, "learning_rate": 2.3537883531621792e-05, "loss": 0.5789, "step": 10227, "task_loss": 1.2446749210357666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5677509307861328, "epoch": 8.65, "learning_rate": 2.3534752661239827e-05, "loss": 0.5158, "step": 10228, "task_loss": 0.4868806004524231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.445784330368042, "epoch": 8.65, "learning_rate": 2.3531621790857862e-05, "loss": 0.5756, "step": 10229, "task_loss": 0.6537379026412964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8189799189567566, "epoch": 8.65, "learning_rate": 2.3528490920475894e-05, "loss": 0.786, "step": 10230, "task_loss": 1.1868191957473755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3963119089603424, "epoch": 8.65, "learning_rate": 2.352536005009393e-05, "loss": 0.6197, "step": 10231, "task_loss": 0.4576316475868225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.365081250667572, "epoch": 8.65, "learning_rate": 2.352222917971196e-05, "loss": 0.6358, "step": 10232, "task_loss": 0.42643409967422485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8328432440757751, "epoch": 8.65, "learning_rate": 2.3519098309329996e-05, "loss": 0.5451, "step": 10233, "task_loss": 0.898066520690918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5133947134017944, "epoch": 8.65, "learning_rate": 2.3515967438948028e-05, "loss": 0.6584, "step": 10234, "task_loss": 0.7849518060684204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7227649688720703, "epoch": 8.65, "learning_rate": 2.3512836568566063e-05, "loss": 0.6963, "step": 10235, "task_loss": 1.0644264221191406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43648916482925415, "epoch": 8.65, "learning_rate": 2.3509705698184095e-05, "loss": 0.4462, "step": 10236, "task_loss": 1.2865588665008545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23773285746574402, "epoch": 8.65, "learning_rate": 2.350657482780213e-05, "loss": 0.4787, "step": 10237, "task_loss": 0.23250730335712433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9015463590621948, "epoch": 8.65, "learning_rate": 2.350344395742016e-05, "loss": 0.6142, "step": 10238, "task_loss": 0.20014029741287231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6011427640914917, "epoch": 8.65, "learning_rate": 2.3500313087038197e-05, "loss": 0.5275, "step": 10239, "task_loss": 0.6751869320869446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5382883548736572, "epoch": 8.66, "learning_rate": 2.3497182216656232e-05, "loss": 0.6847, "step": 10240, "task_loss": 1.5955243110656738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4175835847854614, "epoch": 8.66, "learning_rate": 2.3494051346274264e-05, "loss": 0.6096, "step": 10241, "task_loss": 0.07652056962251663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6698815822601318, "epoch": 8.66, "learning_rate": 2.34909204758923e-05, "loss": 0.6078, "step": 10242, "task_loss": 0.4699294865131378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5716478824615479, "epoch": 8.66, "learning_rate": 2.3487789605510334e-05, "loss": 0.4836, "step": 10243, "task_loss": 0.2379196286201477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0789819955825806, "epoch": 8.66, "learning_rate": 2.348465873512837e-05, "loss": 0.8194, "step": 10244, "task_loss": 1.114412784576416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44963499903678894, "epoch": 8.66, "learning_rate": 2.34815278647464e-05, "loss": 0.505, "step": 10245, "task_loss": 0.5425290465354919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21674859523773193, "epoch": 8.66, "learning_rate": 2.3478396994364436e-05, "loss": 0.6226, "step": 10246, "task_loss": 0.6663046479225159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6089035272598267, "epoch": 8.66, "learning_rate": 2.3475266123982468e-05, "loss": 0.673, "step": 10247, "task_loss": 2.0963926315307617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5344479084014893, "epoch": 8.66, "learning_rate": 2.3472135253600503e-05, "loss": 0.6988, "step": 10248, "task_loss": 0.798233687877655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43697845935821533, "epoch": 8.66, "learning_rate": 2.3469004383218535e-05, "loss": 0.4569, "step": 10249, "task_loss": 0.17511732876300812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48328661918640137, "epoch": 8.66, "learning_rate": 2.346587351283657e-05, "loss": 0.5402, "step": 10250, "task_loss": 0.9041722416877747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48356392979621887, "epoch": 8.66, "learning_rate": 2.3462742642454605e-05, "loss": 0.6926, "step": 10251, "task_loss": 0.11360824108123779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5112522840499878, "epoch": 8.67, "learning_rate": 2.3459611772072637e-05, "loss": 0.7173, "step": 10252, "task_loss": 0.7050277590751648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6018706560134888, "epoch": 8.67, "learning_rate": 2.3456480901690672e-05, "loss": 0.6045, "step": 10253, "task_loss": 1.2196855545043945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3503519296646118, "epoch": 8.67, "learning_rate": 2.3453350031308704e-05, "loss": 0.5124, "step": 10254, "task_loss": 0.38163700699806213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.575977087020874, "epoch": 8.67, "learning_rate": 2.345021916092674e-05, "loss": 0.5124, "step": 10255, "task_loss": 0.8662415146827698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.775056004524231, "epoch": 8.67, "learning_rate": 2.344708829054477e-05, "loss": 0.5932, "step": 10256, "task_loss": 0.7554023265838623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5191013813018799, "epoch": 8.67, "learning_rate": 2.3443957420162806e-05, "loss": 0.6826, "step": 10257, "task_loss": 0.7773323059082031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5732079744338989, "epoch": 8.67, "learning_rate": 2.3440826549780838e-05, "loss": 0.6976, "step": 10258, "task_loss": 0.4773746430873871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38019293546676636, "epoch": 8.67, "learning_rate": 2.3437695679398873e-05, "loss": 0.4847, "step": 10259, "task_loss": 0.2062562108039856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41675451397895813, "epoch": 8.67, "learning_rate": 2.3434564809016908e-05, "loss": 0.4836, "step": 10260, "task_loss": 0.39620476961135864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6373957395553589, "epoch": 8.67, "learning_rate": 2.3431433938634943e-05, "loss": 0.5364, "step": 10261, "task_loss": 1.0597769021987915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48288804292678833, "epoch": 8.67, "learning_rate": 2.3428303068252975e-05, "loss": 0.5169, "step": 10262, "task_loss": 0.35738739371299744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37541988492012024, "epoch": 8.67, "learning_rate": 2.342517219787101e-05, "loss": 0.4372, "step": 10263, "task_loss": 0.31247490644454956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36405372619628906, "epoch": 8.68, "learning_rate": 2.3422041327489046e-05, "loss": 0.4902, "step": 10264, "task_loss": 0.6912161707878113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.106175422668457, "epoch": 8.68, "learning_rate": 2.3418910457107077e-05, "loss": 0.7215, "step": 10265, "task_loss": 1.5190690755844116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6646531820297241, "epoch": 8.68, "learning_rate": 2.3415779586725112e-05, "loss": 0.5573, "step": 10266, "task_loss": 1.307455062866211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6335834860801697, "epoch": 8.68, "learning_rate": 2.3412648716343144e-05, "loss": 0.5666, "step": 10267, "task_loss": 0.8601988554000854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6858842968940735, "epoch": 8.68, "learning_rate": 2.340951784596118e-05, "loss": 0.5506, "step": 10268, "task_loss": 1.8724411725997925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37020719051361084, "epoch": 8.68, "learning_rate": 2.340638697557921e-05, "loss": 0.5036, "step": 10269, "task_loss": 0.38777589797973633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8541181087493896, "epoch": 8.68, "learning_rate": 2.3403256105197246e-05, "loss": 0.5353, "step": 10270, "task_loss": 0.971590518951416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4853457808494568, "epoch": 8.68, "learning_rate": 2.3400125234815278e-05, "loss": 0.501, "step": 10271, "task_loss": 0.8051766753196716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40960627794265747, "epoch": 8.68, "learning_rate": 2.3396994364433313e-05, "loss": 0.455, "step": 10272, "task_loss": 0.75418621301651 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4105125665664673, "epoch": 8.68, "learning_rate": 2.3393863494051345e-05, "loss": 0.611, "step": 10273, "task_loss": 2.1385154724121094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34540629386901855, "epoch": 8.68, "learning_rate": 2.339073262366938e-05, "loss": 0.5775, "step": 10274, "task_loss": 0.14466191828250885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2903979420661926, "epoch": 8.69, "learning_rate": 2.3387601753287412e-05, "loss": 0.537, "step": 10275, "task_loss": 0.2829439640045166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5231016874313354, "epoch": 8.69, "learning_rate": 2.3384470882905447e-05, "loss": 0.53, "step": 10276, "task_loss": 0.5824772119522095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6705288290977478, "epoch": 8.69, "learning_rate": 2.3381340012523482e-05, "loss": 0.6924, "step": 10277, "task_loss": 1.088320255279541 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7023051977157593, "epoch": 8.69, "learning_rate": 2.3378209142141518e-05, "loss": 0.5876, "step": 10278, "task_loss": 1.2503266334533691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5109679698944092, "epoch": 8.69, "learning_rate": 2.3375078271759553e-05, "loss": 0.4983, "step": 10279, "task_loss": 0.5295658707618713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6716022491455078, "epoch": 8.69, "learning_rate": 2.3371947401377585e-05, "loss": 0.824, "step": 10280, "task_loss": 2.075610399246216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8557205200195312, "epoch": 8.69, "learning_rate": 2.336881653099562e-05, "loss": 0.7601, "step": 10281, "task_loss": 0.988161027431488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5889851450920105, "epoch": 8.69, "learning_rate": 2.336568566061365e-05, "loss": 0.5716, "step": 10282, "task_loss": 0.9852766990661621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24255861341953278, "epoch": 8.69, "learning_rate": 2.3362554790231687e-05, "loss": 0.6842, "step": 10283, "task_loss": 0.3701871931552887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5441357493400574, "epoch": 8.69, "learning_rate": 2.335942391984972e-05, "loss": 0.4441, "step": 10284, "task_loss": 0.46320801973342896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46973317861557007, "epoch": 8.69, "learning_rate": 2.3356293049467754e-05, "loss": 0.6034, "step": 10285, "task_loss": 0.7599964141845703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42817240953445435, "epoch": 8.69, "learning_rate": 2.3353162179085785e-05, "loss": 0.4884, "step": 10286, "task_loss": 0.25398361682891846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5245505571365356, "epoch": 8.7, "learning_rate": 2.335003130870382e-05, "loss": 0.7992, "step": 10287, "task_loss": 0.449176162481308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40319257974624634, "epoch": 8.7, "learning_rate": 2.3346900438321856e-05, "loss": 0.6094, "step": 10288, "task_loss": 1.0919468402862549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44393983483314514, "epoch": 8.7, "learning_rate": 2.3343769567939888e-05, "loss": 0.5004, "step": 10289, "task_loss": 0.5630022287368774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.717318058013916, "epoch": 8.7, "learning_rate": 2.3340638697557923e-05, "loss": 0.5617, "step": 10290, "task_loss": 0.705776035785675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7217721939086914, "epoch": 8.7, "learning_rate": 2.3337507827175954e-05, "loss": 0.6057, "step": 10291, "task_loss": 1.0932133197784424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.939380943775177, "epoch": 8.7, "learning_rate": 2.333437695679399e-05, "loss": 0.7829, "step": 10292, "task_loss": 1.6592435836791992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3660343289375305, "epoch": 8.7, "learning_rate": 2.333124608641202e-05, "loss": 0.6972, "step": 10293, "task_loss": 0.5820684432983398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3937099874019623, "epoch": 8.7, "learning_rate": 2.3328115216030057e-05, "loss": 0.7085, "step": 10294, "task_loss": 0.04206370562314987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2999074459075928, "epoch": 8.7, "learning_rate": 2.3324984345648092e-05, "loss": 0.546, "step": 10295, "task_loss": 0.06856942176818848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31668537855148315, "epoch": 8.7, "learning_rate": 2.3321853475266127e-05, "loss": 0.5801, "step": 10296, "task_loss": 0.6736817955970764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7234536409378052, "epoch": 8.7, "learning_rate": 2.331872260488416e-05, "loss": 0.8226, "step": 10297, "task_loss": 1.2139898538589478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5414732694625854, "epoch": 8.7, "learning_rate": 2.3315591734502194e-05, "loss": 0.4729, "step": 10298, "task_loss": 0.4459790289402008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7360880374908447, "epoch": 8.71, "learning_rate": 2.3312460864120226e-05, "loss": 0.6093, "step": 10299, "task_loss": 0.4897862672805786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.497170090675354, "epoch": 8.71, "learning_rate": 2.330932999373826e-05, "loss": 0.4692, "step": 10300, "task_loss": 0.828575074672699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5708574652671814, "epoch": 8.71, "learning_rate": 2.3306199123356296e-05, "loss": 0.6296, "step": 10301, "task_loss": 0.49792754650115967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8382937908172607, "epoch": 8.71, "learning_rate": 2.3303068252974328e-05, "loss": 0.705, "step": 10302, "task_loss": 0.31073951721191406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30283886194229126, "epoch": 8.71, "learning_rate": 2.3299937382592363e-05, "loss": 0.6858, "step": 10303, "task_loss": 0.44411832094192505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196468830108643, "epoch": 8.71, "learning_rate": 2.3296806512210395e-05, "loss": 0.4825, "step": 10304, "task_loss": 0.7928723096847534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6245046854019165, "epoch": 8.71, "learning_rate": 2.329367564182843e-05, "loss": 0.662, "step": 10305, "task_loss": 1.640119194984436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46317601203918457, "epoch": 8.71, "learning_rate": 2.329054477144646e-05, "loss": 0.7112, "step": 10306, "task_loss": 0.7012359499931335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6978142857551575, "epoch": 8.71, "learning_rate": 2.3287413901064497e-05, "loss": 0.7597, "step": 10307, "task_loss": 0.7923853993415833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35569173097610474, "epoch": 8.71, "learning_rate": 2.328428303068253e-05, "loss": 0.6559, "step": 10308, "task_loss": 0.6077647805213928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5507012605667114, "epoch": 8.71, "learning_rate": 2.3281152160300564e-05, "loss": 0.5437, "step": 10309, "task_loss": 0.7900258302688599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4463390111923218, "epoch": 8.71, "learning_rate": 2.32780212899186e-05, "loss": 0.5661, "step": 10310, "task_loss": 0.8078454732894897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37847432494163513, "epoch": 8.72, "learning_rate": 2.327489041953663e-05, "loss": 0.5449, "step": 10311, "task_loss": 0.43028491735458374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4232290983200073, "epoch": 8.72, "learning_rate": 2.3271759549154666e-05, "loss": 0.5566, "step": 10312, "task_loss": 0.6030418276786804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8381743431091309, "epoch": 8.72, "learning_rate": 2.32686286787727e-05, "loss": 0.5797, "step": 10313, "task_loss": 0.6536415219306946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5811944007873535, "epoch": 8.72, "learning_rate": 2.3265497808390736e-05, "loss": 0.4832, "step": 10314, "task_loss": 0.15809014439582825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48007017374038696, "epoch": 8.72, "learning_rate": 2.3262366938008768e-05, "loss": 0.4273, "step": 10315, "task_loss": 1.1659356355667114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5563874244689941, "epoch": 8.72, "learning_rate": 2.3259236067626803e-05, "loss": 0.542, "step": 10316, "task_loss": 0.7981234788894653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3422144651412964, "epoch": 8.72, "learning_rate": 2.3256105197244835e-05, "loss": 0.5809, "step": 10317, "task_loss": 0.95307457447052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8381541967391968, "epoch": 8.72, "learning_rate": 2.325297432686287e-05, "loss": 0.5826, "step": 10318, "task_loss": 0.9266279339790344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35719576478004456, "epoch": 8.72, "learning_rate": 2.3249843456480902e-05, "loss": 0.361, "step": 10319, "task_loss": 0.5242159962654114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3851388096809387, "epoch": 8.72, "learning_rate": 2.3246712586098937e-05, "loss": 0.5971, "step": 10320, "task_loss": 0.9548928141593933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.714460015296936, "epoch": 8.72, "learning_rate": 2.324358171571697e-05, "loss": 0.6114, "step": 10321, "task_loss": 0.8213605880737305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6149166822433472, "epoch": 8.72, "learning_rate": 2.3240450845335004e-05, "loss": 0.5568, "step": 10322, "task_loss": 0.5643141269683838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5283236503601074, "epoch": 8.73, "learning_rate": 2.3237319974953036e-05, "loss": 0.7838, "step": 10323, "task_loss": 0.45607781410217285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6325607299804688, "epoch": 8.73, "learning_rate": 2.323418910457107e-05, "loss": 0.605, "step": 10324, "task_loss": 0.9082795977592468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3709324300289154, "epoch": 8.73, "learning_rate": 2.3231058234189106e-05, "loss": 0.4665, "step": 10325, "task_loss": 0.2649557888507843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5796202421188354, "epoch": 8.73, "learning_rate": 2.3227927363807138e-05, "loss": 0.9118, "step": 10326, "task_loss": 1.5309813022613525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8744503259658813, "epoch": 8.73, "learning_rate": 2.3224796493425173e-05, "loss": 0.5709, "step": 10327, "task_loss": 0.9650465846061707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3466658890247345, "epoch": 8.73, "learning_rate": 2.3221665623043208e-05, "loss": 0.6257, "step": 10328, "task_loss": 0.6811246275901794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6590518951416016, "epoch": 8.73, "learning_rate": 2.3218534752661243e-05, "loss": 0.6909, "step": 10329, "task_loss": 0.8892730474472046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9255754947662354, "epoch": 8.73, "learning_rate": 2.3215403882279275e-05, "loss": 0.6387, "step": 10330, "task_loss": 0.7700848579406738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9049198627471924, "epoch": 8.73, "learning_rate": 2.321227301189731e-05, "loss": 0.7575, "step": 10331, "task_loss": 0.6786505579948425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31144559383392334, "epoch": 8.73, "learning_rate": 2.3209142141515342e-05, "loss": 0.5027, "step": 10332, "task_loss": 0.8281727433204651 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48746463656425476, "epoch": 8.73, "learning_rate": 2.3206011271133377e-05, "loss": 0.5726, "step": 10333, "task_loss": 0.35931044816970825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46268534660339355, "epoch": 8.73, "learning_rate": 2.320288040075141e-05, "loss": 0.5789, "step": 10334, "task_loss": 0.3268965184688568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34380385279655457, "epoch": 8.74, "learning_rate": 2.3199749530369444e-05, "loss": 0.4592, "step": 10335, "task_loss": 0.2025119960308075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5249941349029541, "epoch": 8.74, "learning_rate": 2.3196618659987476e-05, "loss": 0.6895, "step": 10336, "task_loss": 1.230987310409546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1641499996185303, "epoch": 8.74, "learning_rate": 2.319348778960551e-05, "loss": 0.6884, "step": 10337, "task_loss": 1.3084943294525146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7166554927825928, "epoch": 8.74, "learning_rate": 2.3190356919223546e-05, "loss": 0.5877, "step": 10338, "task_loss": 1.3732205629348755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8807579874992371, "epoch": 8.74, "learning_rate": 2.3187226048841578e-05, "loss": 0.6876, "step": 10339, "task_loss": 2.1478803157806396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0290751457214355, "epoch": 8.74, "learning_rate": 2.3184095178459613e-05, "loss": 0.8892, "step": 10340, "task_loss": 0.6659360527992249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4846385717391968, "epoch": 8.74, "learning_rate": 2.3180964308077645e-05, "loss": 0.4953, "step": 10341, "task_loss": 0.9364334344863892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5379605889320374, "epoch": 8.74, "learning_rate": 2.317783343769568e-05, "loss": 0.5873, "step": 10342, "task_loss": 0.419034481048584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5499951839447021, "epoch": 8.74, "learning_rate": 2.3174702567313712e-05, "loss": 0.7221, "step": 10343, "task_loss": 0.5237514972686768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5779290795326233, "epoch": 8.74, "learning_rate": 2.3171571696931747e-05, "loss": 0.7022, "step": 10344, "task_loss": 1.2051113843917847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5470089912414551, "epoch": 8.74, "learning_rate": 2.3168440826549782e-05, "loss": 0.7055, "step": 10345, "task_loss": 0.6371253132820129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5930525064468384, "epoch": 8.75, "learning_rate": 2.3165309956167818e-05, "loss": 0.7203, "step": 10346, "task_loss": 1.8512521982192993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36674487590789795, "epoch": 8.75, "learning_rate": 2.316217908578585e-05, "loss": 0.6722, "step": 10347, "task_loss": 0.5891098380088806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.955354630947113, "epoch": 8.75, "learning_rate": 2.3159048215403885e-05, "loss": 0.6454, "step": 10348, "task_loss": 0.4280858039855957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5089486241340637, "epoch": 8.75, "learning_rate": 2.315591734502192e-05, "loss": 0.6259, "step": 10349, "task_loss": 0.32104092836380005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.379539430141449, "epoch": 8.75, "learning_rate": 2.315278647463995e-05, "loss": 0.5399, "step": 10350, "task_loss": 0.3490810692310333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47443920373916626, "epoch": 8.75, "learning_rate": 2.3149655604257987e-05, "loss": 0.6687, "step": 10351, "task_loss": 1.2928767204284668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7143051028251648, "epoch": 8.75, "learning_rate": 2.314652473387602e-05, "loss": 0.5771, "step": 10352, "task_loss": 0.9570068717002869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6997722387313843, "epoch": 8.75, "learning_rate": 2.3143393863494054e-05, "loss": 0.8043, "step": 10353, "task_loss": 1.3641871213912964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.692549467086792, "epoch": 8.75, "learning_rate": 2.3140262993112085e-05, "loss": 0.7886, "step": 10354, "task_loss": 0.7015074491500854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8476186394691467, "epoch": 8.75, "learning_rate": 2.313713212273012e-05, "loss": 0.5652, "step": 10355, "task_loss": 1.2228002548217773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8332419991493225, "epoch": 8.75, "learning_rate": 2.3134001252348152e-05, "loss": 0.7051, "step": 10356, "task_loss": 0.9708960652351379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6420169472694397, "epoch": 8.75, "learning_rate": 2.3130870381966188e-05, "loss": 0.6006, "step": 10357, "task_loss": 0.7974494099617004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3863992691040039, "epoch": 8.76, "learning_rate": 2.312773951158422e-05, "loss": 0.5049, "step": 10358, "task_loss": 0.25285446643829346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5578740835189819, "epoch": 8.76, "learning_rate": 2.3124608641202254e-05, "loss": 0.6277, "step": 10359, "task_loss": 1.9122505187988281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7356958389282227, "epoch": 8.76, "learning_rate": 2.3121477770820286e-05, "loss": 0.4733, "step": 10360, "task_loss": 1.0393308401107788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38609281182289124, "epoch": 8.76, "learning_rate": 2.311834690043832e-05, "loss": 0.6162, "step": 10361, "task_loss": 0.1925063133239746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37191784381866455, "epoch": 8.76, "learning_rate": 2.3115216030056357e-05, "loss": 0.5569, "step": 10362, "task_loss": 0.9600260853767395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6537123322486877, "epoch": 8.76, "learning_rate": 2.3112085159674392e-05, "loss": 0.7665, "step": 10363, "task_loss": 0.7764924764633179 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3230307698249817, "epoch": 8.76, "learning_rate": 2.3108954289292427e-05, "loss": 0.562, "step": 10364, "task_loss": 0.22137141227722168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43140938878059387, "epoch": 8.76, "learning_rate": 2.310582341891046e-05, "loss": 0.6315, "step": 10365, "task_loss": 0.43146100640296936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5331190824508667, "epoch": 8.76, "learning_rate": 2.3102692548528494e-05, "loss": 0.5398, "step": 10366, "task_loss": 0.8127729296684265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6909347176551819, "epoch": 8.76, "learning_rate": 2.3099561678146526e-05, "loss": 0.5521, "step": 10367, "task_loss": 0.5541678667068481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4124220013618469, "epoch": 8.76, "learning_rate": 2.309643080776456e-05, "loss": 0.8441, "step": 10368, "task_loss": 0.27316808700561523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3350998163223267, "epoch": 8.76, "learning_rate": 2.3093299937382593e-05, "loss": 0.8253, "step": 10369, "task_loss": 1.588733434677124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6323912739753723, "epoch": 8.77, "learning_rate": 2.3090169067000628e-05, "loss": 0.5035, "step": 10370, "task_loss": 1.5203354358673096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30298763513565063, "epoch": 8.77, "learning_rate": 2.308703819661866e-05, "loss": 0.66, "step": 10371, "task_loss": 0.4342939555644989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4886600077152252, "epoch": 8.77, "learning_rate": 2.3083907326236695e-05, "loss": 0.5038, "step": 10372, "task_loss": 0.24454772472381592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6762925386428833, "epoch": 8.77, "learning_rate": 2.3080776455854727e-05, "loss": 0.5587, "step": 10373, "task_loss": 0.8766843676567078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3852609694004059, "epoch": 8.77, "learning_rate": 2.3077645585472762e-05, "loss": 0.4594, "step": 10374, "task_loss": 0.3457113206386566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6160033345222473, "epoch": 8.77, "learning_rate": 2.3074514715090797e-05, "loss": 0.5304, "step": 10375, "task_loss": 0.2684219181537628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5150679349899292, "epoch": 8.77, "learning_rate": 2.307138384470883e-05, "loss": 0.3969, "step": 10376, "task_loss": 0.4745836555957794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3950892686843872, "epoch": 8.77, "learning_rate": 2.3068252974326864e-05, "loss": 0.5512, "step": 10377, "task_loss": 0.3169928193092346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30836284160614014, "epoch": 8.77, "learning_rate": 2.3065122103944896e-05, "loss": 0.4878, "step": 10378, "task_loss": 0.09234236925840378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40844979882240295, "epoch": 8.77, "learning_rate": 2.306199123356293e-05, "loss": 0.4807, "step": 10379, "task_loss": 0.44722384214401245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4780873656272888, "epoch": 8.77, "learning_rate": 2.3058860363180966e-05, "loss": 0.556, "step": 10380, "task_loss": 1.2595882415771484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6175422072410583, "epoch": 8.77, "learning_rate": 2.3055729492799e-05, "loss": 0.5852, "step": 10381, "task_loss": 0.3905195891857147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2845686078071594, "epoch": 8.78, "learning_rate": 2.3052598622417033e-05, "loss": 0.4435, "step": 10382, "task_loss": 0.8841062188148499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030027627944946, "epoch": 8.78, "learning_rate": 2.3049467752035068e-05, "loss": 0.5746, "step": 10383, "task_loss": 0.7060868740081787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6126440763473511, "epoch": 8.78, "learning_rate": 2.30463368816531e-05, "loss": 0.5726, "step": 10384, "task_loss": 0.24917663633823395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6217581033706665, "epoch": 8.78, "learning_rate": 2.3043206011271135e-05, "loss": 0.5418, "step": 10385, "task_loss": 0.5350740551948547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6811140179634094, "epoch": 8.78, "learning_rate": 2.304007514088917e-05, "loss": 0.6088, "step": 10386, "task_loss": 0.7554417848587036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0173475742340088, "epoch": 8.78, "learning_rate": 2.3036944270507202e-05, "loss": 0.5678, "step": 10387, "task_loss": 1.1980688571929932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21973145008087158, "epoch": 8.78, "learning_rate": 2.3033813400125237e-05, "loss": 0.4085, "step": 10388, "task_loss": 0.30830666422843933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6924514770507812, "epoch": 8.78, "learning_rate": 2.303068252974327e-05, "loss": 0.7874, "step": 10389, "task_loss": 1.0641162395477295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.985315203666687, "epoch": 8.78, "learning_rate": 2.3027551659361304e-05, "loss": 0.7095, "step": 10390, "task_loss": 0.6731477379798889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3559461236000061, "epoch": 8.78, "learning_rate": 2.3024420788979336e-05, "loss": 0.5386, "step": 10391, "task_loss": 0.5954186916351318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36798295378685, "epoch": 8.78, "learning_rate": 2.302128991859737e-05, "loss": 0.4759, "step": 10392, "task_loss": 0.628623366355896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7061762809753418, "epoch": 8.78, "learning_rate": 2.3018159048215403e-05, "loss": 0.5152, "step": 10393, "task_loss": 0.7895524501800537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39230602979660034, "epoch": 8.79, "learning_rate": 2.3015028177833438e-05, "loss": 0.6226, "step": 10394, "task_loss": 0.3581845462322235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.534704864025116, "epoch": 8.79, "learning_rate": 2.3011897307451473e-05, "loss": 0.4115, "step": 10395, "task_loss": 0.4903393089771271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45188677310943604, "epoch": 8.79, "learning_rate": 2.300876643706951e-05, "loss": 0.5527, "step": 10396, "task_loss": 0.3470648229122162 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7374054789543152, "epoch": 8.79, "learning_rate": 2.300563556668754e-05, "loss": 0.7796, "step": 10397, "task_loss": 0.5706772804260254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8489549160003662, "epoch": 8.79, "learning_rate": 2.3002504696305575e-05, "loss": 0.5696, "step": 10398, "task_loss": 0.22981072962284088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4644809663295746, "epoch": 8.79, "learning_rate": 2.299937382592361e-05, "loss": 0.602, "step": 10399, "task_loss": 0.4993766248226166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.702807605266571, "epoch": 8.79, "learning_rate": 2.2996242955541642e-05, "loss": 0.5768, "step": 10400, "task_loss": 1.3677982091903687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4047866761684418, "epoch": 8.79, "learning_rate": 2.2993112085159677e-05, "loss": 0.4902, "step": 10401, "task_loss": 0.5207951068878174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3349772095680237, "epoch": 8.79, "learning_rate": 2.298998121477771e-05, "loss": 0.5788, "step": 10402, "task_loss": 0.3992491364479065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36919423937797546, "epoch": 8.79, "learning_rate": 2.2986850344395744e-05, "loss": 0.5172, "step": 10403, "task_loss": 0.43982571363449097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.617461085319519, "epoch": 8.79, "learning_rate": 2.2983719474013776e-05, "loss": 0.5768, "step": 10404, "task_loss": 0.6602907776832581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7570123672485352, "epoch": 8.79, "learning_rate": 2.298058860363181e-05, "loss": 0.6989, "step": 10405, "task_loss": 0.5688965916633606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5079675316810608, "epoch": 8.8, "learning_rate": 2.2977457733249843e-05, "loss": 0.4478, "step": 10406, "task_loss": 1.8784152269363403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6953410506248474, "epoch": 8.8, "learning_rate": 2.2974326862867878e-05, "loss": 0.4968, "step": 10407, "task_loss": 0.766654372215271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2692314684391022, "epoch": 8.8, "learning_rate": 2.297119599248591e-05, "loss": 0.4501, "step": 10408, "task_loss": 0.6639240384101868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5535884499549866, "epoch": 8.8, "learning_rate": 2.2968065122103945e-05, "loss": 0.5346, "step": 10409, "task_loss": 0.2012121081352234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39217817783355713, "epoch": 8.8, "learning_rate": 2.2964934251721977e-05, "loss": 0.3439, "step": 10410, "task_loss": 0.22894737124443054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5869215726852417, "epoch": 8.8, "learning_rate": 2.2961803381340012e-05, "loss": 0.5956, "step": 10411, "task_loss": 1.0420713424682617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6027219295501709, "epoch": 8.8, "learning_rate": 2.2958672510958047e-05, "loss": 0.7561, "step": 10412, "task_loss": 1.8512935638427734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6672438979148865, "epoch": 8.8, "learning_rate": 2.2955541640576082e-05, "loss": 0.5567, "step": 10413, "task_loss": 0.8508520722389221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7304369211196899, "epoch": 8.8, "learning_rate": 2.2952410770194118e-05, "loss": 0.5593, "step": 10414, "task_loss": 0.764739453792572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6121387481689453, "epoch": 8.8, "learning_rate": 2.294927989981215e-05, "loss": 0.6845, "step": 10415, "task_loss": 1.1698776483535767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6708433032035828, "epoch": 8.8, "learning_rate": 2.2946149029430185e-05, "loss": 0.5895, "step": 10416, "task_loss": 0.09957325458526611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5544990301132202, "epoch": 8.81, "learning_rate": 2.2943018159048216e-05, "loss": 0.5826, "step": 10417, "task_loss": 1.5976804494857788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1258759498596191, "epoch": 8.81, "learning_rate": 2.293988728866625e-05, "loss": 0.7656, "step": 10418, "task_loss": 1.4027128219604492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6737262606620789, "epoch": 8.81, "learning_rate": 2.2936756418284283e-05, "loss": 0.6231, "step": 10419, "task_loss": 0.41880524158477783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4934385418891907, "epoch": 8.81, "learning_rate": 2.293362554790232e-05, "loss": 0.583, "step": 10420, "task_loss": 1.0891083478927612 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6571741104125977, "epoch": 8.81, "learning_rate": 2.293049467752035e-05, "loss": 0.514, "step": 10421, "task_loss": 0.5284713506698608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20996138453483582, "epoch": 8.81, "learning_rate": 2.2927363807138385e-05, "loss": 0.4914, "step": 10422, "task_loss": 0.6232519149780273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34344664216041565, "epoch": 8.81, "learning_rate": 2.292423293675642e-05, "loss": 0.605, "step": 10423, "task_loss": 0.7824487686157227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9866288304328918, "epoch": 8.81, "learning_rate": 2.2921102066374452e-05, "loss": 0.6569, "step": 10424, "task_loss": 0.440986692905426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2630934417247772, "epoch": 8.81, "learning_rate": 2.2917971195992488e-05, "loss": 0.4356, "step": 10425, "task_loss": 0.44044554233551025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4033052921295166, "epoch": 8.81, "learning_rate": 2.291484032561052e-05, "loss": 0.6199, "step": 10426, "task_loss": 0.28292712569236755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5967134237289429, "epoch": 8.81, "learning_rate": 2.2911709455228555e-05, "loss": 0.5678, "step": 10427, "task_loss": 1.1858819723129272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5336145758628845, "epoch": 8.81, "learning_rate": 2.2908578584846586e-05, "loss": 0.547, "step": 10428, "task_loss": 0.8411793112754822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49962884187698364, "epoch": 8.82, "learning_rate": 2.290544771446462e-05, "loss": 0.4258, "step": 10429, "task_loss": 0.6744248867034912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0279333591461182, "epoch": 8.82, "learning_rate": 2.2902316844082657e-05, "loss": 0.6781, "step": 10430, "task_loss": 0.8679706454277039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3892497420310974, "epoch": 8.82, "learning_rate": 2.2899185973700692e-05, "loss": 0.4964, "step": 10431, "task_loss": 0.943821132183075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5360161066055298, "epoch": 8.82, "learning_rate": 2.2896055103318724e-05, "loss": 0.7096, "step": 10432, "task_loss": 1.128258466720581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5027453899383545, "epoch": 8.82, "learning_rate": 2.289292423293676e-05, "loss": 0.5644, "step": 10433, "task_loss": 0.41313421726226807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7338376045227051, "epoch": 8.82, "learning_rate": 2.288979336255479e-05, "loss": 0.4578, "step": 10434, "task_loss": 0.43217289447784424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33614733815193176, "epoch": 8.82, "learning_rate": 2.2886662492172826e-05, "loss": 0.5006, "step": 10435, "task_loss": 0.8646050095558167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46052443981170654, "epoch": 8.82, "learning_rate": 2.288353162179086e-05, "loss": 0.5788, "step": 10436, "task_loss": 0.5692204833030701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1146495342254639, "epoch": 8.82, "learning_rate": 2.2880400751408893e-05, "loss": 0.7403, "step": 10437, "task_loss": 1.8941160440444946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6668336391448975, "epoch": 8.82, "learning_rate": 2.2877269881026928e-05, "loss": 0.5534, "step": 10438, "task_loss": 1.4668835401535034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5229263305664062, "epoch": 8.82, "learning_rate": 2.287413901064496e-05, "loss": 0.5167, "step": 10439, "task_loss": 0.5080314874649048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7871178388595581, "epoch": 8.82, "learning_rate": 2.2871008140262995e-05, "loss": 0.5991, "step": 10440, "task_loss": 1.3417716026306152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6589687466621399, "epoch": 8.83, "learning_rate": 2.2867877269881027e-05, "loss": 0.5845, "step": 10441, "task_loss": 0.9358371496200562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4896763563156128, "epoch": 8.83, "learning_rate": 2.2864746399499062e-05, "loss": 0.6071, "step": 10442, "task_loss": 0.6223272681236267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7814544439315796, "epoch": 8.83, "learning_rate": 2.2861615529117094e-05, "loss": 0.5188, "step": 10443, "task_loss": 1.9767159223556519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5879467725753784, "epoch": 8.83, "learning_rate": 2.285848465873513e-05, "loss": 0.6965, "step": 10444, "task_loss": 0.69621342420578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.751725435256958, "epoch": 8.83, "learning_rate": 2.285535378835316e-05, "loss": 0.6899, "step": 10445, "task_loss": 0.2655404210090637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7325441241264343, "epoch": 8.83, "learning_rate": 2.2852222917971196e-05, "loss": 0.5208, "step": 10446, "task_loss": 0.36884403228759766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5618202686309814, "epoch": 8.83, "learning_rate": 2.284909204758923e-05, "loss": 0.7542, "step": 10447, "task_loss": 1.0324609279632568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6579235792160034, "epoch": 8.83, "learning_rate": 2.2845961177207266e-05, "loss": 0.5558, "step": 10448, "task_loss": 0.3519543707370758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5221633911132812, "epoch": 8.83, "learning_rate": 2.28428303068253e-05, "loss": 0.6164, "step": 10449, "task_loss": 0.6858883500099182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7290145754814148, "epoch": 8.83, "learning_rate": 2.2839699436443333e-05, "loss": 0.7059, "step": 10450, "task_loss": 0.8614007830619812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9180335402488708, "epoch": 8.83, "learning_rate": 2.2836568566061368e-05, "loss": 0.7761, "step": 10451, "task_loss": 1.2522287368774414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7286669611930847, "epoch": 8.83, "learning_rate": 2.28334376956794e-05, "loss": 0.6923, "step": 10452, "task_loss": 0.3248380124568939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41725075244903564, "epoch": 8.84, "learning_rate": 2.2830306825297435e-05, "loss": 0.5921, "step": 10453, "task_loss": 0.3158210813999176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8432713747024536, "epoch": 8.84, "learning_rate": 2.2827175954915467e-05, "loss": 0.6266, "step": 10454, "task_loss": 0.9793534278869629 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8957875370979309, "epoch": 8.84, "learning_rate": 2.2824045084533502e-05, "loss": 0.6226, "step": 10455, "task_loss": 0.8345890641212463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4196119010448456, "epoch": 8.84, "learning_rate": 2.2820914214151534e-05, "loss": 0.5604, "step": 10456, "task_loss": 1.2297245264053345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7853177785873413, "epoch": 8.84, "learning_rate": 2.281778334376957e-05, "loss": 0.591, "step": 10457, "task_loss": 1.0229095220565796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5569533705711365, "epoch": 8.84, "learning_rate": 2.28146524733876e-05, "loss": 0.4965, "step": 10458, "task_loss": 1.7872692346572876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5424090027809143, "epoch": 8.84, "learning_rate": 2.2811521603005636e-05, "loss": 0.4701, "step": 10459, "task_loss": 0.7421183586120605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.652808427810669, "epoch": 8.84, "learning_rate": 2.280839073262367e-05, "loss": 0.5357, "step": 10460, "task_loss": 0.9478225111961365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7929708957672119, "epoch": 8.84, "learning_rate": 2.2805259862241703e-05, "loss": 0.6332, "step": 10461, "task_loss": 1.2375218868255615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36287209391593933, "epoch": 8.84, "learning_rate": 2.2802128991859738e-05, "loss": 0.7238, "step": 10462, "task_loss": 0.8427521586418152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3722900152206421, "epoch": 8.84, "learning_rate": 2.279899812147777e-05, "loss": 0.5543, "step": 10463, "task_loss": 0.3778744041919708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6857789158821106, "epoch": 8.84, "learning_rate": 2.2795867251095805e-05, "loss": 0.5769, "step": 10464, "task_loss": 0.5193248987197876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5159193277359009, "epoch": 8.85, "learning_rate": 2.279273638071384e-05, "loss": 0.663, "step": 10465, "task_loss": 0.8453859090805054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47878706455230713, "epoch": 8.85, "learning_rate": 2.2789605510331875e-05, "loss": 0.504, "step": 10466, "task_loss": 0.7330911159515381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2631017565727234, "epoch": 8.85, "learning_rate": 2.2786474639949907e-05, "loss": 0.4961, "step": 10467, "task_loss": 1.042837142944336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5415751934051514, "epoch": 8.85, "learning_rate": 2.2783343769567942e-05, "loss": 0.6242, "step": 10468, "task_loss": 0.6013956665992737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3005352318286896, "epoch": 8.85, "learning_rate": 2.2780212899185974e-05, "loss": 0.5245, "step": 10469, "task_loss": 1.120347499847412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6120573282241821, "epoch": 8.85, "learning_rate": 2.277708202880401e-05, "loss": 0.6977, "step": 10470, "task_loss": 1.1221232414245605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3483387231826782, "epoch": 8.85, "learning_rate": 2.277395115842204e-05, "loss": 0.5292, "step": 10471, "task_loss": 0.6338210701942444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5100565552711487, "epoch": 8.85, "learning_rate": 2.2770820288040076e-05, "loss": 0.6053, "step": 10472, "task_loss": 0.9237958192825317 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48464730381965637, "epoch": 8.85, "learning_rate": 2.276768941765811e-05, "loss": 0.7357, "step": 10473, "task_loss": 0.6939235329627991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.648261547088623, "epoch": 8.85, "learning_rate": 2.2764558547276143e-05, "loss": 0.5596, "step": 10474, "task_loss": 0.41609832644462585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4448244571685791, "epoch": 8.85, "learning_rate": 2.2761427676894178e-05, "loss": 0.5142, "step": 10475, "task_loss": 0.37247776985168457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36651432514190674, "epoch": 8.85, "learning_rate": 2.275829680651221e-05, "loss": 0.4419, "step": 10476, "task_loss": 0.5051535964012146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5358697175979614, "epoch": 8.86, "learning_rate": 2.2755165936130245e-05, "loss": 0.503, "step": 10477, "task_loss": 0.4804258346557617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3328081965446472, "epoch": 8.86, "learning_rate": 2.2752035065748277e-05, "loss": 0.5096, "step": 10478, "task_loss": 0.16120098531246185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8418604731559753, "epoch": 8.86, "learning_rate": 2.2748904195366312e-05, "loss": 0.6516, "step": 10479, "task_loss": 0.3288886845111847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30376172065734863, "epoch": 8.86, "learning_rate": 2.2745773324984347e-05, "loss": 0.5074, "step": 10480, "task_loss": 0.5157417058944702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6025195717811584, "epoch": 8.86, "learning_rate": 2.2742642454602383e-05, "loss": 0.5719, "step": 10481, "task_loss": 1.1164966821670532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4370139539241791, "epoch": 8.86, "learning_rate": 2.2739511584220414e-05, "loss": 0.6286, "step": 10482, "task_loss": 1.3580635786056519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.088862657546997, "epoch": 8.86, "learning_rate": 2.273638071383845e-05, "loss": 0.6754, "step": 10483, "task_loss": 0.3700340688228607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5300573110580444, "epoch": 8.86, "learning_rate": 2.2733249843456485e-05, "loss": 0.5706, "step": 10484, "task_loss": 0.335488885641098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5880345106124878, "epoch": 8.86, "learning_rate": 2.2730118973074516e-05, "loss": 0.6712, "step": 10485, "task_loss": 1.2464828491210938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.529059886932373, "epoch": 8.86, "learning_rate": 2.272698810269255e-05, "loss": 0.521, "step": 10486, "task_loss": 0.683984100818634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7269982099533081, "epoch": 8.86, "learning_rate": 2.2723857232310583e-05, "loss": 0.5556, "step": 10487, "task_loss": 0.8668089509010315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4427383244037628, "epoch": 8.87, "learning_rate": 2.272072636192862e-05, "loss": 0.6575, "step": 10488, "task_loss": 1.0814142227172852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6097993850708008, "epoch": 8.87, "learning_rate": 2.271759549154665e-05, "loss": 0.4867, "step": 10489, "task_loss": 0.8811888098716736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5433340668678284, "epoch": 8.87, "learning_rate": 2.2714464621164685e-05, "loss": 0.597, "step": 10490, "task_loss": 0.35240957140922546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6685360670089722, "epoch": 8.87, "learning_rate": 2.2711333750782717e-05, "loss": 0.6268, "step": 10491, "task_loss": 0.5802634358406067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2729682922363281, "epoch": 8.87, "learning_rate": 2.2708202880400752e-05, "loss": 0.6238, "step": 10492, "task_loss": 0.5911088585853577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.815201997756958, "epoch": 8.87, "learning_rate": 2.2705072010018784e-05, "loss": 0.5418, "step": 10493, "task_loss": 1.1197336912155151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4049554467201233, "epoch": 8.87, "learning_rate": 2.270194113963682e-05, "loss": 0.4615, "step": 10494, "task_loss": 0.8980569839477539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7749455571174622, "epoch": 8.87, "learning_rate": 2.269881026925485e-05, "loss": 0.5951, "step": 10495, "task_loss": 1.2134231328964233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7167612314224243, "epoch": 8.87, "learning_rate": 2.2695679398872886e-05, "loss": 0.6021, "step": 10496, "task_loss": 0.5986979007720947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5996874570846558, "epoch": 8.87, "learning_rate": 2.269254852849092e-05, "loss": 0.5497, "step": 10497, "task_loss": 0.5379956960678101 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3817159831523895, "epoch": 8.87, "learning_rate": 2.2689417658108957e-05, "loss": 0.5444, "step": 10498, "task_loss": 0.3845944106578827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5230769515037537, "epoch": 8.87, "learning_rate": 2.2686286787726992e-05, "loss": 0.6068, "step": 10499, "task_loss": 0.24306488037109375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0672565698623657, "epoch": 8.88, "learning_rate": 2.2683155917345024e-05, "loss": 0.7235, "step": 10500, "task_loss": 1.7998485565185547 }, { "epoch": 8.88, "eval_accuracy": 0.9000396039603961, "eval_loss": 0.39152297377586365, "eval_runtime": 207.1761, "eval_samples_per_second": 121.877, "eval_steps_per_second": 0.956, "step": 10500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6651073694229126, "epoch": 8.88, "learning_rate": 2.268002504696306e-05, "loss": 0.722, "step": 10501, "task_loss": 0.5393909811973572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.497221976518631, "epoch": 8.88, "learning_rate": 2.267689417658109e-05, "loss": 0.6425, "step": 10502, "task_loss": 0.8226983547210693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6366095542907715, "epoch": 8.88, "learning_rate": 2.2673763306199126e-05, "loss": 0.5477, "step": 10503, "task_loss": 0.718218982219696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6796269416809082, "epoch": 8.88, "learning_rate": 2.2670632435817158e-05, "loss": 0.6823, "step": 10504, "task_loss": 1.184224247932434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5871130228042603, "epoch": 8.88, "learning_rate": 2.2667501565435193e-05, "loss": 0.5519, "step": 10505, "task_loss": 0.7667039632797241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4442088007926941, "epoch": 8.88, "learning_rate": 2.2664370695053224e-05, "loss": 0.4419, "step": 10506, "task_loss": 0.2888334393501282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6640299558639526, "epoch": 8.88, "learning_rate": 2.266123982467126e-05, "loss": 0.7325, "step": 10507, "task_loss": 0.8966391682624817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6949887275695801, "epoch": 8.88, "learning_rate": 2.265810895428929e-05, "loss": 0.5114, "step": 10508, "task_loss": 0.7734278440475464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5261613726615906, "epoch": 8.88, "learning_rate": 2.2654978083907327e-05, "loss": 0.5484, "step": 10509, "task_loss": 0.5083299279212952 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5977286100387573, "epoch": 8.88, "learning_rate": 2.2651847213525362e-05, "loss": 0.5967, "step": 10510, "task_loss": 0.2785641849040985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5019561052322388, "epoch": 8.88, "learning_rate": 2.2648716343143394e-05, "loss": 0.5299, "step": 10511, "task_loss": 0.7786123156547546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4069534242153168, "epoch": 8.89, "learning_rate": 2.264558547276143e-05, "loss": 0.5679, "step": 10512, "task_loss": 0.40460067987442017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45377668738365173, "epoch": 8.89, "learning_rate": 2.264245460237946e-05, "loss": 0.5356, "step": 10513, "task_loss": 1.1058604717254639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4052448570728302, "epoch": 8.89, "learning_rate": 2.2639323731997496e-05, "loss": 0.3612, "step": 10514, "task_loss": 0.48617422580718994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7197591662406921, "epoch": 8.89, "learning_rate": 2.263619286161553e-05, "loss": 0.6018, "step": 10515, "task_loss": 1.6527934074401855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1880302429199219, "epoch": 8.89, "learning_rate": 2.2633061991233566e-05, "loss": 0.7131, "step": 10516, "task_loss": 0.5114843845367432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.459593802690506, "epoch": 8.89, "learning_rate": 2.2629931120851598e-05, "loss": 0.5633, "step": 10517, "task_loss": 0.7250383496284485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44248291850090027, "epoch": 8.89, "learning_rate": 2.2626800250469633e-05, "loss": 0.5375, "step": 10518, "task_loss": 0.696547269821167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3824719488620758, "epoch": 8.89, "learning_rate": 2.2623669380087665e-05, "loss": 0.5232, "step": 10519, "task_loss": 0.22857072949409485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6758444905281067, "epoch": 8.89, "learning_rate": 2.26205385097057e-05, "loss": 0.5736, "step": 10520, "task_loss": 0.7699200510978699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5679305791854858, "epoch": 8.89, "learning_rate": 2.2617407639323735e-05, "loss": 0.5964, "step": 10521, "task_loss": 0.6197380423545837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0451138019561768, "epoch": 8.89, "learning_rate": 2.2614276768941767e-05, "loss": 0.6415, "step": 10522, "task_loss": 1.293403148651123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2590227723121643, "epoch": 8.89, "learning_rate": 2.2611145898559802e-05, "loss": 0.5023, "step": 10523, "task_loss": 0.04939115792512894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5097392797470093, "epoch": 8.9, "learning_rate": 2.2608015028177834e-05, "loss": 0.5813, "step": 10524, "task_loss": 0.40550073981285095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5395656824111938, "epoch": 8.9, "learning_rate": 2.260488415779587e-05, "loss": 0.6219, "step": 10525, "task_loss": 0.39081984758377075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.286435067653656, "epoch": 8.9, "learning_rate": 2.26017532874139e-05, "loss": 0.5617, "step": 10526, "task_loss": 0.8928550481796265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38232189416885376, "epoch": 8.9, "learning_rate": 2.2598622417031936e-05, "loss": 0.4948, "step": 10527, "task_loss": 0.459408700466156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35106468200683594, "epoch": 8.9, "learning_rate": 2.2595491546649968e-05, "loss": 0.4116, "step": 10528, "task_loss": 0.5715097784996033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7316780090332031, "epoch": 8.9, "learning_rate": 2.2592360676268003e-05, "loss": 0.5853, "step": 10529, "task_loss": 0.8471717238426208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0662968158721924, "epoch": 8.9, "learning_rate": 2.2589229805886035e-05, "loss": 0.9217, "step": 10530, "task_loss": 1.8822026252746582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7219607830047607, "epoch": 8.9, "learning_rate": 2.258609893550407e-05, "loss": 0.5858, "step": 10531, "task_loss": 1.276758074760437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5173249244689941, "epoch": 8.9, "learning_rate": 2.2582968065122105e-05, "loss": 0.5173, "step": 10532, "task_loss": 1.4584888219833374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0360435247421265, "epoch": 8.9, "learning_rate": 2.257983719474014e-05, "loss": 0.7347, "step": 10533, "task_loss": 0.9073396325111389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42327630519866943, "epoch": 8.9, "learning_rate": 2.2576706324358175e-05, "loss": 0.5072, "step": 10534, "task_loss": 1.5252729654312134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5239620208740234, "epoch": 8.9, "learning_rate": 2.2573575453976207e-05, "loss": 0.5989, "step": 10535, "task_loss": 1.301074743270874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1205112934112549, "epoch": 8.91, "learning_rate": 2.2570444583594242e-05, "loss": 0.7261, "step": 10536, "task_loss": 0.9053485989570618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39859697222709656, "epoch": 8.91, "learning_rate": 2.2567313713212274e-05, "loss": 0.448, "step": 10537, "task_loss": 0.5359454154968262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5049400329589844, "epoch": 8.91, "learning_rate": 2.256418284283031e-05, "loss": 0.6402, "step": 10538, "task_loss": 1.273592233657837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4639715850353241, "epoch": 8.91, "learning_rate": 2.256105197244834e-05, "loss": 0.6661, "step": 10539, "task_loss": 0.879682183265686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.68784099817276, "epoch": 8.91, "learning_rate": 2.2557921102066376e-05, "loss": 0.5217, "step": 10540, "task_loss": 1.7078843116760254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6524635553359985, "epoch": 8.91, "learning_rate": 2.2554790231684408e-05, "loss": 0.7359, "step": 10541, "task_loss": 0.6464616656303406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4324854910373688, "epoch": 8.91, "learning_rate": 2.2551659361302443e-05, "loss": 0.5077, "step": 10542, "task_loss": 0.48503589630126953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.767743706703186, "epoch": 8.91, "learning_rate": 2.2548528490920475e-05, "loss": 0.5525, "step": 10543, "task_loss": 0.6313313841819763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4183396100997925, "epoch": 8.91, "learning_rate": 2.254539762053851e-05, "loss": 0.6781, "step": 10544, "task_loss": 0.7631233930587769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7072739005088806, "epoch": 8.91, "learning_rate": 2.2542266750156542e-05, "loss": 0.4884, "step": 10545, "task_loss": 1.483012080192566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5243905782699585, "epoch": 8.91, "learning_rate": 2.2539135879774577e-05, "loss": 0.5393, "step": 10546, "task_loss": 0.2655244767665863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2809096574783325, "epoch": 8.91, "learning_rate": 2.2536005009392612e-05, "loss": 0.4618, "step": 10547, "task_loss": 0.7709941267967224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5723223090171814, "epoch": 8.92, "learning_rate": 2.2532874139010647e-05, "loss": 0.5094, "step": 10548, "task_loss": 1.709193468093872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0904715061187744, "epoch": 8.92, "learning_rate": 2.252974326862868e-05, "loss": 0.6276, "step": 10549, "task_loss": 0.4644371271133423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7216352820396423, "epoch": 8.92, "learning_rate": 2.2526612398246714e-05, "loss": 0.7253, "step": 10550, "task_loss": 1.0955015420913696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7654957175254822, "epoch": 8.92, "learning_rate": 2.252348152786475e-05, "loss": 0.617, "step": 10551, "task_loss": 0.9037821888923645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32695215940475464, "epoch": 8.92, "learning_rate": 2.252035065748278e-05, "loss": 0.3656, "step": 10552, "task_loss": 0.33592721819877625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6607288122177124, "epoch": 8.92, "learning_rate": 2.2517219787100816e-05, "loss": 0.7694, "step": 10553, "task_loss": 0.45788100361824036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5939805507659912, "epoch": 8.92, "learning_rate": 2.2514088916718848e-05, "loss": 0.5452, "step": 10554, "task_loss": 0.7119861841201782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42948997020721436, "epoch": 8.92, "learning_rate": 2.2510958046336883e-05, "loss": 0.5261, "step": 10555, "task_loss": 0.35702744126319885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.615172803401947, "epoch": 8.92, "learning_rate": 2.2507827175954915e-05, "loss": 0.5194, "step": 10556, "task_loss": 1.215795636177063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4837541878223419, "epoch": 8.92, "learning_rate": 2.250469630557295e-05, "loss": 0.6583, "step": 10557, "task_loss": 1.2889662981033325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5671366453170776, "epoch": 8.92, "learning_rate": 2.2501565435190986e-05, "loss": 0.5183, "step": 10558, "task_loss": 0.9319977760314941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6642827987670898, "epoch": 8.93, "learning_rate": 2.2498434564809017e-05, "loss": 0.7544, "step": 10559, "task_loss": 0.7280251979827881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5016883611679077, "epoch": 8.93, "learning_rate": 2.2495303694427052e-05, "loss": 0.5136, "step": 10560, "task_loss": 0.9828711748123169 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.039698839187622, "epoch": 8.93, "learning_rate": 2.2492172824045084e-05, "loss": 0.6538, "step": 10561, "task_loss": 0.9945589303970337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9994767904281616, "epoch": 8.93, "learning_rate": 2.248904195366312e-05, "loss": 0.85, "step": 10562, "task_loss": 0.9413736462593079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49296313524246216, "epoch": 8.93, "learning_rate": 2.248591108328115e-05, "loss": 0.4795, "step": 10563, "task_loss": 0.5282530784606934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7023472785949707, "epoch": 8.93, "learning_rate": 2.2482780212899186e-05, "loss": 0.4683, "step": 10564, "task_loss": 1.2123581171035767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6587040424346924, "epoch": 8.93, "learning_rate": 2.247964934251722e-05, "loss": 0.5152, "step": 10565, "task_loss": 0.655671238899231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49091798067092896, "epoch": 8.93, "learning_rate": 2.2476518472135257e-05, "loss": 0.6792, "step": 10566, "task_loss": 0.8364027738571167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5026556253433228, "epoch": 8.93, "learning_rate": 2.247338760175329e-05, "loss": 0.6429, "step": 10567, "task_loss": 0.3729636073112488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7575958967208862, "epoch": 8.93, "learning_rate": 2.2470256731371324e-05, "loss": 0.4896, "step": 10568, "task_loss": 0.538507878780365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7647144198417664, "epoch": 8.93, "learning_rate": 2.2467125860989355e-05, "loss": 0.5866, "step": 10569, "task_loss": 0.7485898733139038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.662279486656189, "epoch": 8.93, "learning_rate": 2.246399499060739e-05, "loss": 0.6599, "step": 10570, "task_loss": 0.3507001996040344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5499578714370728, "epoch": 8.94, "learning_rate": 2.2460864120225426e-05, "loss": 0.5126, "step": 10571, "task_loss": 1.3203836679458618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5625450611114502, "epoch": 8.94, "learning_rate": 2.2457733249843458e-05, "loss": 0.5124, "step": 10572, "task_loss": 0.41948699951171875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5856466889381409, "epoch": 8.94, "learning_rate": 2.2454602379461493e-05, "loss": 0.6168, "step": 10573, "task_loss": 0.9739737510681152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3379833698272705, "epoch": 8.94, "learning_rate": 2.2451471509079525e-05, "loss": 0.5611, "step": 10574, "task_loss": 0.21121174097061157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7921472787857056, "epoch": 8.94, "learning_rate": 2.244834063869756e-05, "loss": 0.466, "step": 10575, "task_loss": 0.1358671337366104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5874049067497253, "epoch": 8.94, "learning_rate": 2.244520976831559e-05, "loss": 0.6433, "step": 10576, "task_loss": 0.3686770796775818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3398624062538147, "epoch": 8.94, "learning_rate": 2.2442078897933627e-05, "loss": 0.5573, "step": 10577, "task_loss": 0.2666524052619934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6123543381690979, "epoch": 8.94, "learning_rate": 2.243894802755166e-05, "loss": 0.7215, "step": 10578, "task_loss": 0.7572358250617981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41648241877555847, "epoch": 8.94, "learning_rate": 2.2435817157169694e-05, "loss": 0.5967, "step": 10579, "task_loss": 0.8961945176124573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43926846981048584, "epoch": 8.94, "learning_rate": 2.2432686286787725e-05, "loss": 0.5374, "step": 10580, "task_loss": 0.5121837258338928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5413174033164978, "epoch": 8.94, "learning_rate": 2.242955541640576e-05, "loss": 0.6501, "step": 10581, "task_loss": 0.5615526437759399 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41683468222618103, "epoch": 8.94, "learning_rate": 2.2426424546023796e-05, "loss": 0.5186, "step": 10582, "task_loss": 0.2611750364303589 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43540048599243164, "epoch": 8.95, "learning_rate": 2.242329367564183e-05, "loss": 0.5529, "step": 10583, "task_loss": 0.07233627885580063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5447527766227722, "epoch": 8.95, "learning_rate": 2.2420162805259866e-05, "loss": 0.599, "step": 10584, "task_loss": 0.2399687021970749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47865283489227295, "epoch": 8.95, "learning_rate": 2.2417031934877898e-05, "loss": 0.5795, "step": 10585, "task_loss": 0.20815511047840118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8532282114028931, "epoch": 8.95, "learning_rate": 2.2413901064495933e-05, "loss": 0.6902, "step": 10586, "task_loss": 0.9820976257324219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6111355423927307, "epoch": 8.95, "learning_rate": 2.2410770194113965e-05, "loss": 0.5499, "step": 10587, "task_loss": 1.1582748889923096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.727075457572937, "epoch": 8.95, "learning_rate": 2.2407639323732e-05, "loss": 0.4651, "step": 10588, "task_loss": 0.8232205510139465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4844414293766022, "epoch": 8.95, "learning_rate": 2.2404508453350032e-05, "loss": 0.5418, "step": 10589, "task_loss": 0.725516140460968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5620662569999695, "epoch": 8.95, "learning_rate": 2.2401377582968067e-05, "loss": 0.6722, "step": 10590, "task_loss": 0.3217407166957855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31349092721939087, "epoch": 8.95, "learning_rate": 2.23982467125861e-05, "loss": 0.4698, "step": 10591, "task_loss": 0.36254453659057617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6293317079544067, "epoch": 8.95, "learning_rate": 2.2395115842204134e-05, "loss": 0.6023, "step": 10592, "task_loss": 1.2626954317092896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5917924046516418, "epoch": 8.95, "learning_rate": 2.2391984971822166e-05, "loss": 0.4709, "step": 10593, "task_loss": 0.42773035168647766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.3988544940948486, "epoch": 8.95, "learning_rate": 2.23888541014402e-05, "loss": 0.669, "step": 10594, "task_loss": 1.6082258224487305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8362625241279602, "epoch": 8.96, "learning_rate": 2.2385723231058236e-05, "loss": 0.6273, "step": 10595, "task_loss": 1.260994553565979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30089500546455383, "epoch": 8.96, "learning_rate": 2.2382592360676268e-05, "loss": 0.4206, "step": 10596, "task_loss": 0.41844215989112854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5123028755187988, "epoch": 8.96, "learning_rate": 2.2379461490294303e-05, "loss": 0.5223, "step": 10597, "task_loss": 0.6112334132194519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4751848578453064, "epoch": 8.96, "learning_rate": 2.2376330619912335e-05, "loss": 0.4827, "step": 10598, "task_loss": 0.5803654193878174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4192669093608856, "epoch": 8.96, "learning_rate": 2.237319974953037e-05, "loss": 0.5472, "step": 10599, "task_loss": 0.3756052851676941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5249388813972473, "epoch": 8.96, "learning_rate": 2.2370068879148405e-05, "loss": 0.6098, "step": 10600, "task_loss": 0.7778791189193726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0921502113342285, "epoch": 8.96, "learning_rate": 2.236693800876644e-05, "loss": 0.7025, "step": 10601, "task_loss": 0.9072845578193665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6501781344413757, "epoch": 8.96, "learning_rate": 2.2363807138384472e-05, "loss": 0.5238, "step": 10602, "task_loss": 1.216488003730774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9506369829177856, "epoch": 8.96, "learning_rate": 2.2360676268002507e-05, "loss": 0.6572, "step": 10603, "task_loss": 1.0474634170532227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5189471244812012, "epoch": 8.96, "learning_rate": 2.235754539762054e-05, "loss": 0.6974, "step": 10604, "task_loss": 0.661942720413208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6137993335723877, "epoch": 8.96, "learning_rate": 2.2354414527238574e-05, "loss": 0.5374, "step": 10605, "task_loss": 1.2805614471435547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7540587186813354, "epoch": 8.96, "learning_rate": 2.2351283656856606e-05, "loss": 0.6692, "step": 10606, "task_loss": 0.9661505222320557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4268311560153961, "epoch": 8.97, "learning_rate": 2.234815278647464e-05, "loss": 0.5875, "step": 10607, "task_loss": 0.4295666217803955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5763248205184937, "epoch": 8.97, "learning_rate": 2.2345021916092676e-05, "loss": 0.5764, "step": 10608, "task_loss": 0.7235970497131348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21764075756072998, "epoch": 8.97, "learning_rate": 2.2341891045710708e-05, "loss": 0.4345, "step": 10609, "task_loss": 0.5883426666259766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6024489402770996, "epoch": 8.97, "learning_rate": 2.2338760175328743e-05, "loss": 0.6474, "step": 10610, "task_loss": 0.14490240812301636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4623776078224182, "epoch": 8.97, "learning_rate": 2.2335629304946775e-05, "loss": 0.5211, "step": 10611, "task_loss": 1.1356221437454224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7902412414550781, "epoch": 8.97, "learning_rate": 2.233249843456481e-05, "loss": 0.65, "step": 10612, "task_loss": 0.866962730884552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5067769289016724, "epoch": 8.97, "learning_rate": 2.2329367564182842e-05, "loss": 0.6816, "step": 10613, "task_loss": 0.9006754755973816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39656907320022583, "epoch": 8.97, "learning_rate": 2.2326236693800877e-05, "loss": 0.5064, "step": 10614, "task_loss": 0.3920118510723114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7842472791671753, "epoch": 8.97, "learning_rate": 2.232310582341891e-05, "loss": 0.5775, "step": 10615, "task_loss": 0.8675864934921265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8135489821434021, "epoch": 8.97, "learning_rate": 2.2319974953036944e-05, "loss": 0.662, "step": 10616, "task_loss": 0.7087157964706421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7846254110336304, "epoch": 8.97, "learning_rate": 2.231684408265498e-05, "loss": 0.6719, "step": 10617, "task_loss": 0.5041727423667908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6300644874572754, "epoch": 8.97, "learning_rate": 2.2313713212273014e-05, "loss": 0.4949, "step": 10618, "task_loss": 1.063664436340332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9425626397132874, "epoch": 8.98, "learning_rate": 2.231058234189105e-05, "loss": 0.7076, "step": 10619, "task_loss": 0.4290371835231781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.569407045841217, "epoch": 8.98, "learning_rate": 2.230745147150908e-05, "loss": 0.5536, "step": 10620, "task_loss": 0.32197675108909607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5068245530128479, "epoch": 8.98, "learning_rate": 2.2304320601127116e-05, "loss": 0.6984, "step": 10621, "task_loss": 0.2650056481361389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.816597580909729, "epoch": 8.98, "learning_rate": 2.2301189730745148e-05, "loss": 0.6008, "step": 10622, "task_loss": 1.042345643043518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5248167514801025, "epoch": 8.98, "learning_rate": 2.2298058860363183e-05, "loss": 0.5377, "step": 10623, "task_loss": 0.19034342467784882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7127043604850769, "epoch": 8.98, "learning_rate": 2.2294927989981215e-05, "loss": 0.6927, "step": 10624, "task_loss": 1.2164205312728882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4579421579837799, "epoch": 8.98, "learning_rate": 2.229179711959925e-05, "loss": 0.5201, "step": 10625, "task_loss": 0.5209630727767944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5605078339576721, "epoch": 8.98, "learning_rate": 2.2288666249217282e-05, "loss": 0.4611, "step": 10626, "task_loss": 0.2377406656742096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3966991901397705, "epoch": 8.98, "learning_rate": 2.2285535378835317e-05, "loss": 0.4749, "step": 10627, "task_loss": 0.17085415124893188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4410800337791443, "epoch": 8.98, "learning_rate": 2.228240450845335e-05, "loss": 0.5071, "step": 10628, "task_loss": 0.4847782850265503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4172781705856323, "epoch": 8.98, "learning_rate": 2.2279273638071384e-05, "loss": 0.5878, "step": 10629, "task_loss": 0.4181741178035736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5516926050186157, "epoch": 8.99, "learning_rate": 2.2276142767689416e-05, "loss": 0.6053, "step": 10630, "task_loss": 0.8574026226997375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5714877843856812, "epoch": 8.99, "learning_rate": 2.227301189730745e-05, "loss": 0.5585, "step": 10631, "task_loss": 0.3420681953430176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6054515838623047, "epoch": 8.99, "learning_rate": 2.2269881026925486e-05, "loss": 0.7037, "step": 10632, "task_loss": 0.5843099355697632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4143931269645691, "epoch": 8.99, "learning_rate": 2.226675015654352e-05, "loss": 0.4066, "step": 10633, "task_loss": 0.6431547403335571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3455732762813568, "epoch": 8.99, "learning_rate": 2.2263619286161553e-05, "loss": 0.4412, "step": 10634, "task_loss": 0.4587627351284027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30555930733680725, "epoch": 8.99, "learning_rate": 2.226048841577959e-05, "loss": 0.5902, "step": 10635, "task_loss": 0.21912236511707306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7080652713775635, "epoch": 8.99, "learning_rate": 2.2257357545397624e-05, "loss": 0.5724, "step": 10636, "task_loss": 0.2603958547115326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6323369741439819, "epoch": 8.99, "learning_rate": 2.2254226675015655e-05, "loss": 0.5017, "step": 10637, "task_loss": 0.44254323840141296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4881642758846283, "epoch": 8.99, "learning_rate": 2.225109580463369e-05, "loss": 0.5253, "step": 10638, "task_loss": 1.0547528266906738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.823168158531189, "epoch": 8.99, "learning_rate": 2.2247964934251722e-05, "loss": 0.5557, "step": 10639, "task_loss": 1.7136815786361694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6956143975257874, "epoch": 8.99, "learning_rate": 2.2244834063869758e-05, "loss": 0.6086, "step": 10640, "task_loss": 0.7801706194877625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5329354405403137, "epoch": 8.99, "learning_rate": 2.224170319348779e-05, "loss": 0.49, "step": 10641, "task_loss": 1.409690260887146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5294376611709595, "epoch": 9.0, "learning_rate": 2.2238572323105825e-05, "loss": 0.6152, "step": 10642, "task_loss": 0.17128396034240723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5437154173851013, "epoch": 9.0, "learning_rate": 2.2235441452723856e-05, "loss": 0.6025, "step": 10643, "task_loss": 0.4318736791610718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3560106158256531, "epoch": 9.0, "learning_rate": 2.223231058234189e-05, "loss": 0.5232, "step": 10644, "task_loss": 0.829944908618927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8583235740661621, "epoch": 9.0, "learning_rate": 2.2229179711959927e-05, "loss": 0.5733, "step": 10645, "task_loss": 0.8797690868377686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7663442492485046, "epoch": 9.0, "learning_rate": 2.222604884157796e-05, "loss": 0.6528, "step": 10646, "task_loss": 0.5032749772071838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6132590174674988, "epoch": 9.0, "learning_rate": 2.2222917971195994e-05, "loss": 0.512, "step": 10647, "task_loss": 0.9045599699020386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3559381365776062, "epoch": 9.0, "learning_rate": 2.2219787100814025e-05, "loss": 0.8749, "step": 10648, "task_loss": 0.3542212247848511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42127424478530884, "epoch": 9.0, "learning_rate": 2.221665623043206e-05, "loss": 0.5065, "step": 10649, "task_loss": 0.38799262046813965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9754449129104614, "epoch": 9.0, "learning_rate": 2.2213525360050096e-05, "loss": 0.6681, "step": 10650, "task_loss": 0.7821944952011108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.345196008682251, "epoch": 9.0, "learning_rate": 2.221039448966813e-05, "loss": 0.5193, "step": 10651, "task_loss": 0.8068575859069824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5343503355979919, "epoch": 9.0, "learning_rate": 2.2207263619286163e-05, "loss": 0.5142, "step": 10652, "task_loss": 0.31566309928894043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5753676891326904, "epoch": 9.01, "learning_rate": 2.2204132748904198e-05, "loss": 0.5053, "step": 10653, "task_loss": 0.2249642312526703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4517761468887329, "epoch": 9.01, "learning_rate": 2.220100187852223e-05, "loss": 0.6173, "step": 10654, "task_loss": 0.5889878869056702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4658886790275574, "epoch": 9.01, "learning_rate": 2.2197871008140265e-05, "loss": 0.4458, "step": 10655, "task_loss": 0.08600388467311859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3079991936683655, "epoch": 9.01, "learning_rate": 2.21947401377583e-05, "loss": 0.609, "step": 10656, "task_loss": 0.25085118412971497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7462258338928223, "epoch": 9.01, "learning_rate": 2.2191609267376332e-05, "loss": 0.7535, "step": 10657, "task_loss": 0.7678066492080688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4747820198535919, "epoch": 9.01, "learning_rate": 2.2188478396994367e-05, "loss": 0.4164, "step": 10658, "task_loss": 0.6217344999313354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7664492130279541, "epoch": 9.01, "learning_rate": 2.21853475266124e-05, "loss": 0.7188, "step": 10659, "task_loss": 1.0797388553619385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5773204565048218, "epoch": 9.01, "learning_rate": 2.2182216656230434e-05, "loss": 0.5254, "step": 10660, "task_loss": 1.1168423891067505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.769020676612854, "epoch": 9.01, "learning_rate": 2.2179085785848466e-05, "loss": 0.5083, "step": 10661, "task_loss": 1.6923174858093262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3498965799808502, "epoch": 9.01, "learning_rate": 2.21759549154665e-05, "loss": 0.5245, "step": 10662, "task_loss": 0.5892130136489868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5830617547035217, "epoch": 9.01, "learning_rate": 2.2172824045084533e-05, "loss": 0.5005, "step": 10663, "task_loss": 1.0869790315628052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3650296926498413, "epoch": 9.01, "learning_rate": 2.2169693174702568e-05, "loss": 0.6964, "step": 10664, "task_loss": 0.12098046392202377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48606032133102417, "epoch": 9.02, "learning_rate": 2.21665623043206e-05, "loss": 0.5104, "step": 10665, "task_loss": 1.2455891370773315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5193489193916321, "epoch": 9.02, "learning_rate": 2.2163431433938635e-05, "loss": 0.6493, "step": 10666, "task_loss": 0.5706619620323181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5766449570655823, "epoch": 9.02, "learning_rate": 2.216030056355667e-05, "loss": 0.5503, "step": 10667, "task_loss": 0.9806789755821228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.67018723487854, "epoch": 9.02, "learning_rate": 2.2157169693174705e-05, "loss": 0.6035, "step": 10668, "task_loss": 0.7231374382972717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5076508522033691, "epoch": 9.02, "learning_rate": 2.215403882279274e-05, "loss": 0.5673, "step": 10669, "task_loss": 0.6072719693183899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6134536266326904, "epoch": 9.02, "learning_rate": 2.2150907952410772e-05, "loss": 0.4716, "step": 10670, "task_loss": 0.7222689986228943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40938618779182434, "epoch": 9.02, "learning_rate": 2.2147777082028807e-05, "loss": 0.6085, "step": 10671, "task_loss": 0.40451759099960327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6116071343421936, "epoch": 9.02, "learning_rate": 2.214464621164684e-05, "loss": 0.4851, "step": 10672, "task_loss": 1.1161051988601685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5814194083213806, "epoch": 9.02, "learning_rate": 2.2141515341264874e-05, "loss": 0.6445, "step": 10673, "task_loss": 0.2313188910484314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8266257643699646, "epoch": 9.02, "learning_rate": 2.2138384470882906e-05, "loss": 0.6458, "step": 10674, "task_loss": 0.852421224117279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5609910488128662, "epoch": 9.02, "learning_rate": 2.213525360050094e-05, "loss": 0.6157, "step": 10675, "task_loss": 0.4849820137023926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39059340953826904, "epoch": 9.02, "learning_rate": 2.2132122730118973e-05, "loss": 0.3712, "step": 10676, "task_loss": 0.40398895740509033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46488335728645325, "epoch": 9.03, "learning_rate": 2.2128991859737008e-05, "loss": 0.4816, "step": 10677, "task_loss": 0.16757479310035706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.240903377532959, "epoch": 9.03, "learning_rate": 2.212586098935504e-05, "loss": 0.7353, "step": 10678, "task_loss": 0.9938888549804688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4988916516304016, "epoch": 9.03, "learning_rate": 2.2122730118973075e-05, "loss": 0.5881, "step": 10679, "task_loss": 0.5737087726593018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44180119037628174, "epoch": 9.03, "learning_rate": 2.211959924859111e-05, "loss": 0.5284, "step": 10680, "task_loss": 0.3785021901130676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3575299084186554, "epoch": 9.03, "learning_rate": 2.2116468378209142e-05, "loss": 0.5638, "step": 10681, "task_loss": 0.6537627577781677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.620439350605011, "epoch": 9.03, "learning_rate": 2.2113337507827177e-05, "loss": 0.6045, "step": 10682, "task_loss": 0.9465402960777283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44660520553588867, "epoch": 9.03, "learning_rate": 2.211020663744521e-05, "loss": 0.4567, "step": 10683, "task_loss": 0.2204759567975998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7208359241485596, "epoch": 9.03, "learning_rate": 2.2107075767063244e-05, "loss": 0.5759, "step": 10684, "task_loss": 0.6693023443222046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.373515248298645, "epoch": 9.03, "learning_rate": 2.210394489668128e-05, "loss": 0.4319, "step": 10685, "task_loss": 0.642306387424469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6277673244476318, "epoch": 9.03, "learning_rate": 2.2100814026299314e-05, "loss": 0.556, "step": 10686, "task_loss": 1.1119873523712158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.425438791513443, "epoch": 9.03, "learning_rate": 2.2097683155917346e-05, "loss": 0.5672, "step": 10687, "task_loss": 0.7544173002243042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7341930866241455, "epoch": 9.03, "learning_rate": 2.209455228553538e-05, "loss": 0.5881, "step": 10688, "task_loss": 0.9648576974868774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4986339211463928, "epoch": 9.04, "learning_rate": 2.2091421415153413e-05, "loss": 0.5202, "step": 10689, "task_loss": 0.5140330195426941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49087774753570557, "epoch": 9.04, "learning_rate": 2.208829054477145e-05, "loss": 0.6302, "step": 10690, "task_loss": 1.0153247117996216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9305101633071899, "epoch": 9.04, "learning_rate": 2.208515967438948e-05, "loss": 0.7117, "step": 10691, "task_loss": 1.706640362739563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7804336547851562, "epoch": 9.04, "learning_rate": 2.2082028804007515e-05, "loss": 0.5938, "step": 10692, "task_loss": 1.1137992143630981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5189664363861084, "epoch": 9.04, "learning_rate": 2.207889793362555e-05, "loss": 0.6043, "step": 10693, "task_loss": 0.9979012608528137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5609831809997559, "epoch": 9.04, "learning_rate": 2.2075767063243582e-05, "loss": 0.5434, "step": 10694, "task_loss": 0.6472615003585815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5517375469207764, "epoch": 9.04, "learning_rate": 2.2072636192861617e-05, "loss": 0.5725, "step": 10695, "task_loss": 1.0434991121292114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33626535534858704, "epoch": 9.04, "learning_rate": 2.206950532247965e-05, "loss": 0.3632, "step": 10696, "task_loss": 0.46170279383659363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2594614624977112, "epoch": 9.04, "learning_rate": 2.2066374452097684e-05, "loss": 0.4997, "step": 10697, "task_loss": 0.1571510136127472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5253106355667114, "epoch": 9.04, "learning_rate": 2.2063243581715716e-05, "loss": 0.6367, "step": 10698, "task_loss": 0.6667125225067139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42036378383636475, "epoch": 9.04, "learning_rate": 2.206011271133375e-05, "loss": 0.4143, "step": 10699, "task_loss": 0.4377796947956085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5862522125244141, "epoch": 9.04, "learning_rate": 2.2056981840951786e-05, "loss": 0.4674, "step": 10700, "task_loss": 0.39888280630111694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19524714350700378, "epoch": 9.05, "learning_rate": 2.2053850970569818e-05, "loss": 0.3846, "step": 10701, "task_loss": 0.013180790469050407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8074008822441101, "epoch": 9.05, "learning_rate": 2.2050720100187853e-05, "loss": 0.6737, "step": 10702, "task_loss": 0.5243997573852539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6841822862625122, "epoch": 9.05, "learning_rate": 2.204758922980589e-05, "loss": 0.5365, "step": 10703, "task_loss": 0.43585440516471863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7253317832946777, "epoch": 9.05, "learning_rate": 2.204445835942392e-05, "loss": 0.62, "step": 10704, "task_loss": 0.9110039472579956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4101744592189789, "epoch": 9.05, "learning_rate": 2.2041327489041956e-05, "loss": 0.4906, "step": 10705, "task_loss": 0.716835618019104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7951672077178955, "epoch": 9.05, "learning_rate": 2.203819661865999e-05, "loss": 0.7082, "step": 10706, "task_loss": 0.47645097970962524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5525001287460327, "epoch": 9.05, "learning_rate": 2.2035065748278022e-05, "loss": 0.6664, "step": 10707, "task_loss": 1.0840901136398315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37087222933769226, "epoch": 9.05, "learning_rate": 2.2031934877896058e-05, "loss": 0.5259, "step": 10708, "task_loss": 1.2809826135635376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4934075176715851, "epoch": 9.05, "learning_rate": 2.202880400751409e-05, "loss": 0.4474, "step": 10709, "task_loss": 0.1715051382780075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6977453231811523, "epoch": 9.05, "learning_rate": 2.2025673137132125e-05, "loss": 0.5588, "step": 10710, "task_loss": 0.33562302589416504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6342290043830872, "epoch": 9.05, "learning_rate": 2.2022542266750156e-05, "loss": 0.5673, "step": 10711, "task_loss": 0.8253740072250366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7208288311958313, "epoch": 9.05, "learning_rate": 2.201941139636819e-05, "loss": 0.6204, "step": 10712, "task_loss": 0.7400071024894714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43021851778030396, "epoch": 9.06, "learning_rate": 2.2016280525986223e-05, "loss": 0.4875, "step": 10713, "task_loss": 0.38756582140922546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5973491072654724, "epoch": 9.06, "learning_rate": 2.201314965560426e-05, "loss": 0.5818, "step": 10714, "task_loss": 1.0984511375427246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.843700110912323, "epoch": 9.06, "learning_rate": 2.201001878522229e-05, "loss": 0.5886, "step": 10715, "task_loss": 1.9912712574005127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4342159926891327, "epoch": 9.06, "learning_rate": 2.2006887914840325e-05, "loss": 0.57, "step": 10716, "task_loss": 0.37605413794517517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6684862375259399, "epoch": 9.06, "learning_rate": 2.200375704445836e-05, "loss": 0.7656, "step": 10717, "task_loss": 1.0897732973098755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3676321506500244, "epoch": 9.06, "learning_rate": 2.2000626174076396e-05, "loss": 0.5399, "step": 10718, "task_loss": 0.2690333425998688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8932292461395264, "epoch": 9.06, "learning_rate": 2.1997495303694428e-05, "loss": 0.8428, "step": 10719, "task_loss": 0.9153185486793518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5293970108032227, "epoch": 9.06, "learning_rate": 2.1994364433312463e-05, "loss": 0.4228, "step": 10720, "task_loss": 0.4464450180530548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41897884011268616, "epoch": 9.06, "learning_rate": 2.1991233562930498e-05, "loss": 0.5878, "step": 10721, "task_loss": 0.5197316408157349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3814024329185486, "epoch": 9.06, "learning_rate": 2.198810269254853e-05, "loss": 0.4517, "step": 10722, "task_loss": 0.12296129763126373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35739654302597046, "epoch": 9.06, "learning_rate": 2.1984971822166565e-05, "loss": 0.4619, "step": 10723, "task_loss": 0.599518358707428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21526549756526947, "epoch": 9.07, "learning_rate": 2.1981840951784597e-05, "loss": 0.5929, "step": 10724, "task_loss": 0.10506422072649002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6124686002731323, "epoch": 9.07, "learning_rate": 2.1978710081402632e-05, "loss": 0.5799, "step": 10725, "task_loss": 0.9221330285072327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4512573778629303, "epoch": 9.07, "learning_rate": 2.1975579211020664e-05, "loss": 0.573, "step": 10726, "task_loss": 0.6633158922195435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.662458598613739, "epoch": 9.07, "learning_rate": 2.19724483406387e-05, "loss": 0.8251, "step": 10727, "task_loss": 1.026606559753418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4396263360977173, "epoch": 9.07, "learning_rate": 2.196931747025673e-05, "loss": 0.5427, "step": 10728, "task_loss": 0.5733851194381714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8989805579185486, "epoch": 9.07, "learning_rate": 2.1966186599874766e-05, "loss": 0.7215, "step": 10729, "task_loss": 0.6479354500770569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7944360971450806, "epoch": 9.07, "learning_rate": 2.19630557294928e-05, "loss": 0.5838, "step": 10730, "task_loss": 1.4490487575531006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6620262861251831, "epoch": 9.07, "learning_rate": 2.1959924859110833e-05, "loss": 0.4789, "step": 10731, "task_loss": 0.5538578033447266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2748720049858093, "epoch": 9.07, "learning_rate": 2.1956793988728868e-05, "loss": 0.5074, "step": 10732, "task_loss": 0.22586020827293396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7101326584815979, "epoch": 9.07, "learning_rate": 2.19536631183469e-05, "loss": 0.5584, "step": 10733, "task_loss": 1.0976570844650269 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7758088111877441, "epoch": 9.07, "learning_rate": 2.1950532247964935e-05, "loss": 0.6309, "step": 10734, "task_loss": 0.7096821069717407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2668735980987549, "epoch": 9.07, "learning_rate": 2.194740137758297e-05, "loss": 0.4418, "step": 10735, "task_loss": 0.028588177636265755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6448503136634827, "epoch": 9.08, "learning_rate": 2.1944270507201005e-05, "loss": 0.5638, "step": 10736, "task_loss": 0.7795564532279968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.363076388835907, "epoch": 9.08, "learning_rate": 2.1941139636819037e-05, "loss": 0.5031, "step": 10737, "task_loss": 1.0536030530929565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6188504099845886, "epoch": 9.08, "learning_rate": 2.1938008766437072e-05, "loss": 0.5019, "step": 10738, "task_loss": 1.670802354812622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40334996581077576, "epoch": 9.08, "learning_rate": 2.1934877896055104e-05, "loss": 0.6482, "step": 10739, "task_loss": 0.7416701316833496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38880765438079834, "epoch": 9.08, "learning_rate": 2.193174702567314e-05, "loss": 0.4128, "step": 10740, "task_loss": 0.4301588237285614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48531949520111084, "epoch": 9.08, "learning_rate": 2.192861615529117e-05, "loss": 0.6416, "step": 10741, "task_loss": 0.9696143865585327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2410479485988617, "epoch": 9.08, "learning_rate": 2.1925485284909206e-05, "loss": 0.509, "step": 10742, "task_loss": 0.7337374091148376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38391023874282837, "epoch": 9.08, "learning_rate": 2.192235441452724e-05, "loss": 0.6589, "step": 10743, "task_loss": 0.5074435472488403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7482444047927856, "epoch": 9.08, "learning_rate": 2.1919223544145273e-05, "loss": 0.5049, "step": 10744, "task_loss": 0.6857324838638306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0258764028549194, "epoch": 9.08, "learning_rate": 2.1916092673763308e-05, "loss": 0.6469, "step": 10745, "task_loss": 0.7519462704658508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39565032720565796, "epoch": 9.08, "learning_rate": 2.191296180338134e-05, "loss": 0.4717, "step": 10746, "task_loss": 0.3737684488296509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6768993735313416, "epoch": 9.08, "learning_rate": 2.1909830932999375e-05, "loss": 0.5193, "step": 10747, "task_loss": 1.4171452522277832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5779070854187012, "epoch": 9.09, "learning_rate": 2.1906700062617407e-05, "loss": 0.5546, "step": 10748, "task_loss": 0.24212244153022766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5206707715988159, "epoch": 9.09, "learning_rate": 2.1903569192235442e-05, "loss": 0.4775, "step": 10749, "task_loss": 0.6817193031311035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5161746144294739, "epoch": 9.09, "learning_rate": 2.1900438321853474e-05, "loss": 0.5047, "step": 10750, "task_loss": 0.4374638795852661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3993494510650635, "epoch": 9.09, "learning_rate": 2.189730745147151e-05, "loss": 0.5669, "step": 10751, "task_loss": 0.4136887192726135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6239427328109741, "epoch": 9.09, "learning_rate": 2.1894176581089544e-05, "loss": 0.5883, "step": 10752, "task_loss": 0.42890116572380066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4716903865337372, "epoch": 9.09, "learning_rate": 2.189104571070758e-05, "loss": 0.5294, "step": 10753, "task_loss": 0.1877465397119522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6594790816307068, "epoch": 9.09, "learning_rate": 2.1887914840325614e-05, "loss": 0.543, "step": 10754, "task_loss": 1.039831280708313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26273614168167114, "epoch": 9.09, "learning_rate": 2.1884783969943646e-05, "loss": 0.5257, "step": 10755, "task_loss": 0.5761433243751526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5388988256454468, "epoch": 9.09, "learning_rate": 2.188165309956168e-05, "loss": 0.5764, "step": 10756, "task_loss": 0.4108578860759735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4670754671096802, "epoch": 9.09, "learning_rate": 2.1878522229179713e-05, "loss": 0.5036, "step": 10757, "task_loss": 0.34167376160621643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.403389036655426, "epoch": 9.09, "learning_rate": 2.187539135879775e-05, "loss": 0.7198, "step": 10758, "task_loss": 0.8958787322044373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4573601186275482, "epoch": 9.09, "learning_rate": 2.187226048841578e-05, "loss": 0.5556, "step": 10759, "task_loss": 0.46117568016052246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5540484189987183, "epoch": 9.1, "learning_rate": 2.1869129618033815e-05, "loss": 0.5166, "step": 10760, "task_loss": 0.07666939496994019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3661864995956421, "epoch": 9.1, "learning_rate": 2.1865998747651847e-05, "loss": 0.6611, "step": 10761, "task_loss": 0.12748773396015167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5516582727432251, "epoch": 9.1, "learning_rate": 2.1862867877269882e-05, "loss": 0.5767, "step": 10762, "task_loss": 0.7275339961051941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4251628816127777, "epoch": 9.1, "learning_rate": 2.1859737006887914e-05, "loss": 0.4644, "step": 10763, "task_loss": 0.7907235026359558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46637824177742004, "epoch": 9.1, "learning_rate": 2.185660613650595e-05, "loss": 0.4974, "step": 10764, "task_loss": 0.4630821645259857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5285096168518066, "epoch": 9.1, "learning_rate": 2.185347526612398e-05, "loss": 0.6476, "step": 10765, "task_loss": 1.2324399948120117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.686110258102417, "epoch": 9.1, "learning_rate": 2.1850344395742016e-05, "loss": 0.6657, "step": 10766, "task_loss": 0.5411072373390198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5917903184890747, "epoch": 9.1, "learning_rate": 2.184721352536005e-05, "loss": 0.5373, "step": 10767, "task_loss": 0.7401444911956787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.779970645904541, "epoch": 9.1, "learning_rate": 2.1844082654978083e-05, "loss": 0.6011, "step": 10768, "task_loss": 0.5900400876998901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.674863874912262, "epoch": 9.1, "learning_rate": 2.1840951784596118e-05, "loss": 0.613, "step": 10769, "task_loss": 0.5725447535514832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4761059284210205, "epoch": 9.1, "learning_rate": 2.1837820914214153e-05, "loss": 0.4667, "step": 10770, "task_loss": 0.17457948625087738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41168659925460815, "epoch": 9.1, "learning_rate": 2.183469004383219e-05, "loss": 0.5801, "step": 10771, "task_loss": 0.3928980231285095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5978654623031616, "epoch": 9.11, "learning_rate": 2.183155917345022e-05, "loss": 0.4444, "step": 10772, "task_loss": 0.9575604200363159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6103721857070923, "epoch": 9.11, "learning_rate": 2.1828428303068256e-05, "loss": 0.4757, "step": 10773, "task_loss": 1.303750991821289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9381407499313354, "epoch": 9.11, "learning_rate": 2.1825297432686287e-05, "loss": 0.6625, "step": 10774, "task_loss": 0.9570679068565369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.634233295917511, "epoch": 9.11, "learning_rate": 2.1822166562304323e-05, "loss": 0.7508, "step": 10775, "task_loss": 0.6843137741088867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0867902040481567, "epoch": 9.11, "learning_rate": 2.1819035691922354e-05, "loss": 0.521, "step": 10776, "task_loss": 0.6992771625518799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2684599161148071, "epoch": 9.11, "learning_rate": 2.181590482154039e-05, "loss": 0.7285, "step": 10777, "task_loss": 1.7703239917755127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5671638250350952, "epoch": 9.11, "learning_rate": 2.181277395115842e-05, "loss": 0.5465, "step": 10778, "task_loss": 0.8016722798347473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5883300304412842, "epoch": 9.11, "learning_rate": 2.1809643080776456e-05, "loss": 0.6282, "step": 10779, "task_loss": 0.7338662147521973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4316498637199402, "epoch": 9.11, "learning_rate": 2.180651221039449e-05, "loss": 0.6505, "step": 10780, "task_loss": 1.1932809352874756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7861335277557373, "epoch": 9.11, "learning_rate": 2.1803381340012523e-05, "loss": 0.6865, "step": 10781, "task_loss": 1.3743152618408203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5921609401702881, "epoch": 9.11, "learning_rate": 2.180025046963056e-05, "loss": 0.4248, "step": 10782, "task_loss": 0.9239550828933716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4874712824821472, "epoch": 9.11, "learning_rate": 2.179711959924859e-05, "loss": 0.4386, "step": 10783, "task_loss": 0.5710985660552979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6906114816665649, "epoch": 9.12, "learning_rate": 2.1793988728866625e-05, "loss": 0.4537, "step": 10784, "task_loss": 0.5026059150695801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.649724006652832, "epoch": 9.12, "learning_rate": 2.179085785848466e-05, "loss": 0.594, "step": 10785, "task_loss": 0.932915210723877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5842980742454529, "epoch": 9.12, "learning_rate": 2.1787726988102692e-05, "loss": 0.5681, "step": 10786, "task_loss": 0.41702306270599365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48341017961502075, "epoch": 9.12, "learning_rate": 2.1784596117720728e-05, "loss": 0.5353, "step": 10787, "task_loss": 0.6144500970840454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.863166093826294, "epoch": 9.12, "learning_rate": 2.1781465247338763e-05, "loss": 0.6704, "step": 10788, "task_loss": 0.511526882648468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7441883683204651, "epoch": 9.12, "learning_rate": 2.1778334376956795e-05, "loss": 0.5553, "step": 10789, "task_loss": 0.8669180274009705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6544240713119507, "epoch": 9.12, "learning_rate": 2.177520350657483e-05, "loss": 0.4782, "step": 10790, "task_loss": 0.8321774005889893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5973882675170898, "epoch": 9.12, "learning_rate": 2.1772072636192865e-05, "loss": 0.5071, "step": 10791, "task_loss": 0.21759475767612457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5827220678329468, "epoch": 9.12, "learning_rate": 2.1768941765810897e-05, "loss": 0.533, "step": 10792, "task_loss": 0.979975163936615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5982214212417603, "epoch": 9.12, "learning_rate": 2.1765810895428932e-05, "loss": 0.5663, "step": 10793, "task_loss": 1.1487809419631958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6295074224472046, "epoch": 9.12, "learning_rate": 2.1762680025046964e-05, "loss": 0.6524, "step": 10794, "task_loss": 0.4949423670768738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33164307475090027, "epoch": 9.13, "learning_rate": 2.1759549154665e-05, "loss": 0.5022, "step": 10795, "task_loss": 1.3993710279464722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.298446387052536, "epoch": 9.13, "learning_rate": 2.175641828428303e-05, "loss": 0.5687, "step": 10796, "task_loss": 0.371734082698822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5934202671051025, "epoch": 9.13, "learning_rate": 2.1753287413901066e-05, "loss": 0.5674, "step": 10797, "task_loss": 0.38397929072380066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.527718186378479, "epoch": 9.13, "learning_rate": 2.1750156543519098e-05, "loss": 0.7384, "step": 10798, "task_loss": 0.8474376201629639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6568619012832642, "epoch": 9.13, "learning_rate": 2.1747025673137133e-05, "loss": 0.4963, "step": 10799, "task_loss": 0.9808559417724609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4427694082260132, "epoch": 9.13, "learning_rate": 2.1743894802755164e-05, "loss": 0.6409, "step": 10800, "task_loss": 0.4030267894268036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5686321258544922, "epoch": 9.13, "learning_rate": 2.17407639323732e-05, "loss": 0.4984, "step": 10801, "task_loss": 0.6529268622398376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6223354935646057, "epoch": 9.13, "learning_rate": 2.1737633061991235e-05, "loss": 0.6462, "step": 10802, "task_loss": 2.0150671005249023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25852081179618835, "epoch": 9.13, "learning_rate": 2.173450219160927e-05, "loss": 0.4578, "step": 10803, "task_loss": 0.7179861068725586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44670596718788147, "epoch": 9.13, "learning_rate": 2.1731371321227305e-05, "loss": 0.4917, "step": 10804, "task_loss": 0.264409601688385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9250242114067078, "epoch": 9.13, "learning_rate": 2.1728240450845337e-05, "loss": 0.72, "step": 10805, "task_loss": 0.7922685742378235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39515572786331177, "epoch": 9.13, "learning_rate": 2.1725109580463372e-05, "loss": 0.635, "step": 10806, "task_loss": 0.8682048320770264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6662571430206299, "epoch": 9.14, "learning_rate": 2.1721978710081404e-05, "loss": 0.4507, "step": 10807, "task_loss": 0.7511534690856934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4091273844242096, "epoch": 9.14, "learning_rate": 2.171884783969944e-05, "loss": 0.6347, "step": 10808, "task_loss": 0.4866512715816498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4008023142814636, "epoch": 9.14, "learning_rate": 2.171571696931747e-05, "loss": 0.4735, "step": 10809, "task_loss": 0.5016112923622131 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6034723520278931, "epoch": 9.14, "learning_rate": 2.1712586098935506e-05, "loss": 0.6535, "step": 10810, "task_loss": 0.9633305072784424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6281410455703735, "epoch": 9.14, "learning_rate": 2.1709455228553538e-05, "loss": 0.5247, "step": 10811, "task_loss": 1.0469446182250977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39322221279144287, "epoch": 9.14, "learning_rate": 2.1706324358171573e-05, "loss": 0.4792, "step": 10812, "task_loss": 0.2486800104379654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4052883982658386, "epoch": 9.14, "learning_rate": 2.1703193487789605e-05, "loss": 0.3835, "step": 10813, "task_loss": 0.16745254397392273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.876307487487793, "epoch": 9.14, "learning_rate": 2.170006261740764e-05, "loss": 0.5622, "step": 10814, "task_loss": 0.42031747102737427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4564352035522461, "epoch": 9.14, "learning_rate": 2.1696931747025675e-05, "loss": 0.4624, "step": 10815, "task_loss": 0.8987082839012146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39727982878685, "epoch": 9.14, "learning_rate": 2.1693800876643707e-05, "loss": 0.4268, "step": 10816, "task_loss": 0.788692831993103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.385593980550766, "epoch": 9.14, "learning_rate": 2.1690670006261742e-05, "loss": 0.5844, "step": 10817, "task_loss": 0.720504105091095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49342530965805054, "epoch": 9.14, "learning_rate": 2.1687539135879774e-05, "loss": 0.6521, "step": 10818, "task_loss": 0.6234811544418335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4022221565246582, "epoch": 9.15, "learning_rate": 2.168440826549781e-05, "loss": 0.5156, "step": 10819, "task_loss": 0.20833280682563782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5778744220733643, "epoch": 9.15, "learning_rate": 2.1681277395115844e-05, "loss": 0.7791, "step": 10820, "task_loss": 0.701546847820282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3403276205062866, "epoch": 9.15, "learning_rate": 2.167814652473388e-05, "loss": 0.4568, "step": 10821, "task_loss": 0.27805864810943604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8931653499603271, "epoch": 9.15, "learning_rate": 2.167501565435191e-05, "loss": 0.6631, "step": 10822, "task_loss": 0.8117380738258362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6478039622306824, "epoch": 9.15, "learning_rate": 2.1671884783969946e-05, "loss": 0.6529, "step": 10823, "task_loss": 0.5761736035346985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41512200236320496, "epoch": 9.15, "learning_rate": 2.1668753913587978e-05, "loss": 0.435, "step": 10824, "task_loss": 0.30861127376556396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8149711489677429, "epoch": 9.15, "learning_rate": 2.1665623043206013e-05, "loss": 0.6513, "step": 10825, "task_loss": 0.7156995534896851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22911061346530914, "epoch": 9.15, "learning_rate": 2.1662492172824045e-05, "loss": 0.547, "step": 10826, "task_loss": 0.24998728930950165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7221283316612244, "epoch": 9.15, "learning_rate": 2.165936130244208e-05, "loss": 0.5939, "step": 10827, "task_loss": 1.4320980310440063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.331041544675827, "epoch": 9.15, "learning_rate": 2.1656230432060115e-05, "loss": 0.4862, "step": 10828, "task_loss": 0.37661001086235046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3271137475967407, "epoch": 9.15, "learning_rate": 2.1653099561678147e-05, "loss": 0.4876, "step": 10829, "task_loss": 0.1133360043168068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6177870035171509, "epoch": 9.15, "learning_rate": 2.1649968691296182e-05, "loss": 0.5236, "step": 10830, "task_loss": 1.8078876733779907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2890002429485321, "epoch": 9.16, "learning_rate": 2.1646837820914214e-05, "loss": 0.5598, "step": 10831, "task_loss": 0.5552334785461426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6078407764434814, "epoch": 9.16, "learning_rate": 2.164370695053225e-05, "loss": 0.6224, "step": 10832, "task_loss": 1.142133355140686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4893971085548401, "epoch": 9.16, "learning_rate": 2.164057608015028e-05, "loss": 0.6012, "step": 10833, "task_loss": 0.9203267097473145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38470345735549927, "epoch": 9.16, "learning_rate": 2.1637445209768316e-05, "loss": 0.3464, "step": 10834, "task_loss": 0.6152617335319519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3799683749675751, "epoch": 9.16, "learning_rate": 2.1634314339386348e-05, "loss": 0.6634, "step": 10835, "task_loss": 0.38013580441474915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5191724896430969, "epoch": 9.16, "learning_rate": 2.1631183469004383e-05, "loss": 0.7478, "step": 10836, "task_loss": 0.7657111883163452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0082024335861206, "epoch": 9.16, "learning_rate": 2.1628052598622418e-05, "loss": 0.568, "step": 10837, "task_loss": 1.3488949537277222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32939714193344116, "epoch": 9.16, "learning_rate": 2.1624921728240453e-05, "loss": 0.6781, "step": 10838, "task_loss": 0.3038693070411682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5220847129821777, "epoch": 9.16, "learning_rate": 2.1621790857858485e-05, "loss": 0.5591, "step": 10839, "task_loss": 0.4311464726924896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6362727880477905, "epoch": 9.16, "learning_rate": 2.161865998747652e-05, "loss": 0.534, "step": 10840, "task_loss": 0.48234328627586365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6126822233200073, "epoch": 9.16, "learning_rate": 2.1615529117094556e-05, "loss": 0.6748, "step": 10841, "task_loss": 0.8049226999282837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3340612053871155, "epoch": 9.16, "learning_rate": 2.1612398246712587e-05, "loss": 0.6478, "step": 10842, "task_loss": 0.33988818526268005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5327385067939758, "epoch": 9.17, "learning_rate": 2.1609267376330623e-05, "loss": 0.4681, "step": 10843, "task_loss": 1.181688666343689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6318315863609314, "epoch": 9.17, "learning_rate": 2.1606136505948654e-05, "loss": 0.8218, "step": 10844, "task_loss": 0.5662764310836792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.278198778629303, "epoch": 9.17, "learning_rate": 2.160300563556669e-05, "loss": 0.5392, "step": 10845, "task_loss": 0.08471614122390747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5648735165596008, "epoch": 9.17, "learning_rate": 2.159987476518472e-05, "loss": 0.523, "step": 10846, "task_loss": 0.7913439273834229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5863088369369507, "epoch": 9.17, "learning_rate": 2.1596743894802756e-05, "loss": 0.4841, "step": 10847, "task_loss": 0.7385919690132141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43614083528518677, "epoch": 9.17, "learning_rate": 2.1593613024420788e-05, "loss": 0.4151, "step": 10848, "task_loss": 0.45695215463638306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4713941812515259, "epoch": 9.17, "learning_rate": 2.1590482154038823e-05, "loss": 0.4631, "step": 10849, "task_loss": 0.4377646744251251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5819191932678223, "epoch": 9.17, "learning_rate": 2.1587351283656855e-05, "loss": 0.517, "step": 10850, "task_loss": 1.1084933280944824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6137416362762451, "epoch": 9.17, "learning_rate": 2.158422041327489e-05, "loss": 0.5187, "step": 10851, "task_loss": 0.7242845296859741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.64979088306427, "epoch": 9.17, "learning_rate": 2.1581089542892926e-05, "loss": 0.4929, "step": 10852, "task_loss": 0.7142972946166992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6068898439407349, "epoch": 9.17, "learning_rate": 2.1577958672510957e-05, "loss": 0.6551, "step": 10853, "task_loss": 1.2356534004211426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.749579906463623, "epoch": 9.17, "learning_rate": 2.1574827802128992e-05, "loss": 0.6339, "step": 10854, "task_loss": 0.998931884765625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42090779542922974, "epoch": 9.18, "learning_rate": 2.1571696931747028e-05, "loss": 0.5442, "step": 10855, "task_loss": 0.8587653636932373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7019698619842529, "epoch": 9.18, "learning_rate": 2.1568566061365063e-05, "loss": 0.5892, "step": 10856, "task_loss": 1.4216864109039307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3970860242843628, "epoch": 9.18, "learning_rate": 2.1565435190983095e-05, "loss": 0.5143, "step": 10857, "task_loss": 0.21091079711914062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5980845093727112, "epoch": 9.18, "learning_rate": 2.156230432060113e-05, "loss": 0.5236, "step": 10858, "task_loss": 0.3505527675151825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3183314800262451, "epoch": 9.18, "learning_rate": 2.155917345021916e-05, "loss": 0.5074, "step": 10859, "task_loss": 0.52081298828125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32852256298065186, "epoch": 9.18, "learning_rate": 2.1556042579837197e-05, "loss": 0.4619, "step": 10860, "task_loss": 0.3196869194507599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3657461702823639, "epoch": 9.18, "learning_rate": 2.155291170945523e-05, "loss": 0.4814, "step": 10861, "task_loss": 0.2634945809841156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4061799645423889, "epoch": 9.18, "learning_rate": 2.1549780839073264e-05, "loss": 0.5277, "step": 10862, "task_loss": 0.25397977232933044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5767486095428467, "epoch": 9.18, "learning_rate": 2.1546649968691295e-05, "loss": 0.4107, "step": 10863, "task_loss": 0.5107080340385437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34271687269210815, "epoch": 9.18, "learning_rate": 2.154351909830933e-05, "loss": 0.471, "step": 10864, "task_loss": 1.546038031578064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8187350034713745, "epoch": 9.18, "learning_rate": 2.1540388227927366e-05, "loss": 0.5822, "step": 10865, "task_loss": 0.5808723568916321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38430267572402954, "epoch": 9.19, "learning_rate": 2.1537257357545398e-05, "loss": 0.4973, "step": 10866, "task_loss": 1.3076395988464355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3123360872268677, "epoch": 9.19, "learning_rate": 2.1534126487163433e-05, "loss": 0.5981, "step": 10867, "task_loss": 0.4841679632663727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4629463851451874, "epoch": 9.19, "learning_rate": 2.1530995616781465e-05, "loss": 0.5404, "step": 10868, "task_loss": 0.5289047360420227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47404739260673523, "epoch": 9.19, "learning_rate": 2.15278647463995e-05, "loss": 0.6401, "step": 10869, "task_loss": 0.3344303071498871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5069639682769775, "epoch": 9.19, "learning_rate": 2.1524733876017535e-05, "loss": 0.4664, "step": 10870, "task_loss": 0.6204138994216919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45518067479133606, "epoch": 9.19, "learning_rate": 2.1521603005635567e-05, "loss": 0.5207, "step": 10871, "task_loss": 1.2183525562286377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.622827410697937, "epoch": 9.19, "learning_rate": 2.1518472135253602e-05, "loss": 0.7393, "step": 10872, "task_loss": 0.7857483625411987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28328555822372437, "epoch": 9.19, "learning_rate": 2.1515341264871637e-05, "loss": 0.5001, "step": 10873, "task_loss": 0.1773114949464798 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43933287262916565, "epoch": 9.19, "learning_rate": 2.151221039448967e-05, "loss": 0.6023, "step": 10874, "task_loss": 1.2644399404525757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40510454773902893, "epoch": 9.19, "learning_rate": 2.1509079524107704e-05, "loss": 0.471, "step": 10875, "task_loss": 0.5119494199752808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4197666645050049, "epoch": 9.19, "learning_rate": 2.1505948653725736e-05, "loss": 0.6526, "step": 10876, "task_loss": 1.0912022590637207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8527094125747681, "epoch": 9.19, "learning_rate": 2.150281778334377e-05, "loss": 0.5265, "step": 10877, "task_loss": 0.7256184220314026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46835994720458984, "epoch": 9.2, "learning_rate": 2.1499686912961806e-05, "loss": 0.562, "step": 10878, "task_loss": 0.09248249232769012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9534649848937988, "epoch": 9.2, "learning_rate": 2.1496556042579838e-05, "loss": 0.638, "step": 10879, "task_loss": 2.0141379833221436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4206981062889099, "epoch": 9.2, "learning_rate": 2.1493425172197873e-05, "loss": 0.637, "step": 10880, "task_loss": 0.11708416044712067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5128781199455261, "epoch": 9.2, "learning_rate": 2.1490294301815905e-05, "loss": 0.5002, "step": 10881, "task_loss": 0.6460906863212585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6144223809242249, "epoch": 9.2, "learning_rate": 2.148716343143394e-05, "loss": 0.5401, "step": 10882, "task_loss": 0.7632240056991577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5146634578704834, "epoch": 9.2, "learning_rate": 2.1484032561051972e-05, "loss": 0.665, "step": 10883, "task_loss": 0.6760275363922119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6023615598678589, "epoch": 9.2, "learning_rate": 2.1480901690670007e-05, "loss": 0.5366, "step": 10884, "task_loss": 0.7749180793762207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37847283482551575, "epoch": 9.2, "learning_rate": 2.147777082028804e-05, "loss": 0.4887, "step": 10885, "task_loss": 0.04881501570343971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.561476469039917, "epoch": 9.2, "learning_rate": 2.1474639949906074e-05, "loss": 0.6617, "step": 10886, "task_loss": 0.7930432558059692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6836427450180054, "epoch": 9.2, "learning_rate": 2.147150907952411e-05, "loss": 0.5758, "step": 10887, "task_loss": 1.0476616621017456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45852354168891907, "epoch": 9.2, "learning_rate": 2.1468378209142144e-05, "loss": 0.5085, "step": 10888, "task_loss": 1.0976526737213135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47180649638175964, "epoch": 9.2, "learning_rate": 2.146524733876018e-05, "loss": 0.5225, "step": 10889, "task_loss": 0.32783204317092896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5903705358505249, "epoch": 9.21, "learning_rate": 2.146211646837821e-05, "loss": 0.5872, "step": 10890, "task_loss": 0.947970449924469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49364399909973145, "epoch": 9.21, "learning_rate": 2.1458985597996246e-05, "loss": 0.6807, "step": 10891, "task_loss": 0.1938280612230301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1805577278137207, "epoch": 9.21, "learning_rate": 2.1455854727614278e-05, "loss": 0.6971, "step": 10892, "task_loss": 0.7009936571121216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6477997303009033, "epoch": 9.21, "learning_rate": 2.1452723857232313e-05, "loss": 0.5428, "step": 10893, "task_loss": 0.8989834785461426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43821266293525696, "epoch": 9.21, "learning_rate": 2.1449592986850345e-05, "loss": 0.3353, "step": 10894, "task_loss": 0.4989883303642273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5481058359146118, "epoch": 9.21, "learning_rate": 2.144646211646838e-05, "loss": 0.606, "step": 10895, "task_loss": 0.42359623312950134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49851876497268677, "epoch": 9.21, "learning_rate": 2.1443331246086412e-05, "loss": 0.5283, "step": 10896, "task_loss": 0.9212086796760559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5051236152648926, "epoch": 9.21, "learning_rate": 2.1440200375704447e-05, "loss": 0.4761, "step": 10897, "task_loss": 0.5388107299804688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3500555455684662, "epoch": 9.21, "learning_rate": 2.143706950532248e-05, "loss": 0.417, "step": 10898, "task_loss": 0.30302706360816956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7048258185386658, "epoch": 9.21, "learning_rate": 2.1433938634940514e-05, "loss": 0.6708, "step": 10899, "task_loss": 1.018332839012146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4283768832683563, "epoch": 9.21, "learning_rate": 2.1430807764558546e-05, "loss": 0.4669, "step": 10900, "task_loss": 0.6607311964035034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.68162602186203, "epoch": 9.21, "learning_rate": 2.142767689417658e-05, "loss": 0.6109, "step": 10901, "task_loss": 0.9482491612434387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4650007486343384, "epoch": 9.22, "learning_rate": 2.1424546023794616e-05, "loss": 0.5162, "step": 10902, "task_loss": 0.6551792025566101 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3212777078151703, "epoch": 9.22, "learning_rate": 2.1421415153412648e-05, "loss": 0.4835, "step": 10903, "task_loss": 0.30251339077949524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6223666667938232, "epoch": 9.22, "learning_rate": 2.1418284283030683e-05, "loss": 0.6115, "step": 10904, "task_loss": 0.6684866547584534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33913445472717285, "epoch": 9.22, "learning_rate": 2.141515341264872e-05, "loss": 0.4681, "step": 10905, "task_loss": 0.6508069038391113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4184630811214447, "epoch": 9.22, "learning_rate": 2.1412022542266754e-05, "loss": 0.4815, "step": 10906, "task_loss": 1.0698230266571045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5623041391372681, "epoch": 9.22, "learning_rate": 2.1408891671884785e-05, "loss": 0.5631, "step": 10907, "task_loss": 0.6564867496490479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.519415020942688, "epoch": 9.22, "learning_rate": 2.140576080150282e-05, "loss": 0.4941, "step": 10908, "task_loss": 0.4757525324821472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7588217258453369, "epoch": 9.22, "learning_rate": 2.1402629931120852e-05, "loss": 0.5354, "step": 10909, "task_loss": 1.059624433517456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5385714173316956, "epoch": 9.22, "learning_rate": 2.1399499060738887e-05, "loss": 0.5874, "step": 10910, "task_loss": 1.1655223369598389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4426650404930115, "epoch": 9.22, "learning_rate": 2.139636819035692e-05, "loss": 0.7159, "step": 10911, "task_loss": 0.9254082441329956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44767943024635315, "epoch": 9.22, "learning_rate": 2.1393237319974954e-05, "loss": 0.5912, "step": 10912, "task_loss": 0.4576539993286133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6747550368309021, "epoch": 9.22, "learning_rate": 2.139010644959299e-05, "loss": 0.48, "step": 10913, "task_loss": 0.8630843758583069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5697304606437683, "epoch": 9.23, "learning_rate": 2.138697557921102e-05, "loss": 0.535, "step": 10914, "task_loss": 1.1150908470153809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2988319993019104, "epoch": 9.23, "learning_rate": 2.1383844708829056e-05, "loss": 0.4021, "step": 10915, "task_loss": 0.5637917518615723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8878071308135986, "epoch": 9.23, "learning_rate": 2.1380713838447088e-05, "loss": 0.6069, "step": 10916, "task_loss": 1.326325535774231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5750740766525269, "epoch": 9.23, "learning_rate": 2.1377582968065123e-05, "loss": 0.5801, "step": 10917, "task_loss": 1.168007254600525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3625173568725586, "epoch": 9.23, "learning_rate": 2.1374452097683155e-05, "loss": 0.6885, "step": 10918, "task_loss": 0.24815595149993896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5042824745178223, "epoch": 9.23, "learning_rate": 2.137132122730119e-05, "loss": 0.6116, "step": 10919, "task_loss": 1.0247471332550049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5384512543678284, "epoch": 9.23, "learning_rate": 2.1368190356919222e-05, "loss": 0.5173, "step": 10920, "task_loss": 0.910983145236969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8035609722137451, "epoch": 9.23, "learning_rate": 2.1365059486537257e-05, "loss": 0.6316, "step": 10921, "task_loss": 0.8519521355628967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43562668561935425, "epoch": 9.23, "learning_rate": 2.1361928616155293e-05, "loss": 0.5967, "step": 10922, "task_loss": 0.5298208594322205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8232869505882263, "epoch": 9.23, "learning_rate": 2.1358797745773328e-05, "loss": 0.7342, "step": 10923, "task_loss": 1.173535704612732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22864538431167603, "epoch": 9.23, "learning_rate": 2.135566687539136e-05, "loss": 0.3646, "step": 10924, "task_loss": 0.12770842015743256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6853054761886597, "epoch": 9.23, "learning_rate": 2.1352536005009395e-05, "loss": 0.4477, "step": 10925, "task_loss": 0.8653681874275208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4213179349899292, "epoch": 9.24, "learning_rate": 2.134940513462743e-05, "loss": 0.4544, "step": 10926, "task_loss": 0.6201899647712708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5321208238601685, "epoch": 9.24, "learning_rate": 2.134627426424546e-05, "loss": 0.3299, "step": 10927, "task_loss": 0.06832588464021683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38499921560287476, "epoch": 9.24, "learning_rate": 2.1343143393863497e-05, "loss": 0.5256, "step": 10928, "task_loss": 0.6429681777954102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2602386176586151, "epoch": 9.24, "learning_rate": 2.134001252348153e-05, "loss": 0.3604, "step": 10929, "task_loss": 0.3457573354244232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42229723930358887, "epoch": 9.24, "learning_rate": 2.1336881653099564e-05, "loss": 0.586, "step": 10930, "task_loss": 0.7372218370437622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5491085052490234, "epoch": 9.24, "learning_rate": 2.1333750782717595e-05, "loss": 0.5898, "step": 10931, "task_loss": 1.244326114654541 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6268080472946167, "epoch": 9.24, "learning_rate": 2.133061991233563e-05, "loss": 0.5697, "step": 10932, "task_loss": 1.1351968050003052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47410494089126587, "epoch": 9.24, "learning_rate": 2.1327489041953662e-05, "loss": 0.4326, "step": 10933, "task_loss": 0.6800298094749451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5500917434692383, "epoch": 9.24, "learning_rate": 2.1324358171571698e-05, "loss": 0.4065, "step": 10934, "task_loss": 0.23375087976455688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38254427909851074, "epoch": 9.24, "learning_rate": 2.132122730118973e-05, "loss": 0.4711, "step": 10935, "task_loss": 0.7151814699172974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8779199123382568, "epoch": 9.24, "learning_rate": 2.1318096430807765e-05, "loss": 0.68, "step": 10936, "task_loss": 0.8926211595535278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8139522075653076, "epoch": 9.24, "learning_rate": 2.13149655604258e-05, "loss": 0.6005, "step": 10937, "task_loss": 0.8239250183105469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4297165870666504, "epoch": 9.25, "learning_rate": 2.131183469004383e-05, "loss": 0.5709, "step": 10938, "task_loss": 0.3510212004184723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9361065030097961, "epoch": 9.25, "learning_rate": 2.1308703819661867e-05, "loss": 0.7119, "step": 10939, "task_loss": 1.0182311534881592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5342442989349365, "epoch": 9.25, "learning_rate": 2.1305572949279902e-05, "loss": 0.6202, "step": 10940, "task_loss": 0.3514188230037689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4866122901439667, "epoch": 9.25, "learning_rate": 2.1302442078897937e-05, "loss": 0.5528, "step": 10941, "task_loss": 1.4298650026321411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5200307369232178, "epoch": 9.25, "learning_rate": 2.129931120851597e-05, "loss": 0.5023, "step": 10942, "task_loss": 0.39539921283721924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.657658576965332, "epoch": 9.25, "learning_rate": 2.1296180338134004e-05, "loss": 0.5227, "step": 10943, "task_loss": 0.972457766532898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6207591891288757, "epoch": 9.25, "learning_rate": 2.1293049467752036e-05, "loss": 0.503, "step": 10944, "task_loss": 0.5119209289550781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7298190593719482, "epoch": 9.25, "learning_rate": 2.128991859737007e-05, "loss": 0.4535, "step": 10945, "task_loss": 1.1515848636627197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.712441623210907, "epoch": 9.25, "learning_rate": 2.1286787726988103e-05, "loss": 0.6854, "step": 10946, "task_loss": 1.038048505783081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5885166525840759, "epoch": 9.25, "learning_rate": 2.1283656856606138e-05, "loss": 0.4894, "step": 10947, "task_loss": 0.6588267087936401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6577605605125427, "epoch": 9.25, "learning_rate": 2.128052598622417e-05, "loss": 0.5615, "step": 10948, "task_loss": 1.1200001239776611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5500950217247009, "epoch": 9.26, "learning_rate": 2.1277395115842205e-05, "loss": 0.6323, "step": 10949, "task_loss": 1.4862384796142578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37843790650367737, "epoch": 9.26, "learning_rate": 2.127426424546024e-05, "loss": 0.5844, "step": 10950, "task_loss": 1.0339802503585815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3949783444404602, "epoch": 9.26, "learning_rate": 2.1271133375078272e-05, "loss": 0.6939, "step": 10951, "task_loss": 0.8240654468536377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3597279191017151, "epoch": 9.26, "learning_rate": 2.1268002504696307e-05, "loss": 0.4818, "step": 10952, "task_loss": 1.1835144758224487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6641346216201782, "epoch": 9.26, "learning_rate": 2.126487163431434e-05, "loss": 0.6901, "step": 10953, "task_loss": 0.9962551593780518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6204301714897156, "epoch": 9.26, "learning_rate": 2.1261740763932374e-05, "loss": 0.6701, "step": 10954, "task_loss": 0.9233690500259399 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39548370242118835, "epoch": 9.26, "learning_rate": 2.125860989355041e-05, "loss": 0.4887, "step": 10955, "task_loss": 0.8051570057868958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4969893991947174, "epoch": 9.26, "learning_rate": 2.1255479023168444e-05, "loss": 0.4946, "step": 10956, "task_loss": 1.5789693593978882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46021026372909546, "epoch": 9.26, "learning_rate": 2.1252348152786476e-05, "loss": 0.6607, "step": 10957, "task_loss": 0.6119105815887451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4000053107738495, "epoch": 9.26, "learning_rate": 2.124921728240451e-05, "loss": 0.6101, "step": 10958, "task_loss": 0.23214876651763916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.843894362449646, "epoch": 9.26, "learning_rate": 2.1246086412022543e-05, "loss": 0.5743, "step": 10959, "task_loss": 0.24309709668159485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3611828088760376, "epoch": 9.26, "learning_rate": 2.1242955541640578e-05, "loss": 0.4667, "step": 10960, "task_loss": 0.16456349194049835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42099568247795105, "epoch": 9.27, "learning_rate": 2.123982467125861e-05, "loss": 0.5479, "step": 10961, "task_loss": 1.4007837772369385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41204509139060974, "epoch": 9.27, "learning_rate": 2.1236693800876645e-05, "loss": 0.473, "step": 10962, "task_loss": 0.29872259497642517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6341841816902161, "epoch": 9.27, "learning_rate": 2.123356293049468e-05, "loss": 0.5099, "step": 10963, "task_loss": 0.43977656960487366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5356093645095825, "epoch": 9.27, "learning_rate": 2.1230432060112712e-05, "loss": 0.432, "step": 10964, "task_loss": 0.51674485206604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33958911895751953, "epoch": 9.27, "learning_rate": 2.1227301189730747e-05, "loss": 0.4868, "step": 10965, "task_loss": 0.402753084897995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6170613169670105, "epoch": 9.27, "learning_rate": 2.122417031934878e-05, "loss": 0.5812, "step": 10966, "task_loss": 0.9869973659515381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7055357694625854, "epoch": 9.27, "learning_rate": 2.1221039448966814e-05, "loss": 0.569, "step": 10967, "task_loss": 0.5293646454811096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47522932291030884, "epoch": 9.27, "learning_rate": 2.1217908578584846e-05, "loss": 0.5239, "step": 10968, "task_loss": 1.0056711435317993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.523888349533081, "epoch": 9.27, "learning_rate": 2.121477770820288e-05, "loss": 0.4695, "step": 10969, "task_loss": 0.8720892071723938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.57622230052948, "epoch": 9.27, "learning_rate": 2.1211646837820913e-05, "loss": 0.5384, "step": 10970, "task_loss": 0.7073364853858948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46024948358535767, "epoch": 9.27, "learning_rate": 2.1208515967438948e-05, "loss": 0.5464, "step": 10971, "task_loss": 1.3565375804901123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28800565004348755, "epoch": 9.27, "learning_rate": 2.1205385097056983e-05, "loss": 0.4767, "step": 10972, "task_loss": 0.11874084174633026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2343231588602066, "epoch": 9.28, "learning_rate": 2.120225422667502e-05, "loss": 0.3965, "step": 10973, "task_loss": 0.3304547965526581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5238041281700134, "epoch": 9.28, "learning_rate": 2.119912335629305e-05, "loss": 0.5479, "step": 10974, "task_loss": 0.38163554668426514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6606035232543945, "epoch": 9.28, "learning_rate": 2.1195992485911085e-05, "loss": 0.6772, "step": 10975, "task_loss": 1.386702060699463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6085480451583862, "epoch": 9.28, "learning_rate": 2.119286161552912e-05, "loss": 0.5927, "step": 10976, "task_loss": 0.5691719651222229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5871055722236633, "epoch": 9.28, "learning_rate": 2.1189730745147152e-05, "loss": 0.532, "step": 10977, "task_loss": 0.5200388431549072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42819955945014954, "epoch": 9.28, "learning_rate": 2.1186599874765187e-05, "loss": 0.6269, "step": 10978, "task_loss": 0.5926833152770996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9046127200126648, "epoch": 9.28, "learning_rate": 2.118346900438322e-05, "loss": 0.5692, "step": 10979, "task_loss": 0.7268161177635193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6609224677085876, "epoch": 9.28, "learning_rate": 2.1180338134001254e-05, "loss": 0.5185, "step": 10980, "task_loss": 0.8057270646095276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9564125537872314, "epoch": 9.28, "learning_rate": 2.1177207263619286e-05, "loss": 0.589, "step": 10981, "task_loss": 1.8332239389419556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4116284251213074, "epoch": 9.28, "learning_rate": 2.117407639323732e-05, "loss": 0.4688, "step": 10982, "task_loss": 0.10509831458330154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.914676308631897, "epoch": 9.28, "learning_rate": 2.1170945522855353e-05, "loss": 0.7386, "step": 10983, "task_loss": 1.3399028778076172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5786545872688293, "epoch": 9.28, "learning_rate": 2.1167814652473388e-05, "loss": 0.5958, "step": 10984, "task_loss": 0.9887636303901672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22716110944747925, "epoch": 9.29, "learning_rate": 2.116468378209142e-05, "loss": 0.4696, "step": 10985, "task_loss": 0.1664201021194458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.547349214553833, "epoch": 9.29, "learning_rate": 2.1161552911709455e-05, "loss": 0.6009, "step": 10986, "task_loss": 0.5663459897041321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7757431268692017, "epoch": 9.29, "learning_rate": 2.115842204132749e-05, "loss": 0.8008, "step": 10987, "task_loss": 0.7495940923690796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42252543568611145, "epoch": 9.29, "learning_rate": 2.1155291170945522e-05, "loss": 0.5237, "step": 10988, "task_loss": 0.758827269077301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.639927089214325, "epoch": 9.29, "learning_rate": 2.1152160300563557e-05, "loss": 0.7558, "step": 10989, "task_loss": 0.17115819454193115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6819243431091309, "epoch": 9.29, "learning_rate": 2.1149029430181593e-05, "loss": 0.6426, "step": 10990, "task_loss": 0.8094286918640137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4794903099536896, "epoch": 9.29, "learning_rate": 2.1145898559799628e-05, "loss": 0.5282, "step": 10991, "task_loss": 1.317718267440796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44821491837501526, "epoch": 9.29, "learning_rate": 2.114276768941766e-05, "loss": 0.5447, "step": 10992, "task_loss": 0.48795199394226074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7978559136390686, "epoch": 9.29, "learning_rate": 2.1139636819035695e-05, "loss": 0.5369, "step": 10993, "task_loss": 0.9549655318260193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.347912073135376, "epoch": 9.29, "learning_rate": 2.1136505948653726e-05, "loss": 0.4704, "step": 10994, "task_loss": 0.7751814126968384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8824114799499512, "epoch": 9.29, "learning_rate": 2.113337507827176e-05, "loss": 0.5175, "step": 10995, "task_loss": 0.7911750078201294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3624292314052582, "epoch": 9.29, "learning_rate": 2.1130244207889793e-05, "loss": 0.4222, "step": 10996, "task_loss": 0.6262227296829224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5484211444854736, "epoch": 9.3, "learning_rate": 2.112711333750783e-05, "loss": 0.5618, "step": 10997, "task_loss": 0.6928092837333679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.721632719039917, "epoch": 9.3, "learning_rate": 2.112398246712586e-05, "loss": 0.5204, "step": 10998, "task_loss": 1.150126576423645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5535643696784973, "epoch": 9.3, "learning_rate": 2.1120851596743896e-05, "loss": 0.4267, "step": 10999, "task_loss": 0.3417198657989502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7252011299133301, "epoch": 9.3, "learning_rate": 2.111772072636193e-05, "loss": 0.5323, "step": 11000, "task_loss": 0.466799795627594 }, { "epoch": 9.3, "eval_accuracy": 0.9016633663366337, "eval_loss": 0.3793487250804901, "eval_runtime": 206.1337, "eval_samples_per_second": 122.493, "eval_steps_per_second": 0.961, "step": 11000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9623677730560303, "epoch": 9.3, "learning_rate": 2.1114589855979962e-05, "loss": 0.7295, "step": 11001, "task_loss": 0.47625887393951416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5299513339996338, "epoch": 9.3, "learning_rate": 2.1111458985597998e-05, "loss": 0.6378, "step": 11002, "task_loss": 1.240906834602356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5188761353492737, "epoch": 9.3, "learning_rate": 2.110832811521603e-05, "loss": 0.602, "step": 11003, "task_loss": 1.153619408607483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0214561223983765, "epoch": 9.3, "learning_rate": 2.1105197244834065e-05, "loss": 0.5579, "step": 11004, "task_loss": 1.2658144235610962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6547983884811401, "epoch": 9.3, "learning_rate": 2.1102066374452096e-05, "loss": 0.5196, "step": 11005, "task_loss": 0.5332574844360352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7137526273727417, "epoch": 9.3, "learning_rate": 2.109893550407013e-05, "loss": 0.7594, "step": 11006, "task_loss": 0.2760656476020813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0427443981170654, "epoch": 9.3, "learning_rate": 2.1095804633688167e-05, "loss": 0.7596, "step": 11007, "task_loss": 0.5587009787559509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8365087509155273, "epoch": 9.3, "learning_rate": 2.1092673763306202e-05, "loss": 0.4569, "step": 11008, "task_loss": 1.0348951816558838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21145832538604736, "epoch": 9.31, "learning_rate": 2.1089542892924234e-05, "loss": 0.5034, "step": 11009, "task_loss": 0.4591176211833954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4173012673854828, "epoch": 9.31, "learning_rate": 2.108641202254227e-05, "loss": 0.4327, "step": 11010, "task_loss": 0.49352967739105225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21527692675590515, "epoch": 9.31, "learning_rate": 2.10832811521603e-05, "loss": 0.4777, "step": 11011, "task_loss": 0.16324567794799805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.807902455329895, "epoch": 9.31, "learning_rate": 2.1080150281778336e-05, "loss": 0.6917, "step": 11012, "task_loss": 0.6279479265213013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35721275210380554, "epoch": 9.31, "learning_rate": 2.107701941139637e-05, "loss": 0.5393, "step": 11013, "task_loss": 0.17685876786708832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4954848885536194, "epoch": 9.31, "learning_rate": 2.1073888541014403e-05, "loss": 0.5564, "step": 11014, "task_loss": 0.5993374586105347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5603012442588806, "epoch": 9.31, "learning_rate": 2.1070757670632438e-05, "loss": 0.5717, "step": 11015, "task_loss": 0.8402811288833618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5526353716850281, "epoch": 9.31, "learning_rate": 2.106762680025047e-05, "loss": 0.5885, "step": 11016, "task_loss": 0.946648895740509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6028943061828613, "epoch": 9.31, "learning_rate": 2.1064495929868505e-05, "loss": 0.488, "step": 11017, "task_loss": 0.46591758728027344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6554441452026367, "epoch": 9.31, "learning_rate": 2.1061365059486537e-05, "loss": 0.5368, "step": 11018, "task_loss": 1.2420228719711304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37560293078422546, "epoch": 9.31, "learning_rate": 2.1058234189104572e-05, "loss": 0.5778, "step": 11019, "task_loss": 0.549220085144043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4517941474914551, "epoch": 9.32, "learning_rate": 2.1055103318722604e-05, "loss": 0.6208, "step": 11020, "task_loss": 0.7067399621009827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46670758724212646, "epoch": 9.32, "learning_rate": 2.105197244834064e-05, "loss": 0.5431, "step": 11021, "task_loss": 0.41013893485069275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4980430603027344, "epoch": 9.32, "learning_rate": 2.1048841577958674e-05, "loss": 0.6122, "step": 11022, "task_loss": 0.5611720085144043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7161940336227417, "epoch": 9.32, "learning_rate": 2.1045710707576706e-05, "loss": 0.5064, "step": 11023, "task_loss": 0.36962953209877014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.863868236541748, "epoch": 9.32, "learning_rate": 2.104257983719474e-05, "loss": 0.6388, "step": 11024, "task_loss": 0.8607770800590515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7176629304885864, "epoch": 9.32, "learning_rate": 2.1039448966812776e-05, "loss": 0.4785, "step": 11025, "task_loss": 0.9822527766227722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4039854407310486, "epoch": 9.32, "learning_rate": 2.103631809643081e-05, "loss": 0.5962, "step": 11026, "task_loss": 0.787648618221283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6130164861679077, "epoch": 9.32, "learning_rate": 2.1033187226048843e-05, "loss": 0.4608, "step": 11027, "task_loss": 1.0217814445495605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8617525696754456, "epoch": 9.32, "learning_rate": 2.1030056355666878e-05, "loss": 0.5892, "step": 11028, "task_loss": 0.692875862121582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45631664991378784, "epoch": 9.32, "learning_rate": 2.102692548528491e-05, "loss": 0.5351, "step": 11029, "task_loss": 1.0149005651474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.762970507144928, "epoch": 9.32, "learning_rate": 2.1023794614902945e-05, "loss": 0.5343, "step": 11030, "task_loss": 0.49640247225761414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6087266206741333, "epoch": 9.32, "learning_rate": 2.1020663744520977e-05, "loss": 0.5938, "step": 11031, "task_loss": 0.7062137126922607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.708909273147583, "epoch": 9.33, "learning_rate": 2.1017532874139012e-05, "loss": 0.5723, "step": 11032, "task_loss": 1.1298727989196777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4437698721885681, "epoch": 9.33, "learning_rate": 2.1014402003757044e-05, "loss": 0.5547, "step": 11033, "task_loss": 0.40451911091804504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7378613352775574, "epoch": 9.33, "learning_rate": 2.101127113337508e-05, "loss": 0.7219, "step": 11034, "task_loss": 0.8343433141708374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5952461957931519, "epoch": 9.33, "learning_rate": 2.100814026299311e-05, "loss": 0.4644, "step": 11035, "task_loss": 0.8161760568618774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5592974424362183, "epoch": 9.33, "learning_rate": 2.1005009392611146e-05, "loss": 0.5877, "step": 11036, "task_loss": 1.0932706594467163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3117920458316803, "epoch": 9.33, "learning_rate": 2.100187852222918e-05, "loss": 0.4377, "step": 11037, "task_loss": 0.29006892442703247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6203903555870056, "epoch": 9.33, "learning_rate": 2.0998747651847213e-05, "loss": 0.5023, "step": 11038, "task_loss": 1.1338294744491577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4800589084625244, "epoch": 9.33, "learning_rate": 2.0995616781465248e-05, "loss": 0.5078, "step": 11039, "task_loss": 0.641564667224884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5979856252670288, "epoch": 9.33, "learning_rate": 2.0992485911083283e-05, "loss": 0.5564, "step": 11040, "task_loss": 0.6688344478607178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35318467020988464, "epoch": 9.33, "learning_rate": 2.098935504070132e-05, "loss": 0.5852, "step": 11041, "task_loss": 0.7330173850059509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5211511850357056, "epoch": 9.33, "learning_rate": 2.098622417031935e-05, "loss": 0.5376, "step": 11042, "task_loss": 0.6433767676353455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6728024482727051, "epoch": 9.33, "learning_rate": 2.0983093299937385e-05, "loss": 0.6269, "step": 11043, "task_loss": 1.2280324697494507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4758075177669525, "epoch": 9.34, "learning_rate": 2.0979962429555417e-05, "loss": 0.5447, "step": 11044, "task_loss": 1.008735179901123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5366891026496887, "epoch": 9.34, "learning_rate": 2.0976831559173452e-05, "loss": 0.5905, "step": 11045, "task_loss": 0.6761773824691772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.506938099861145, "epoch": 9.34, "learning_rate": 2.0973700688791484e-05, "loss": 0.5271, "step": 11046, "task_loss": 0.4266812205314636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8562105298042297, "epoch": 9.34, "learning_rate": 2.097056981840952e-05, "loss": 0.6629, "step": 11047, "task_loss": 0.41501477360725403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43721240758895874, "epoch": 9.34, "learning_rate": 2.0967438948027554e-05, "loss": 0.5214, "step": 11048, "task_loss": 0.39509543776512146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3374711573123932, "epoch": 9.34, "learning_rate": 2.0964308077645586e-05, "loss": 0.4913, "step": 11049, "task_loss": 0.1561509370803833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3942636251449585, "epoch": 9.34, "learning_rate": 2.096117720726362e-05, "loss": 0.433, "step": 11050, "task_loss": 0.42468327283859253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9151959419250488, "epoch": 9.34, "learning_rate": 2.0958046336881653e-05, "loss": 0.6767, "step": 11051, "task_loss": 0.22582069039344788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46117955446243286, "epoch": 9.34, "learning_rate": 2.095491546649969e-05, "loss": 0.6329, "step": 11052, "task_loss": 0.6117879748344421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5360149145126343, "epoch": 9.34, "learning_rate": 2.095178459611772e-05, "loss": 0.5467, "step": 11053, "task_loss": 0.6518809199333191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5953665375709534, "epoch": 9.34, "learning_rate": 2.0948653725735755e-05, "loss": 0.5713, "step": 11054, "task_loss": 0.6232239007949829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7215033769607544, "epoch": 9.34, "learning_rate": 2.0945522855353787e-05, "loss": 0.5918, "step": 11055, "task_loss": 0.654564619064331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6707335710525513, "epoch": 9.35, "learning_rate": 2.0942391984971822e-05, "loss": 0.7316, "step": 11056, "task_loss": 1.3850572109222412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5717812776565552, "epoch": 9.35, "learning_rate": 2.0939261114589857e-05, "loss": 0.526, "step": 11057, "task_loss": 0.8842477202415466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4381035268306732, "epoch": 9.35, "learning_rate": 2.0936130244207893e-05, "loss": 0.4372, "step": 11058, "task_loss": 0.27524876594543457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3029240369796753, "epoch": 9.35, "learning_rate": 2.0932999373825924e-05, "loss": 0.496, "step": 11059, "task_loss": 1.4768670797348022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6320303678512573, "epoch": 9.35, "learning_rate": 2.092986850344396e-05, "loss": 0.7019, "step": 11060, "task_loss": 0.28157472610473633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7205538749694824, "epoch": 9.35, "learning_rate": 2.0926737633061995e-05, "loss": 0.5669, "step": 11061, "task_loss": 0.4825977683067322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31320565938949585, "epoch": 9.35, "learning_rate": 2.0923606762680026e-05, "loss": 0.5136, "step": 11062, "task_loss": 0.2732275724411011 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6477102041244507, "epoch": 9.35, "learning_rate": 2.092047589229806e-05, "loss": 0.5565, "step": 11063, "task_loss": 0.9247978925704956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4339419901371002, "epoch": 9.35, "learning_rate": 2.0917345021916093e-05, "loss": 0.4949, "step": 11064, "task_loss": 0.6933647990226746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3923567533493042, "epoch": 9.35, "learning_rate": 2.091421415153413e-05, "loss": 0.4984, "step": 11065, "task_loss": 0.3636775612831116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6771771907806396, "epoch": 9.35, "learning_rate": 2.091108328115216e-05, "loss": 0.6058, "step": 11066, "task_loss": 0.5175504088401794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7323749661445618, "epoch": 9.35, "learning_rate": 2.0907952410770196e-05, "loss": 0.468, "step": 11067, "task_loss": 0.9864889979362488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7646000385284424, "epoch": 9.36, "learning_rate": 2.0904821540388227e-05, "loss": 0.6323, "step": 11068, "task_loss": 0.8265000581741333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5117040872573853, "epoch": 9.36, "learning_rate": 2.0901690670006262e-05, "loss": 0.4761, "step": 11069, "task_loss": 0.5577810406684875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3715049624443054, "epoch": 9.36, "learning_rate": 2.0898559799624294e-05, "loss": 0.3479, "step": 11070, "task_loss": 0.3219952881336212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5503859519958496, "epoch": 9.36, "learning_rate": 2.089542892924233e-05, "loss": 0.5115, "step": 11071, "task_loss": 0.8405212163925171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9880281686782837, "epoch": 9.36, "learning_rate": 2.089229805886036e-05, "loss": 0.5531, "step": 11072, "task_loss": 0.9278154969215393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6758714914321899, "epoch": 9.36, "learning_rate": 2.0889167188478396e-05, "loss": 0.6636, "step": 11073, "task_loss": 0.5740950107574463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7018527984619141, "epoch": 9.36, "learning_rate": 2.088603631809643e-05, "loss": 0.6872, "step": 11074, "task_loss": 0.706409215927124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4733871817588806, "epoch": 9.36, "learning_rate": 2.0882905447714467e-05, "loss": 0.4948, "step": 11075, "task_loss": 0.3632488250732422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7951631546020508, "epoch": 9.36, "learning_rate": 2.0879774577332502e-05, "loss": 0.5792, "step": 11076, "task_loss": 0.9810490608215332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6763271689414978, "epoch": 9.36, "learning_rate": 2.0876643706950534e-05, "loss": 0.5198, "step": 11077, "task_loss": 0.47952109575271606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9819873571395874, "epoch": 9.36, "learning_rate": 2.087351283656857e-05, "loss": 0.5393, "step": 11078, "task_loss": 0.53044193983078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29400351643562317, "epoch": 9.36, "learning_rate": 2.08703819661866e-05, "loss": 0.4113, "step": 11079, "task_loss": 0.08749697357416153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7865314483642578, "epoch": 9.37, "learning_rate": 2.0867251095804636e-05, "loss": 0.6471, "step": 11080, "task_loss": 1.2800530195236206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3541516661643982, "epoch": 9.37, "learning_rate": 2.0864120225422668e-05, "loss": 0.4732, "step": 11081, "task_loss": 0.5420963168144226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5974559187889099, "epoch": 9.37, "learning_rate": 2.0860989355040703e-05, "loss": 0.7211, "step": 11082, "task_loss": 1.3383153676986694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34306657314300537, "epoch": 9.37, "learning_rate": 2.0857858484658735e-05, "loss": 0.6532, "step": 11083, "task_loss": 0.5032360553741455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0112818479537964, "epoch": 9.37, "learning_rate": 2.085472761427677e-05, "loss": 0.6956, "step": 11084, "task_loss": 1.3690439462661743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4321610927581787, "epoch": 9.37, "learning_rate": 2.0851596743894805e-05, "loss": 0.4818, "step": 11085, "task_loss": 0.9114605188369751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4049995243549347, "epoch": 9.37, "learning_rate": 2.0848465873512837e-05, "loss": 0.6429, "step": 11086, "task_loss": 0.4219440221786499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4149618148803711, "epoch": 9.37, "learning_rate": 2.0845335003130872e-05, "loss": 0.4862, "step": 11087, "task_loss": 0.6596091389656067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4334532618522644, "epoch": 9.37, "learning_rate": 2.0842204132748904e-05, "loss": 0.5101, "step": 11088, "task_loss": 0.8211643695831299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4860309362411499, "epoch": 9.37, "learning_rate": 2.083907326236694e-05, "loss": 0.6933, "step": 11089, "task_loss": 0.8992518782615662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5162341594696045, "epoch": 9.37, "learning_rate": 2.083594239198497e-05, "loss": 0.5416, "step": 11090, "task_loss": 0.7724950909614563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4974701702594757, "epoch": 9.38, "learning_rate": 2.0832811521603006e-05, "loss": 0.6465, "step": 11091, "task_loss": 1.1691416501998901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46752429008483887, "epoch": 9.38, "learning_rate": 2.082968065122104e-05, "loss": 0.6577, "step": 11092, "task_loss": 0.3868715465068817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.435374915599823, "epoch": 9.38, "learning_rate": 2.0826549780839076e-05, "loss": 0.4674, "step": 11093, "task_loss": 0.4035654366016388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4691745936870575, "epoch": 9.38, "learning_rate": 2.0823418910457108e-05, "loss": 0.4161, "step": 11094, "task_loss": 0.21432554721832275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4180178940296173, "epoch": 9.38, "learning_rate": 2.0820288040075143e-05, "loss": 0.5664, "step": 11095, "task_loss": 1.5704598426818848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32594358921051025, "epoch": 9.38, "learning_rate": 2.0817157169693175e-05, "loss": 0.4196, "step": 11096, "task_loss": 0.30951428413391113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8523607850074768, "epoch": 9.38, "learning_rate": 2.081402629931121e-05, "loss": 0.6172, "step": 11097, "task_loss": 0.35583725571632385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7431790232658386, "epoch": 9.38, "learning_rate": 2.0810895428929245e-05, "loss": 0.6748, "step": 11098, "task_loss": 0.5901286005973816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4838095009326935, "epoch": 9.38, "learning_rate": 2.0807764558547277e-05, "loss": 0.5914, "step": 11099, "task_loss": 0.6219283938407898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3776724338531494, "epoch": 9.38, "learning_rate": 2.0804633688165312e-05, "loss": 0.5027, "step": 11100, "task_loss": 0.06119833514094353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5480364561080933, "epoch": 9.38, "learning_rate": 2.0801502817783344e-05, "loss": 0.6317, "step": 11101, "task_loss": 0.8233464360237122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5016152262687683, "epoch": 9.38, "learning_rate": 2.079837194740138e-05, "loss": 0.5048, "step": 11102, "task_loss": 0.6097140312194824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3437802195549011, "epoch": 9.39, "learning_rate": 2.079524107701941e-05, "loss": 0.5907, "step": 11103, "task_loss": 0.6641770601272583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7022411823272705, "epoch": 9.39, "learning_rate": 2.0792110206637446e-05, "loss": 0.7239, "step": 11104, "task_loss": 1.261233925819397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29116353392601013, "epoch": 9.39, "learning_rate": 2.0788979336255478e-05, "loss": 0.4761, "step": 11105, "task_loss": 0.6239451766014099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7933984994888306, "epoch": 9.39, "learning_rate": 2.0785848465873513e-05, "loss": 0.6029, "step": 11106, "task_loss": 1.4060593843460083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7882918119430542, "epoch": 9.39, "learning_rate": 2.0782717595491548e-05, "loss": 0.5567, "step": 11107, "task_loss": 1.1282553672790527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6464877128601074, "epoch": 9.39, "learning_rate": 2.0779586725109583e-05, "loss": 0.6224, "step": 11108, "task_loss": 0.4683595895767212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5174087285995483, "epoch": 9.39, "learning_rate": 2.0776455854727615e-05, "loss": 0.5928, "step": 11109, "task_loss": 0.652855634689331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6765828132629395, "epoch": 9.39, "learning_rate": 2.077332498434565e-05, "loss": 0.7043, "step": 11110, "task_loss": 0.6625597476959229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.641205906867981, "epoch": 9.39, "learning_rate": 2.0770194113963685e-05, "loss": 0.5151, "step": 11111, "task_loss": 1.2532798051834106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9924364686012268, "epoch": 9.39, "learning_rate": 2.0767063243581717e-05, "loss": 0.6296, "step": 11112, "task_loss": 1.076135516166687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3466476798057556, "epoch": 9.39, "learning_rate": 2.0763932373199752e-05, "loss": 0.4321, "step": 11113, "task_loss": 0.057596608996391296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.167325496673584, "epoch": 9.39, "learning_rate": 2.0760801502817784e-05, "loss": 0.6919, "step": 11114, "task_loss": 1.9425065517425537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38657069206237793, "epoch": 9.4, "learning_rate": 2.075767063243582e-05, "loss": 0.4662, "step": 11115, "task_loss": 0.3130972683429718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4128304719924927, "epoch": 9.4, "learning_rate": 2.075453976205385e-05, "loss": 0.5073, "step": 11116, "task_loss": 0.4561305046081543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45484593510627747, "epoch": 9.4, "learning_rate": 2.0751408891671886e-05, "loss": 0.4957, "step": 11117, "task_loss": 0.5043544173240662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49775877594947815, "epoch": 9.4, "learning_rate": 2.0748278021289918e-05, "loss": 0.4081, "step": 11118, "task_loss": 0.7010271549224854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5666275024414062, "epoch": 9.4, "learning_rate": 2.0745147150907953e-05, "loss": 0.5958, "step": 11119, "task_loss": 0.48447495698928833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2698635756969452, "epoch": 9.4, "learning_rate": 2.0742016280525985e-05, "loss": 0.5203, "step": 11120, "task_loss": 0.46916186809539795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3930051922798157, "epoch": 9.4, "learning_rate": 2.073888541014402e-05, "loss": 0.4115, "step": 11121, "task_loss": 0.15671266615390778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4613761603832245, "epoch": 9.4, "learning_rate": 2.0735754539762055e-05, "loss": 0.5368, "step": 11122, "task_loss": 1.1582274436950684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5390015244483948, "epoch": 9.4, "learning_rate": 2.0732623669380087e-05, "loss": 0.4798, "step": 11123, "task_loss": 0.33388006687164307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4931383430957794, "epoch": 9.4, "learning_rate": 2.0729492798998122e-05, "loss": 0.6285, "step": 11124, "task_loss": 0.2984165847301483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4035928547382355, "epoch": 9.4, "learning_rate": 2.0726361928616157e-05, "loss": 0.6148, "step": 11125, "task_loss": 0.5724236965179443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8856648802757263, "epoch": 9.4, "learning_rate": 2.0723231058234193e-05, "loss": 0.7353, "step": 11126, "task_loss": 1.3263977766036987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45751631259918213, "epoch": 9.41, "learning_rate": 2.0720100187852224e-05, "loss": 0.4895, "step": 11127, "task_loss": 0.534850537776947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7168595790863037, "epoch": 9.41, "learning_rate": 2.071696931747026e-05, "loss": 0.6856, "step": 11128, "task_loss": 0.9173924326896667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41320526599884033, "epoch": 9.41, "learning_rate": 2.071383844708829e-05, "loss": 0.5215, "step": 11129, "task_loss": 0.08458413183689117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4893134832382202, "epoch": 9.41, "learning_rate": 2.0710707576706327e-05, "loss": 0.5068, "step": 11130, "task_loss": 0.5590797066688538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5925166010856628, "epoch": 9.41, "learning_rate": 2.0707576706324358e-05, "loss": 0.6259, "step": 11131, "task_loss": 1.722000241279602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6151952743530273, "epoch": 9.41, "learning_rate": 2.0704445835942393e-05, "loss": 0.6433, "step": 11132, "task_loss": 1.0767526626586914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4521305561065674, "epoch": 9.41, "learning_rate": 2.0701314965560425e-05, "loss": 0.3665, "step": 11133, "task_loss": 0.3849320411682129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5420761108398438, "epoch": 9.41, "learning_rate": 2.069818409517846e-05, "loss": 0.5009, "step": 11134, "task_loss": 0.4980718791484833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39783182740211487, "epoch": 9.41, "learning_rate": 2.0695053224796496e-05, "loss": 0.498, "step": 11135, "task_loss": 0.8538029789924622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19452226161956787, "epoch": 9.41, "learning_rate": 2.0691922354414527e-05, "loss": 0.5402, "step": 11136, "task_loss": 0.9968996644020081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3843739628791809, "epoch": 9.41, "learning_rate": 2.0688791484032563e-05, "loss": 0.4961, "step": 11137, "task_loss": 0.4816380441188812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5740887522697449, "epoch": 9.41, "learning_rate": 2.0685660613650594e-05, "loss": 0.4517, "step": 11138, "task_loss": 0.779881477355957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4797784984111786, "epoch": 9.42, "learning_rate": 2.068252974326863e-05, "loss": 0.5475, "step": 11139, "task_loss": 0.4979042708873749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6636371612548828, "epoch": 9.42, "learning_rate": 2.067939887288666e-05, "loss": 0.6405, "step": 11140, "task_loss": 1.0886609554290771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33401864767074585, "epoch": 9.42, "learning_rate": 2.0676268002504696e-05, "loss": 0.5531, "step": 11141, "task_loss": 0.5187253952026367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49431905150413513, "epoch": 9.42, "learning_rate": 2.067313713212273e-05, "loss": 0.6524, "step": 11142, "task_loss": 0.41980189085006714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40769973397254944, "epoch": 9.42, "learning_rate": 2.0670006261740767e-05, "loss": 0.6014, "step": 11143, "task_loss": 0.2820325791835785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28555408120155334, "epoch": 9.42, "learning_rate": 2.06668753913588e-05, "loss": 0.5948, "step": 11144, "task_loss": 0.6144871711730957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6515889167785645, "epoch": 9.42, "learning_rate": 2.0663744520976834e-05, "loss": 0.5624, "step": 11145, "task_loss": 0.6896404027938843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5608714818954468, "epoch": 9.42, "learning_rate": 2.0660613650594866e-05, "loss": 0.7595, "step": 11146, "task_loss": 0.9495664834976196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7966141700744629, "epoch": 9.42, "learning_rate": 2.06574827802129e-05, "loss": 0.5496, "step": 11147, "task_loss": 0.6745889782905579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8736964464187622, "epoch": 9.42, "learning_rate": 2.0654351909830936e-05, "loss": 0.632, "step": 11148, "task_loss": 1.0100066661834717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48804205656051636, "epoch": 9.42, "learning_rate": 2.0651221039448968e-05, "loss": 0.547, "step": 11149, "task_loss": 0.8020251989364624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7795400619506836, "epoch": 9.42, "learning_rate": 2.0648090169067003e-05, "loss": 0.7048, "step": 11150, "task_loss": 1.5756973028182983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2908931374549866, "epoch": 9.43, "learning_rate": 2.0644959298685035e-05, "loss": 0.4951, "step": 11151, "task_loss": 0.741003692150116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46741753816604614, "epoch": 9.43, "learning_rate": 2.064182842830307e-05, "loss": 0.6507, "step": 11152, "task_loss": 0.8526996374130249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6701887249946594, "epoch": 9.43, "learning_rate": 2.06386975579211e-05, "loss": 0.646, "step": 11153, "task_loss": 0.35950085520744324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6020390391349792, "epoch": 9.43, "learning_rate": 2.0635566687539137e-05, "loss": 0.6058, "step": 11154, "task_loss": 0.6441352367401123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30649513006210327, "epoch": 9.43, "learning_rate": 2.063243581715717e-05, "loss": 0.3837, "step": 11155, "task_loss": 0.30779770016670227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5969529747962952, "epoch": 9.43, "learning_rate": 2.0629304946775204e-05, "loss": 0.5962, "step": 11156, "task_loss": 0.34430497884750366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5877125263214111, "epoch": 9.43, "learning_rate": 2.0626174076393235e-05, "loss": 0.6276, "step": 11157, "task_loss": 1.0099871158599854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8669157028198242, "epoch": 9.43, "learning_rate": 2.062304320601127e-05, "loss": 0.6058, "step": 11158, "task_loss": 0.815217137336731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5813777446746826, "epoch": 9.43, "learning_rate": 2.0619912335629306e-05, "loss": 0.6693, "step": 11159, "task_loss": 0.8008521199226379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4250643253326416, "epoch": 9.43, "learning_rate": 2.061678146524734e-05, "loss": 0.4787, "step": 11160, "task_loss": 0.22783493995666504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5116851329803467, "epoch": 9.43, "learning_rate": 2.0613650594865376e-05, "loss": 0.4807, "step": 11161, "task_loss": 0.6268780827522278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5664005279541016, "epoch": 9.44, "learning_rate": 2.0610519724483408e-05, "loss": 0.5501, "step": 11162, "task_loss": 0.6575700044631958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8029319643974304, "epoch": 9.44, "learning_rate": 2.0607388854101443e-05, "loss": 0.4959, "step": 11163, "task_loss": 1.3831977844238281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2780126929283142, "epoch": 9.44, "learning_rate": 2.0604257983719475e-05, "loss": 0.5843, "step": 11164, "task_loss": 0.7233217358589172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47447964549064636, "epoch": 9.44, "learning_rate": 2.060112711333751e-05, "loss": 0.5014, "step": 11165, "task_loss": 0.808947741985321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35247746109962463, "epoch": 9.44, "learning_rate": 2.0597996242955542e-05, "loss": 0.5165, "step": 11166, "task_loss": 0.856187105178833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4989609122276306, "epoch": 9.44, "learning_rate": 2.0594865372573577e-05, "loss": 0.3989, "step": 11167, "task_loss": 0.2812872529029846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6363271474838257, "epoch": 9.44, "learning_rate": 2.059173450219161e-05, "loss": 0.5384, "step": 11168, "task_loss": 0.40860167145729065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3949642777442932, "epoch": 9.44, "learning_rate": 2.0588603631809644e-05, "loss": 0.5155, "step": 11169, "task_loss": 1.3494837284088135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6772137880325317, "epoch": 9.44, "learning_rate": 2.0585472761427676e-05, "loss": 0.6666, "step": 11170, "task_loss": 0.5143532752990723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41721010208129883, "epoch": 9.44, "learning_rate": 2.058234189104571e-05, "loss": 0.461, "step": 11171, "task_loss": 0.9379950165748596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28648096323013306, "epoch": 9.44, "learning_rate": 2.0579211020663746e-05, "loss": 0.4923, "step": 11172, "task_loss": 0.4109518527984619 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.087327003479004, "epoch": 9.44, "learning_rate": 2.0576080150281778e-05, "loss": 0.6823, "step": 11173, "task_loss": 1.3821860551834106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3695640563964844, "epoch": 9.45, "learning_rate": 2.0572949279899813e-05, "loss": 0.6587, "step": 11174, "task_loss": 1.2057639360427856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36415567994117737, "epoch": 9.45, "learning_rate": 2.0569818409517845e-05, "loss": 0.3601, "step": 11175, "task_loss": 0.840754508972168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5446778535842896, "epoch": 9.45, "learning_rate": 2.056668753913588e-05, "loss": 0.7173, "step": 11176, "task_loss": 1.0040769577026367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48376786708831787, "epoch": 9.45, "learning_rate": 2.0563556668753915e-05, "loss": 0.3964, "step": 11177, "task_loss": 0.14819926023483276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3539237380027771, "epoch": 9.45, "learning_rate": 2.056042579837195e-05, "loss": 0.5168, "step": 11178, "task_loss": 0.5203588604927063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6554898619651794, "epoch": 9.45, "learning_rate": 2.0557294927989982e-05, "loss": 0.6501, "step": 11179, "task_loss": 0.4861893355846405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7293578386306763, "epoch": 9.45, "learning_rate": 2.0554164057608017e-05, "loss": 0.7038, "step": 11180, "task_loss": 0.2777087092399597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30142372846603394, "epoch": 9.45, "learning_rate": 2.055103318722605e-05, "loss": 0.4636, "step": 11181, "task_loss": 0.8148493766784668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5827466249465942, "epoch": 9.45, "learning_rate": 2.0547902316844084e-05, "loss": 0.6779, "step": 11182, "task_loss": 1.4460619688034058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34403252601623535, "epoch": 9.45, "learning_rate": 2.054477144646212e-05, "loss": 0.4579, "step": 11183, "task_loss": 0.43237173557281494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4394824206829071, "epoch": 9.45, "learning_rate": 2.054164057608015e-05, "loss": 0.6069, "step": 11184, "task_loss": 0.5421826839447021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44951555132865906, "epoch": 9.45, "learning_rate": 2.0538509705698186e-05, "loss": 0.5955, "step": 11185, "task_loss": 0.6515569090843201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2653992772102356, "epoch": 9.46, "learning_rate": 2.0535378835316218e-05, "loss": 0.4407, "step": 11186, "task_loss": 0.19472892582416534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5540260672569275, "epoch": 9.46, "learning_rate": 2.0532247964934253e-05, "loss": 0.5034, "step": 11187, "task_loss": 0.7031592726707458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4597819149494171, "epoch": 9.46, "learning_rate": 2.0529117094552285e-05, "loss": 0.5303, "step": 11188, "task_loss": 0.7496867179870605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5061589479446411, "epoch": 9.46, "learning_rate": 2.052598622417032e-05, "loss": 0.6001, "step": 11189, "task_loss": 1.4593271017074585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4713425636291504, "epoch": 9.46, "learning_rate": 2.0522855353788352e-05, "loss": 0.3868, "step": 11190, "task_loss": 0.8462479710578918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4412979781627655, "epoch": 9.46, "learning_rate": 2.0519724483406387e-05, "loss": 0.4891, "step": 11191, "task_loss": 0.3657672107219696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45641687512397766, "epoch": 9.46, "learning_rate": 2.0516593613024422e-05, "loss": 0.55, "step": 11192, "task_loss": 0.6818670034408569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47182220220565796, "epoch": 9.46, "learning_rate": 2.0513462742642457e-05, "loss": 0.5442, "step": 11193, "task_loss": 0.6109762787818909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4033227562904358, "epoch": 9.46, "learning_rate": 2.051033187226049e-05, "loss": 0.4849, "step": 11194, "task_loss": 1.9705026149749756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5045433640480042, "epoch": 9.46, "learning_rate": 2.0507201001878524e-05, "loss": 0.4905, "step": 11195, "task_loss": 1.2357059717178345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43137305974960327, "epoch": 9.46, "learning_rate": 2.050407013149656e-05, "loss": 0.506, "step": 11196, "task_loss": 0.5525730848312378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.425146222114563, "epoch": 9.46, "learning_rate": 2.050093926111459e-05, "loss": 0.479, "step": 11197, "task_loss": 0.8354417085647583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4417227506637573, "epoch": 9.47, "learning_rate": 2.0497808390732627e-05, "loss": 0.5521, "step": 11198, "task_loss": 0.575149416923523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6016185283660889, "epoch": 9.47, "learning_rate": 2.049467752035066e-05, "loss": 0.5683, "step": 11199, "task_loss": 0.8019603490829468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37276986241340637, "epoch": 9.47, "learning_rate": 2.0491546649968694e-05, "loss": 0.5705, "step": 11200, "task_loss": 0.6273150444030762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2565717697143555, "epoch": 9.47, "learning_rate": 2.0488415779586725e-05, "loss": 0.6476, "step": 11201, "task_loss": 0.8085762858390808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4125361740589142, "epoch": 9.47, "learning_rate": 2.048528490920476e-05, "loss": 0.5315, "step": 11202, "task_loss": 0.03842768445611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.522205114364624, "epoch": 9.47, "learning_rate": 2.0482154038822792e-05, "loss": 0.5319, "step": 11203, "task_loss": 1.3147358894348145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0645502805709839, "epoch": 9.47, "learning_rate": 2.0479023168440827e-05, "loss": 0.6899, "step": 11204, "task_loss": 0.9629391431808472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23412391543388367, "epoch": 9.47, "learning_rate": 2.047589229805886e-05, "loss": 0.5156, "step": 11205, "task_loss": 0.041542328894138336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47442683577537537, "epoch": 9.47, "learning_rate": 2.0472761427676894e-05, "loss": 0.5059, "step": 11206, "task_loss": 0.17096945643424988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48154622316360474, "epoch": 9.47, "learning_rate": 2.0469630557294926e-05, "loss": 0.4433, "step": 11207, "task_loss": 0.2676759958267212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5564737319946289, "epoch": 9.47, "learning_rate": 2.046649968691296e-05, "loss": 0.6553, "step": 11208, "task_loss": 1.0841857194900513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4644075930118561, "epoch": 9.47, "learning_rate": 2.0463368816530996e-05, "loss": 0.5635, "step": 11209, "task_loss": 1.195984959602356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9521697163581848, "epoch": 9.48, "learning_rate": 2.046023794614903e-05, "loss": 0.6213, "step": 11210, "task_loss": 1.0427703857421875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5253634452819824, "epoch": 9.48, "learning_rate": 2.0457107075767067e-05, "loss": 0.7321, "step": 11211, "task_loss": 0.3801705837249756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7676039338111877, "epoch": 9.48, "learning_rate": 2.04539762053851e-05, "loss": 0.6139, "step": 11212, "task_loss": 0.46647512912750244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7973577380180359, "epoch": 9.48, "learning_rate": 2.0450845335003134e-05, "loss": 0.5921, "step": 11213, "task_loss": 0.5874646902084351 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39039987325668335, "epoch": 9.48, "learning_rate": 2.0447714464621166e-05, "loss": 0.4306, "step": 11214, "task_loss": 0.6466406583786011 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35016417503356934, "epoch": 9.48, "learning_rate": 2.04445835942392e-05, "loss": 0.4811, "step": 11215, "task_loss": 0.2757631540298462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7128924131393433, "epoch": 9.48, "learning_rate": 2.0441452723857232e-05, "loss": 0.5388, "step": 11216, "task_loss": 1.1916460990905762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4340100884437561, "epoch": 9.48, "learning_rate": 2.0438321853475268e-05, "loss": 0.6452, "step": 11217, "task_loss": 0.5195159912109375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30924955010414124, "epoch": 9.48, "learning_rate": 2.04351909830933e-05, "loss": 0.4619, "step": 11218, "task_loss": 0.22463753819465637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48033684492111206, "epoch": 9.48, "learning_rate": 2.0432060112711335e-05, "loss": 0.5002, "step": 11219, "task_loss": 0.8660057783126831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4784981310367584, "epoch": 9.48, "learning_rate": 2.042892924232937e-05, "loss": 0.4923, "step": 11220, "task_loss": 0.43277043104171753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5962539315223694, "epoch": 9.48, "learning_rate": 2.04257983719474e-05, "loss": 0.6437, "step": 11221, "task_loss": 1.3364439010620117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42677050828933716, "epoch": 9.49, "learning_rate": 2.0422667501565437e-05, "loss": 0.462, "step": 11222, "task_loss": 0.44228678941726685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5165603160858154, "epoch": 9.49, "learning_rate": 2.041953663118347e-05, "loss": 0.5183, "step": 11223, "task_loss": 0.8416460752487183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8686898350715637, "epoch": 9.49, "learning_rate": 2.0416405760801504e-05, "loss": 0.6762, "step": 11224, "task_loss": 1.426055908203125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5603064298629761, "epoch": 9.49, "learning_rate": 2.0413274890419535e-05, "loss": 0.6242, "step": 11225, "task_loss": 0.9819295406341553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8008278012275696, "epoch": 9.49, "learning_rate": 2.041014402003757e-05, "loss": 0.5675, "step": 11226, "task_loss": 0.7535830736160278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6229571104049683, "epoch": 9.49, "learning_rate": 2.0407013149655606e-05, "loss": 0.5382, "step": 11227, "task_loss": 1.226574182510376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5085885524749756, "epoch": 9.49, "learning_rate": 2.040388227927364e-05, "loss": 0.5595, "step": 11228, "task_loss": 0.08829927444458008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.909530758857727, "epoch": 9.49, "learning_rate": 2.0400751408891673e-05, "loss": 0.6246, "step": 11229, "task_loss": 0.6141870617866516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46837905049324036, "epoch": 9.49, "learning_rate": 2.0397620538509708e-05, "loss": 0.5745, "step": 11230, "task_loss": 0.4805211126804352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4878770709037781, "epoch": 9.49, "learning_rate": 2.039448966812774e-05, "loss": 0.5486, "step": 11231, "task_loss": 0.435899555683136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.591152548789978, "epoch": 9.49, "learning_rate": 2.0391358797745775e-05, "loss": 0.4804, "step": 11232, "task_loss": 0.9772195219993591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8437149524688721, "epoch": 9.5, "learning_rate": 2.038822792736381e-05, "loss": 0.5483, "step": 11233, "task_loss": 1.0704753398895264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3982442021369934, "epoch": 9.5, "learning_rate": 2.0385097056981842e-05, "loss": 0.5742, "step": 11234, "task_loss": 0.342947393655777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4527246356010437, "epoch": 9.5, "learning_rate": 2.0381966186599877e-05, "loss": 0.3943, "step": 11235, "task_loss": 0.7738919258117676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5183579921722412, "epoch": 9.5, "learning_rate": 2.037883531621791e-05, "loss": 0.4989, "step": 11236, "task_loss": 0.6759052872657776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47781139612197876, "epoch": 9.5, "learning_rate": 2.0375704445835944e-05, "loss": 0.5244, "step": 11237, "task_loss": 0.3410857319831848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5395419597625732, "epoch": 9.5, "learning_rate": 2.0372573575453976e-05, "loss": 0.4331, "step": 11238, "task_loss": 1.0231642723083496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6995275020599365, "epoch": 9.5, "learning_rate": 2.036944270507201e-05, "loss": 0.4635, "step": 11239, "task_loss": 1.849626898765564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7694113254547119, "epoch": 9.5, "learning_rate": 2.0366311834690043e-05, "loss": 0.7127, "step": 11240, "task_loss": 0.4858366549015045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5516558289527893, "epoch": 9.5, "learning_rate": 2.0363180964308078e-05, "loss": 0.5298, "step": 11241, "task_loss": 0.8462251424789429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4129839837551117, "epoch": 9.5, "learning_rate": 2.036005009392611e-05, "loss": 0.5402, "step": 11242, "task_loss": 0.9844934940338135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5057719349861145, "epoch": 9.5, "learning_rate": 2.0356919223544145e-05, "loss": 0.5255, "step": 11243, "task_loss": 1.0806610584259033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5012354254722595, "epoch": 9.5, "learning_rate": 2.035378835316218e-05, "loss": 0.5816, "step": 11244, "task_loss": 0.4043217599391937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.485105961561203, "epoch": 9.51, "learning_rate": 2.0350657482780215e-05, "loss": 0.5614, "step": 11245, "task_loss": 0.8010050654411316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26766180992126465, "epoch": 9.51, "learning_rate": 2.034752661239825e-05, "loss": 0.426, "step": 11246, "task_loss": 0.21514928340911865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4594436585903168, "epoch": 9.51, "learning_rate": 2.0344395742016282e-05, "loss": 0.5699, "step": 11247, "task_loss": 0.6386981010437012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.669175922870636, "epoch": 9.51, "learning_rate": 2.0341264871634317e-05, "loss": 0.5406, "step": 11248, "task_loss": 0.6877296566963196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6813497543334961, "epoch": 9.51, "learning_rate": 2.033813400125235e-05, "loss": 0.5729, "step": 11249, "task_loss": 0.3061593174934387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.643942654132843, "epoch": 9.51, "learning_rate": 2.0335003130870384e-05, "loss": 0.5507, "step": 11250, "task_loss": 0.8842963576316833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5125287175178528, "epoch": 9.51, "learning_rate": 2.0331872260488416e-05, "loss": 0.7541, "step": 11251, "task_loss": 0.431420236825943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8309155106544495, "epoch": 9.51, "learning_rate": 2.032874139010645e-05, "loss": 0.516, "step": 11252, "task_loss": 0.5819403529167175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5720171332359314, "epoch": 9.51, "learning_rate": 2.0325610519724483e-05, "loss": 0.5099, "step": 11253, "task_loss": 1.0121551752090454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9654394388198853, "epoch": 9.51, "learning_rate": 2.0322479649342518e-05, "loss": 0.8474, "step": 11254, "task_loss": 1.7843842506408691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5626282095909119, "epoch": 9.51, "learning_rate": 2.031934877896055e-05, "loss": 0.5058, "step": 11255, "task_loss": 0.3338312804698944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4526161253452301, "epoch": 9.51, "learning_rate": 2.0316217908578585e-05, "loss": 0.5002, "step": 11256, "task_loss": 0.589309573173523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6108368635177612, "epoch": 9.52, "learning_rate": 2.031308703819662e-05, "loss": 0.6554, "step": 11257, "task_loss": 0.5494696497917175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3843992352485657, "epoch": 9.52, "learning_rate": 2.0309956167814652e-05, "loss": 0.5586, "step": 11258, "task_loss": 0.33973264694213867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3833247721195221, "epoch": 9.52, "learning_rate": 2.0306825297432687e-05, "loss": 0.4404, "step": 11259, "task_loss": 0.3187718689441681 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33413970470428467, "epoch": 9.52, "learning_rate": 2.0303694427050722e-05, "loss": 0.5028, "step": 11260, "task_loss": 0.45236736536026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43233710527420044, "epoch": 9.52, "learning_rate": 2.0300563556668754e-05, "loss": 0.5918, "step": 11261, "task_loss": 0.4479560852050781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.570199728012085, "epoch": 9.52, "learning_rate": 2.029743268628679e-05, "loss": 0.5973, "step": 11262, "task_loss": 0.44882750511169434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1317425966262817, "epoch": 9.52, "learning_rate": 2.0294301815904824e-05, "loss": 0.6899, "step": 11263, "task_loss": 0.5014103055000305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5074343085289001, "epoch": 9.52, "learning_rate": 2.0291170945522856e-05, "loss": 0.4688, "step": 11264, "task_loss": 0.7278651595115662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3263566493988037, "epoch": 9.52, "learning_rate": 2.028804007514089e-05, "loss": 0.4407, "step": 11265, "task_loss": 0.23368915915489197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5953338742256165, "epoch": 9.52, "learning_rate": 2.0284909204758923e-05, "loss": 0.565, "step": 11266, "task_loss": 0.62763911485672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9048928022384644, "epoch": 9.52, "learning_rate": 2.028177833437696e-05, "loss": 0.5995, "step": 11267, "task_loss": 1.269993782043457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7121292352676392, "epoch": 9.52, "learning_rate": 2.027864746399499e-05, "loss": 0.5877, "step": 11268, "task_loss": 0.5164435505867004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7803394198417664, "epoch": 9.53, "learning_rate": 2.0275516593613025e-05, "loss": 0.5333, "step": 11269, "task_loss": 0.8488005995750427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3182833790779114, "epoch": 9.53, "learning_rate": 2.027238572323106e-05, "loss": 0.5553, "step": 11270, "task_loss": 0.43040159344673157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3431887626647949, "epoch": 9.53, "learning_rate": 2.0269254852849092e-05, "loss": 0.5505, "step": 11271, "task_loss": 1.0213106870651245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2924032509326935, "epoch": 9.53, "learning_rate": 2.0266123982467127e-05, "loss": 0.4471, "step": 11272, "task_loss": 0.11700939387083054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5719513893127441, "epoch": 9.53, "learning_rate": 2.026299311208516e-05, "loss": 0.5696, "step": 11273, "task_loss": 0.919419527053833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8549978733062744, "epoch": 9.53, "learning_rate": 2.0259862241703194e-05, "loss": 0.4942, "step": 11274, "task_loss": 0.6792372465133667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40721091628074646, "epoch": 9.53, "learning_rate": 2.0256731371321226e-05, "loss": 0.6359, "step": 11275, "task_loss": 0.41607290506362915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.509833037853241, "epoch": 9.53, "learning_rate": 2.025360050093926e-05, "loss": 0.5002, "step": 11276, "task_loss": 0.29974329471588135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47756558656692505, "epoch": 9.53, "learning_rate": 2.0250469630557297e-05, "loss": 0.597, "step": 11277, "task_loss": 0.13461117446422577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5137880444526672, "epoch": 9.53, "learning_rate": 2.024733876017533e-05, "loss": 0.3495, "step": 11278, "task_loss": 0.7981230020523071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3073747158050537, "epoch": 9.53, "learning_rate": 2.0244207889793363e-05, "loss": 0.4068, "step": 11279, "task_loss": 1.2586581707000732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3990315794944763, "epoch": 9.53, "learning_rate": 2.02410770194114e-05, "loss": 0.3591, "step": 11280, "task_loss": 0.4683440923690796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6902435421943665, "epoch": 9.54, "learning_rate": 2.0237946149029434e-05, "loss": 0.4756, "step": 11281, "task_loss": 2.1279184818267822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5766195058822632, "epoch": 9.54, "learning_rate": 2.0234815278647466e-05, "loss": 0.6811, "step": 11282, "task_loss": 0.36171600222587585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6444180607795715, "epoch": 9.54, "learning_rate": 2.02316844082655e-05, "loss": 0.5424, "step": 11283, "task_loss": 1.2326406240463257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5545305013656616, "epoch": 9.54, "learning_rate": 2.0228553537883533e-05, "loss": 0.5175, "step": 11284, "task_loss": 0.4413456916809082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4815494418144226, "epoch": 9.54, "learning_rate": 2.0225422667501568e-05, "loss": 0.5348, "step": 11285, "task_loss": 1.0468723773956299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39721786975860596, "epoch": 9.54, "learning_rate": 2.02222917971196e-05, "loss": 0.4798, "step": 11286, "task_loss": 0.7300109267234802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3317987322807312, "epoch": 9.54, "learning_rate": 2.0219160926737635e-05, "loss": 0.4862, "step": 11287, "task_loss": 0.3727075457572937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7487756609916687, "epoch": 9.54, "learning_rate": 2.0216030056355666e-05, "loss": 0.5735, "step": 11288, "task_loss": 1.5934579372406006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4770052433013916, "epoch": 9.54, "learning_rate": 2.02128991859737e-05, "loss": 0.5302, "step": 11289, "task_loss": 0.8193758726119995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8135160207748413, "epoch": 9.54, "learning_rate": 2.0209768315591733e-05, "loss": 0.6125, "step": 11290, "task_loss": 1.385048508644104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4124804735183716, "epoch": 9.54, "learning_rate": 2.020663744520977e-05, "loss": 0.5294, "step": 11291, "task_loss": 0.5508372783660889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45812928676605225, "epoch": 9.54, "learning_rate": 2.02035065748278e-05, "loss": 0.5024, "step": 11292, "task_loss": 1.4612799882888794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7161959409713745, "epoch": 9.55, "learning_rate": 2.0200375704445836e-05, "loss": 0.7186, "step": 11293, "task_loss": 0.593767523765564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5923593640327454, "epoch": 9.55, "learning_rate": 2.019724483406387e-05, "loss": 0.5744, "step": 11294, "task_loss": 0.5305915474891663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4596346616744995, "epoch": 9.55, "learning_rate": 2.0194113963681906e-05, "loss": 0.5155, "step": 11295, "task_loss": 0.030906230211257935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9170129299163818, "epoch": 9.55, "learning_rate": 2.019098309329994e-05, "loss": 0.6267, "step": 11296, "task_loss": 0.7111082673072815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27494263648986816, "epoch": 9.55, "learning_rate": 2.0187852222917973e-05, "loss": 0.4485, "step": 11297, "task_loss": 0.7393298149108887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5825574398040771, "epoch": 9.55, "learning_rate": 2.0184721352536008e-05, "loss": 0.6271, "step": 11298, "task_loss": 0.948905348777771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.550849437713623, "epoch": 9.55, "learning_rate": 2.018159048215404e-05, "loss": 0.5161, "step": 11299, "task_loss": 1.1391692161560059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4766658544540405, "epoch": 9.55, "learning_rate": 2.0178459611772075e-05, "loss": 0.5951, "step": 11300, "task_loss": 0.5447161197662354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5767871141433716, "epoch": 9.55, "learning_rate": 2.0175328741390107e-05, "loss": 0.704, "step": 11301, "task_loss": 1.2369147539138794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29297247529029846, "epoch": 9.55, "learning_rate": 2.0172197871008142e-05, "loss": 0.5667, "step": 11302, "task_loss": 0.4745354950428009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.668285608291626, "epoch": 9.55, "learning_rate": 2.0169067000626174e-05, "loss": 0.4708, "step": 11303, "task_loss": 0.5044649839401245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.611952543258667, "epoch": 9.56, "learning_rate": 2.016593613024421e-05, "loss": 0.6633, "step": 11304, "task_loss": 0.3492443859577179 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4914405345916748, "epoch": 9.56, "learning_rate": 2.016280525986224e-05, "loss": 0.4708, "step": 11305, "task_loss": 0.5976226329803467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4179247319698334, "epoch": 9.56, "learning_rate": 2.0159674389480276e-05, "loss": 0.6021, "step": 11306, "task_loss": 0.6358379125595093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3733099699020386, "epoch": 9.56, "learning_rate": 2.015654351909831e-05, "loss": 0.4851, "step": 11307, "task_loss": 0.44281184673309326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3228073716163635, "epoch": 9.56, "learning_rate": 2.0153412648716343e-05, "loss": 0.5902, "step": 11308, "task_loss": 0.6290004849433899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8264307975769043, "epoch": 9.56, "learning_rate": 2.0150281778334378e-05, "loss": 0.647, "step": 11309, "task_loss": 0.6387498378753662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2290351688861847, "epoch": 9.56, "learning_rate": 2.014715090795241e-05, "loss": 0.4258, "step": 11310, "task_loss": 0.2221429944038391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26905184984207153, "epoch": 9.56, "learning_rate": 2.0144020037570445e-05, "loss": 0.5779, "step": 11311, "task_loss": 0.562477707862854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37551283836364746, "epoch": 9.56, "learning_rate": 2.014088916718848e-05, "loss": 0.4678, "step": 11312, "task_loss": 1.0091471672058105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45429477095603943, "epoch": 9.56, "learning_rate": 2.0137758296806515e-05, "loss": 0.5257, "step": 11313, "task_loss": 0.5815458297729492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.521268367767334, "epoch": 9.56, "learning_rate": 2.0134627426424547e-05, "loss": 0.5847, "step": 11314, "task_loss": 0.88334059715271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3778032660484314, "epoch": 9.56, "learning_rate": 2.0131496556042582e-05, "loss": 0.4916, "step": 11315, "task_loss": 0.8135557174682617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6381540298461914, "epoch": 9.57, "learning_rate": 2.0128365685660614e-05, "loss": 0.5555, "step": 11316, "task_loss": 0.7313001751899719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8293584585189819, "epoch": 9.57, "learning_rate": 2.012523481527865e-05, "loss": 0.5821, "step": 11317, "task_loss": 0.6817441582679749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36975759267807007, "epoch": 9.57, "learning_rate": 2.0122103944896684e-05, "loss": 0.4044, "step": 11318, "task_loss": 0.776977002620697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7780752182006836, "epoch": 9.57, "learning_rate": 2.0118973074514716e-05, "loss": 0.5407, "step": 11319, "task_loss": 0.5055119395256042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4348961412906647, "epoch": 9.57, "learning_rate": 2.011584220413275e-05, "loss": 0.5218, "step": 11320, "task_loss": 0.6405312418937683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29853588342666626, "epoch": 9.57, "learning_rate": 2.0112711333750783e-05, "loss": 0.4847, "step": 11321, "task_loss": 0.25549012422561646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6479320526123047, "epoch": 9.57, "learning_rate": 2.0109580463368818e-05, "loss": 0.5004, "step": 11322, "task_loss": 0.2166755497455597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6850980520248413, "epoch": 9.57, "learning_rate": 2.010644959298685e-05, "loss": 0.4998, "step": 11323, "task_loss": 0.5309674739837646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4504559636116028, "epoch": 9.57, "learning_rate": 2.0103318722604885e-05, "loss": 0.698, "step": 11324, "task_loss": 0.7981002330780029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5005103349685669, "epoch": 9.57, "learning_rate": 2.0100187852222917e-05, "loss": 0.4912, "step": 11325, "task_loss": 0.5281919836997986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6549570560455322, "epoch": 9.57, "learning_rate": 2.0097056981840952e-05, "loss": 0.5737, "step": 11326, "task_loss": 1.2755305767059326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4173576533794403, "epoch": 9.57, "learning_rate": 2.0093926111458984e-05, "loss": 0.5211, "step": 11327, "task_loss": 0.059527430683374405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4661264717578888, "epoch": 9.58, "learning_rate": 2.009079524107702e-05, "loss": 0.5053, "step": 11328, "task_loss": 0.6722946763038635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5963429808616638, "epoch": 9.58, "learning_rate": 2.0087664370695054e-05, "loss": 0.509, "step": 11329, "task_loss": 0.2937360405921936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46133750677108765, "epoch": 9.58, "learning_rate": 2.008453350031309e-05, "loss": 0.4829, "step": 11330, "task_loss": 0.6214385032653809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6129201054573059, "epoch": 9.58, "learning_rate": 2.0081402629931125e-05, "loss": 0.4936, "step": 11331, "task_loss": 0.690991222858429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46535080671310425, "epoch": 9.58, "learning_rate": 2.0078271759549156e-05, "loss": 0.503, "step": 11332, "task_loss": 0.3150765895843506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5911092758178711, "epoch": 9.58, "learning_rate": 2.007514088916719e-05, "loss": 0.5326, "step": 11333, "task_loss": 0.9012985229492188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.299363911151886, "epoch": 9.58, "learning_rate": 2.0072010018785223e-05, "loss": 0.3726, "step": 11334, "task_loss": 0.10905496031045914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6393782496452332, "epoch": 9.58, "learning_rate": 2.006887914840326e-05, "loss": 0.4518, "step": 11335, "task_loss": 0.490482896566391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45151785016059875, "epoch": 9.58, "learning_rate": 2.006574827802129e-05, "loss": 0.5218, "step": 11336, "task_loss": 0.45781105756759644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3321381211280823, "epoch": 9.58, "learning_rate": 2.0062617407639325e-05, "loss": 0.4988, "step": 11337, "task_loss": 0.034162819385528564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5909457206726074, "epoch": 9.58, "learning_rate": 2.0059486537257357e-05, "loss": 0.7291, "step": 11338, "task_loss": 1.3133918046951294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30439698696136475, "epoch": 9.58, "learning_rate": 2.0056355666875392e-05, "loss": 0.4496, "step": 11339, "task_loss": 0.614002525806427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8164217472076416, "epoch": 9.59, "learning_rate": 2.0053224796493424e-05, "loss": 0.602, "step": 11340, "task_loss": 0.20148888230323792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4913516938686371, "epoch": 9.59, "learning_rate": 2.005009392611146e-05, "loss": 0.6404, "step": 11341, "task_loss": 1.1758928298950195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5877797603607178, "epoch": 9.59, "learning_rate": 2.004696305572949e-05, "loss": 0.6737, "step": 11342, "task_loss": 1.0060975551605225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29597675800323486, "epoch": 9.59, "learning_rate": 2.0043832185347526e-05, "loss": 0.4731, "step": 11343, "task_loss": 0.46094024181365967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5970086455345154, "epoch": 9.59, "learning_rate": 2.004070131496556e-05, "loss": 0.5276, "step": 11344, "task_loss": 1.1381738185882568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5001522302627563, "epoch": 9.59, "learning_rate": 2.0037570444583597e-05, "loss": 0.6534, "step": 11345, "task_loss": 0.6212742328643799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8302507996559143, "epoch": 9.59, "learning_rate": 2.003443957420163e-05, "loss": 0.7201, "step": 11346, "task_loss": 1.1527330875396729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37067896127700806, "epoch": 9.59, "learning_rate": 2.0031308703819663e-05, "loss": 0.4902, "step": 11347, "task_loss": 0.32233926653862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5700045824050903, "epoch": 9.59, "learning_rate": 2.00281778334377e-05, "loss": 0.431, "step": 11348, "task_loss": 1.066907525062561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6190651059150696, "epoch": 9.59, "learning_rate": 2.002504696305573e-05, "loss": 0.6122, "step": 11349, "task_loss": 0.6019753813743591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6600720882415771, "epoch": 9.59, "learning_rate": 2.0021916092673766e-05, "loss": 0.4786, "step": 11350, "task_loss": 1.2669169902801514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3609832227230072, "epoch": 9.59, "learning_rate": 2.0018785222291797e-05, "loss": 0.5739, "step": 11351, "task_loss": 0.3393155634403229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6657270789146423, "epoch": 9.6, "learning_rate": 2.0015654351909833e-05, "loss": 0.5574, "step": 11352, "task_loss": 0.428869366645813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5954182147979736, "epoch": 9.6, "learning_rate": 2.0012523481527864e-05, "loss": 0.6477, "step": 11353, "task_loss": 0.8581010103225708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5802528858184814, "epoch": 9.6, "learning_rate": 2.00093926111459e-05, "loss": 0.532, "step": 11354, "task_loss": 0.3015058934688568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6894526481628418, "epoch": 9.6, "learning_rate": 2.0006261740763935e-05, "loss": 0.6186, "step": 11355, "task_loss": 1.2999649047851562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5244376063346863, "epoch": 9.6, "learning_rate": 2.0003130870381966e-05, "loss": 0.5651, "step": 11356, "task_loss": 1.0293217897415161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4024331867694855, "epoch": 9.6, "learning_rate": 2e-05, "loss": 0.4329, "step": 11357, "task_loss": 0.9846789240837097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6993475556373596, "epoch": 9.6, "learning_rate": 1.9996869129618033e-05, "loss": 0.5814, "step": 11358, "task_loss": 0.7011184096336365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39042922854423523, "epoch": 9.6, "learning_rate": 1.999373825923607e-05, "loss": 0.4291, "step": 11359, "task_loss": 0.7307684421539307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4727763533592224, "epoch": 9.6, "learning_rate": 1.99906073888541e-05, "loss": 0.4748, "step": 11360, "task_loss": 0.7980494499206543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4121585190296173, "epoch": 9.6, "learning_rate": 1.9987476518472136e-05, "loss": 0.5562, "step": 11361, "task_loss": 0.840019941329956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3546028733253479, "epoch": 9.6, "learning_rate": 1.998434564809017e-05, "loss": 0.3664, "step": 11362, "task_loss": 0.18202438950538635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5574893951416016, "epoch": 9.6, "learning_rate": 1.9981214777708206e-05, "loss": 0.4995, "step": 11363, "task_loss": 0.31637024879455566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8587394952774048, "epoch": 9.61, "learning_rate": 1.9978083907326238e-05, "loss": 0.6671, "step": 11364, "task_loss": 0.821760356426239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7169572710990906, "epoch": 9.61, "learning_rate": 1.9974953036944273e-05, "loss": 0.5679, "step": 11365, "task_loss": 0.5971095561981201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8587714433670044, "epoch": 9.61, "learning_rate": 1.9971822166562305e-05, "loss": 0.6032, "step": 11366, "task_loss": 0.6792311072349548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5023744106292725, "epoch": 9.61, "learning_rate": 1.996869129618034e-05, "loss": 0.3885, "step": 11367, "task_loss": 0.4597201645374298 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37267717719078064, "epoch": 9.61, "learning_rate": 1.9965560425798375e-05, "loss": 0.395, "step": 11368, "task_loss": 0.4538896977901459 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3950776159763336, "epoch": 9.61, "learning_rate": 1.9962429555416407e-05, "loss": 0.4433, "step": 11369, "task_loss": 0.2128416895866394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30139949917793274, "epoch": 9.61, "learning_rate": 1.9959298685034442e-05, "loss": 0.6499, "step": 11370, "task_loss": 0.03256037086248398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3988201320171356, "epoch": 9.61, "learning_rate": 1.9956167814652474e-05, "loss": 0.5219, "step": 11371, "task_loss": 0.07429192960262299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7661631107330322, "epoch": 9.61, "learning_rate": 1.995303694427051e-05, "loss": 0.6075, "step": 11372, "task_loss": 0.6280069351196289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43364042043685913, "epoch": 9.61, "learning_rate": 1.994990607388854e-05, "loss": 0.4231, "step": 11373, "task_loss": 0.5607591271400452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4030305743217468, "epoch": 9.61, "learning_rate": 1.9946775203506576e-05, "loss": 0.4984, "step": 11374, "task_loss": 0.825931966304779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5823071002960205, "epoch": 9.61, "learning_rate": 1.9943644333124608e-05, "loss": 0.5691, "step": 11375, "task_loss": 0.3750426173210144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.540593683719635, "epoch": 9.62, "learning_rate": 1.9940513462742643e-05, "loss": 0.5081, "step": 11376, "task_loss": 0.7101706862449646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5890624523162842, "epoch": 9.62, "learning_rate": 1.9937382592360675e-05, "loss": 0.482, "step": 11377, "task_loss": 0.7091532349586487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30429673194885254, "epoch": 9.62, "learning_rate": 1.993425172197871e-05, "loss": 0.4243, "step": 11378, "task_loss": 0.4652881324291229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7892046570777893, "epoch": 9.62, "learning_rate": 1.9931120851596745e-05, "loss": 0.639, "step": 11379, "task_loss": 0.7526462078094482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6391066312789917, "epoch": 9.62, "learning_rate": 1.992798998121478e-05, "loss": 0.6883, "step": 11380, "task_loss": 0.7828980684280396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5865297317504883, "epoch": 9.62, "learning_rate": 1.9924859110832815e-05, "loss": 0.5833, "step": 11381, "task_loss": 0.7035923004150391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.567879319190979, "epoch": 9.62, "learning_rate": 1.9921728240450847e-05, "loss": 0.61, "step": 11382, "task_loss": 0.6984516978263855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9057669639587402, "epoch": 9.62, "learning_rate": 1.9918597370068882e-05, "loss": 0.6658, "step": 11383, "task_loss": 1.1531678438186646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5202196836471558, "epoch": 9.62, "learning_rate": 1.9915466499686914e-05, "loss": 0.5621, "step": 11384, "task_loss": 0.9357413053512573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5346084833145142, "epoch": 9.62, "learning_rate": 1.991233562930495e-05, "loss": 0.5921, "step": 11385, "task_loss": 0.5446224212646484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7095826268196106, "epoch": 9.62, "learning_rate": 1.990920475892298e-05, "loss": 0.5197, "step": 11386, "task_loss": 1.0137746334075928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3225402235984802, "epoch": 9.63, "learning_rate": 1.9906073888541016e-05, "loss": 0.5191, "step": 11387, "task_loss": 0.5679959058761597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5170493125915527, "epoch": 9.63, "learning_rate": 1.9902943018159048e-05, "loss": 0.5496, "step": 11388, "task_loss": 1.2868270874023438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7718780040740967, "epoch": 9.63, "learning_rate": 1.9899812147777083e-05, "loss": 0.6505, "step": 11389, "task_loss": 1.0573375225067139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29062721133232117, "epoch": 9.63, "learning_rate": 1.9896681277395115e-05, "loss": 0.3428, "step": 11390, "task_loss": 0.5122820138931274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9616600275039673, "epoch": 9.63, "learning_rate": 1.989355040701315e-05, "loss": 0.7967, "step": 11391, "task_loss": 2.0062406063079834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44944846630096436, "epoch": 9.63, "learning_rate": 1.9890419536631185e-05, "loss": 0.4908, "step": 11392, "task_loss": 0.22287482023239136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26687294244766235, "epoch": 9.63, "learning_rate": 1.9887288666249217e-05, "loss": 0.5849, "step": 11393, "task_loss": 0.669321596622467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5120130777359009, "epoch": 9.63, "learning_rate": 1.9884157795867252e-05, "loss": 0.4471, "step": 11394, "task_loss": 0.9217043519020081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3826237618923187, "epoch": 9.63, "learning_rate": 1.9881026925485284e-05, "loss": 0.6156, "step": 11395, "task_loss": 0.28106215596199036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23861142992973328, "epoch": 9.63, "learning_rate": 1.987789605510332e-05, "loss": 0.4561, "step": 11396, "task_loss": 0.038684550672769547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7591334581375122, "epoch": 9.63, "learning_rate": 1.9874765184721354e-05, "loss": 0.5904, "step": 11397, "task_loss": 0.3196423649787903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5163205862045288, "epoch": 9.63, "learning_rate": 1.987163431433939e-05, "loss": 0.6226, "step": 11398, "task_loss": 0.9571279287338257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5670987963676453, "epoch": 9.64, "learning_rate": 1.986850344395742e-05, "loss": 0.624, "step": 11399, "task_loss": 0.2808457314968109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2723139524459839, "epoch": 9.64, "learning_rate": 1.9865372573575456e-05, "loss": 0.4162, "step": 11400, "task_loss": 0.5949050188064575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4680194854736328, "epoch": 9.64, "learning_rate": 1.9862241703193488e-05, "loss": 0.5586, "step": 11401, "task_loss": 0.5704677104949951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3888252377510071, "epoch": 9.64, "learning_rate": 1.9859110832811523e-05, "loss": 0.4729, "step": 11402, "task_loss": 0.29112181067466736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4751502275466919, "epoch": 9.64, "learning_rate": 1.9855979962429555e-05, "loss": 0.5727, "step": 11403, "task_loss": 0.5086777806282043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4468209445476532, "epoch": 9.64, "learning_rate": 1.985284909204759e-05, "loss": 0.6018, "step": 11404, "task_loss": 0.4473933279514313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39009177684783936, "epoch": 9.64, "learning_rate": 1.9849718221665625e-05, "loss": 0.4668, "step": 11405, "task_loss": 0.6395938992500305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5462759733200073, "epoch": 9.64, "learning_rate": 1.9846587351283657e-05, "loss": 0.5388, "step": 11406, "task_loss": 0.5903078317642212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3698870539665222, "epoch": 9.64, "learning_rate": 1.9843456480901692e-05, "loss": 0.6363, "step": 11407, "task_loss": 0.5422508716583252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6243787407875061, "epoch": 9.64, "learning_rate": 1.9840325610519724e-05, "loss": 0.7642, "step": 11408, "task_loss": 0.47818905115127563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42326682806015015, "epoch": 9.64, "learning_rate": 1.983719474013776e-05, "loss": 0.4042, "step": 11409, "task_loss": 0.9985173940658569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39241090416908264, "epoch": 9.64, "learning_rate": 1.983406386975579e-05, "loss": 0.5053, "step": 11410, "task_loss": 0.22116412222385406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2376750260591507, "epoch": 9.65, "learning_rate": 1.9830932999373826e-05, "loss": 0.4206, "step": 11411, "task_loss": 0.10352277010679245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28262197971343994, "epoch": 9.65, "learning_rate": 1.982780212899186e-05, "loss": 0.4008, "step": 11412, "task_loss": 0.901085615158081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5719962120056152, "epoch": 9.65, "learning_rate": 1.9824671258609893e-05, "loss": 0.5483, "step": 11413, "task_loss": 0.932773232460022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36694175004959106, "epoch": 9.65, "learning_rate": 1.982154038822793e-05, "loss": 0.3879, "step": 11414, "task_loss": 0.05230306461453438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4778267741203308, "epoch": 9.65, "learning_rate": 1.9818409517845964e-05, "loss": 0.5518, "step": 11415, "task_loss": 0.8740503787994385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35519886016845703, "epoch": 9.65, "learning_rate": 1.9815278647464e-05, "loss": 0.6519, "step": 11416, "task_loss": 0.2849069833755493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4767727255821228, "epoch": 9.65, "learning_rate": 1.981214777708203e-05, "loss": 0.5227, "step": 11417, "task_loss": 0.49215203523635864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.66871178150177, "epoch": 9.65, "learning_rate": 1.9809016906700066e-05, "loss": 0.4706, "step": 11418, "task_loss": 0.315182626247406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4731830060482025, "epoch": 9.65, "learning_rate": 1.9805886036318097e-05, "loss": 0.6245, "step": 11419, "task_loss": 0.30815568566322327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8387899398803711, "epoch": 9.65, "learning_rate": 1.9802755165936133e-05, "loss": 0.8308, "step": 11420, "task_loss": 1.1443027257919312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5440508127212524, "epoch": 9.65, "learning_rate": 1.9799624295554164e-05, "loss": 0.4785, "step": 11421, "task_loss": 0.9492506384849548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47129207849502563, "epoch": 9.65, "learning_rate": 1.97964934251722e-05, "loss": 0.4853, "step": 11422, "task_loss": 0.12125322967767715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29167166352272034, "epoch": 9.66, "learning_rate": 1.979336255479023e-05, "loss": 0.4712, "step": 11423, "task_loss": 0.20951198041439056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43718647956848145, "epoch": 9.66, "learning_rate": 1.9790231684408267e-05, "loss": 0.4521, "step": 11424, "task_loss": 0.5266664028167725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6764252781867981, "epoch": 9.66, "learning_rate": 1.9787100814026298e-05, "loss": 0.5871, "step": 11425, "task_loss": 0.9662127494812012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3502879738807678, "epoch": 9.66, "learning_rate": 1.9783969943644333e-05, "loss": 0.5569, "step": 11426, "task_loss": 0.40940719842910767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36334964632987976, "epoch": 9.66, "learning_rate": 1.9780839073262365e-05, "loss": 0.4527, "step": 11427, "task_loss": 0.5619573593139648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33798685669898987, "epoch": 9.66, "learning_rate": 1.97777082028804e-05, "loss": 0.6085, "step": 11428, "task_loss": 0.25684118270874023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4906187057495117, "epoch": 9.66, "learning_rate": 1.9774577332498436e-05, "loss": 0.5594, "step": 11429, "task_loss": 1.056955337524414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4191204011440277, "epoch": 9.66, "learning_rate": 1.977144646211647e-05, "loss": 0.5165, "step": 11430, "task_loss": 0.9844582080841064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5088682174682617, "epoch": 9.66, "learning_rate": 1.9768315591734503e-05, "loss": 0.5848, "step": 11431, "task_loss": 0.5589878559112549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4178157448768616, "epoch": 9.66, "learning_rate": 1.9765184721352538e-05, "loss": 0.4082, "step": 11432, "task_loss": 0.46547937393188477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5044748187065125, "epoch": 9.66, "learning_rate": 1.9762053850970573e-05, "loss": 0.5164, "step": 11433, "task_loss": 1.6895028352737427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6376059055328369, "epoch": 9.66, "learning_rate": 1.9758922980588605e-05, "loss": 0.5582, "step": 11434, "task_loss": 0.8395466804504395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41408947110176086, "epoch": 9.67, "learning_rate": 1.975579211020664e-05, "loss": 0.4587, "step": 11435, "task_loss": 0.22425857186317444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3364698588848114, "epoch": 9.67, "learning_rate": 1.975266123982467e-05, "loss": 0.5395, "step": 11436, "task_loss": 0.45064428448677063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.546302080154419, "epoch": 9.67, "learning_rate": 1.9749530369442707e-05, "loss": 0.5053, "step": 11437, "task_loss": 0.9727370142936707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2698715329170227, "epoch": 9.67, "learning_rate": 1.974639949906074e-05, "loss": 0.4315, "step": 11438, "task_loss": 0.15951062738895416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5442439913749695, "epoch": 9.67, "learning_rate": 1.9743268628678774e-05, "loss": 0.5499, "step": 11439, "task_loss": 1.5045188665390015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6771121025085449, "epoch": 9.67, "learning_rate": 1.9740137758296805e-05, "loss": 0.6006, "step": 11440, "task_loss": 0.3773405849933624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6824321746826172, "epoch": 9.67, "learning_rate": 1.973700688791484e-05, "loss": 0.5864, "step": 11441, "task_loss": 0.7986275553703308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2848937511444092, "epoch": 9.67, "learning_rate": 1.9733876017532876e-05, "loss": 0.5197, "step": 11442, "task_loss": 0.6212865710258484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4321041703224182, "epoch": 9.67, "learning_rate": 1.9730745147150908e-05, "loss": 0.5595, "step": 11443, "task_loss": 0.5271742343902588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7331293821334839, "epoch": 9.67, "learning_rate": 1.9727614276768943e-05, "loss": 0.632, "step": 11444, "task_loss": 1.1029493808746338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5061719417572021, "epoch": 9.67, "learning_rate": 1.9724483406386975e-05, "loss": 0.4792, "step": 11445, "task_loss": 1.148537516593933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5293819904327393, "epoch": 9.67, "learning_rate": 1.972135253600501e-05, "loss": 0.6463, "step": 11446, "task_loss": 0.4697108268737793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7695452570915222, "epoch": 9.68, "learning_rate": 1.9718221665623045e-05, "loss": 0.6162, "step": 11447, "task_loss": 1.2598845958709717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.510299801826477, "epoch": 9.68, "learning_rate": 1.971509079524108e-05, "loss": 0.4908, "step": 11448, "task_loss": 0.19200201332569122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43731582164764404, "epoch": 9.68, "learning_rate": 1.9711959924859112e-05, "loss": 0.4977, "step": 11449, "task_loss": 0.26628032326698303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20925530791282654, "epoch": 9.68, "learning_rate": 1.9708829054477147e-05, "loss": 0.53, "step": 11450, "task_loss": 0.8568643927574158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4500838816165924, "epoch": 9.68, "learning_rate": 1.970569818409518e-05, "loss": 0.4915, "step": 11451, "task_loss": 1.0174129009246826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3408641815185547, "epoch": 9.68, "learning_rate": 1.9702567313713214e-05, "loss": 0.5995, "step": 11452, "task_loss": 0.5018743276596069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7217057347297668, "epoch": 9.68, "learning_rate": 1.969943644333125e-05, "loss": 0.6, "step": 11453, "task_loss": 1.083902359008789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6915547847747803, "epoch": 9.68, "learning_rate": 1.969630557294928e-05, "loss": 0.5558, "step": 11454, "task_loss": 0.4923596680164337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.726566731929779, "epoch": 9.68, "learning_rate": 1.9693174702567316e-05, "loss": 0.6207, "step": 11455, "task_loss": 0.4797140657901764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44524961709976196, "epoch": 9.68, "learning_rate": 1.9690043832185348e-05, "loss": 0.6729, "step": 11456, "task_loss": 0.638375461101532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8899569511413574, "epoch": 9.68, "learning_rate": 1.9686912961803383e-05, "loss": 0.5606, "step": 11457, "task_loss": 0.3166080415248871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31994301080703735, "epoch": 9.69, "learning_rate": 1.9683782091421415e-05, "loss": 0.4632, "step": 11458, "task_loss": 0.4644120931625366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5142444372177124, "epoch": 9.69, "learning_rate": 1.968065122103945e-05, "loss": 0.4427, "step": 11459, "task_loss": 0.26330187916755676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4225069284439087, "epoch": 9.69, "learning_rate": 1.9677520350657482e-05, "loss": 0.4853, "step": 11460, "task_loss": 0.860550582408905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33833760023117065, "epoch": 9.69, "learning_rate": 1.9674389480275517e-05, "loss": 0.4997, "step": 11461, "task_loss": 1.2459710836410522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3163233995437622, "epoch": 9.69, "learning_rate": 1.967125860989355e-05, "loss": 0.5421, "step": 11462, "task_loss": 0.5153548121452332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8136933445930481, "epoch": 9.69, "learning_rate": 1.9668127739511584e-05, "loss": 0.6041, "step": 11463, "task_loss": 0.427778422832489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5104421377182007, "epoch": 9.69, "learning_rate": 1.966499686912962e-05, "loss": 0.6406, "step": 11464, "task_loss": 0.314953476190567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6073487997055054, "epoch": 9.69, "learning_rate": 1.9661865998747654e-05, "loss": 0.5725, "step": 11465, "task_loss": 0.7612685561180115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4365811347961426, "epoch": 9.69, "learning_rate": 1.965873512836569e-05, "loss": 0.4758, "step": 11466, "task_loss": 0.33550599217414856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26628217101097107, "epoch": 9.69, "learning_rate": 1.965560425798372e-05, "loss": 0.5183, "step": 11467, "task_loss": 0.13872650265693665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8433094024658203, "epoch": 9.69, "learning_rate": 1.9652473387601756e-05, "loss": 0.6506, "step": 11468, "task_loss": 0.9560882449150085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46675586700439453, "epoch": 9.69, "learning_rate": 1.9649342517219788e-05, "loss": 0.621, "step": 11469, "task_loss": 1.1390206813812256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7161898612976074, "epoch": 9.7, "learning_rate": 1.9646211646837823e-05, "loss": 0.4415, "step": 11470, "task_loss": 0.8804476261138916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41045141220092773, "epoch": 9.7, "learning_rate": 1.9643080776455855e-05, "loss": 0.5036, "step": 11471, "task_loss": 0.4508552849292755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4772064685821533, "epoch": 9.7, "learning_rate": 1.963994990607389e-05, "loss": 0.4534, "step": 11472, "task_loss": 0.2750204801559448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4010735750198364, "epoch": 9.7, "learning_rate": 1.9636819035691922e-05, "loss": 0.591, "step": 11473, "task_loss": 0.421718955039978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2796887755393982, "epoch": 9.7, "learning_rate": 1.9633688165309957e-05, "loss": 0.4288, "step": 11474, "task_loss": 0.27300184965133667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6914469599723816, "epoch": 9.7, "learning_rate": 1.963055729492799e-05, "loss": 0.5993, "step": 11475, "task_loss": 0.4673765301704407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6049767732620239, "epoch": 9.7, "learning_rate": 1.9627426424546024e-05, "loss": 0.528, "step": 11476, "task_loss": 0.584034264087677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4021163582801819, "epoch": 9.7, "learning_rate": 1.9624295554164056e-05, "loss": 0.4697, "step": 11477, "task_loss": 0.34257805347442627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3405894637107849, "epoch": 9.7, "learning_rate": 1.962116468378209e-05, "loss": 0.4369, "step": 11478, "task_loss": 0.33075231313705444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6115224361419678, "epoch": 9.7, "learning_rate": 1.9618033813400126e-05, "loss": 0.5725, "step": 11479, "task_loss": 0.9345648288726807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7578883767127991, "epoch": 9.7, "learning_rate": 1.9614902943018158e-05, "loss": 0.5698, "step": 11480, "task_loss": 1.1056239604949951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5493202209472656, "epoch": 9.7, "learning_rate": 1.9611772072636193e-05, "loss": 0.4806, "step": 11481, "task_loss": 0.928357720375061 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6181823015213013, "epoch": 9.71, "learning_rate": 1.960864120225423e-05, "loss": 0.5571, "step": 11482, "task_loss": 0.59356689453125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42304903268814087, "epoch": 9.71, "learning_rate": 1.9605510331872264e-05, "loss": 0.6062, "step": 11483, "task_loss": 1.228961706161499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.564358115196228, "epoch": 9.71, "learning_rate": 1.9602379461490295e-05, "loss": 0.6745, "step": 11484, "task_loss": 1.1679726839065552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4858130216598511, "epoch": 9.71, "learning_rate": 1.959924859110833e-05, "loss": 0.5082, "step": 11485, "task_loss": 0.831802248954773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44415363669395447, "epoch": 9.71, "learning_rate": 1.9596117720726362e-05, "loss": 0.5701, "step": 11486, "task_loss": 1.9158393144607544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4670506715774536, "epoch": 9.71, "learning_rate": 1.9592986850344397e-05, "loss": 0.703, "step": 11487, "task_loss": 0.5403284430503845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8512698411941528, "epoch": 9.71, "learning_rate": 1.958985597996243e-05, "loss": 0.5785, "step": 11488, "task_loss": 1.3781265020370483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7522706389427185, "epoch": 9.71, "learning_rate": 1.9586725109580464e-05, "loss": 0.6375, "step": 11489, "task_loss": 0.8087536096572876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4348558187484741, "epoch": 9.71, "learning_rate": 1.95835942391985e-05, "loss": 0.5778, "step": 11490, "task_loss": 0.37708336114883423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3197459578514099, "epoch": 9.71, "learning_rate": 1.958046336881653e-05, "loss": 0.4858, "step": 11491, "task_loss": 0.18276098370552063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5182896852493286, "epoch": 9.71, "learning_rate": 1.9577332498434567e-05, "loss": 0.5928, "step": 11492, "task_loss": 0.5755763053894043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6586048007011414, "epoch": 9.71, "learning_rate": 1.95742016280526e-05, "loss": 0.6127, "step": 11493, "task_loss": 0.8938744068145752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5987610816955566, "epoch": 9.72, "learning_rate": 1.9571070757670633e-05, "loss": 0.6644, "step": 11494, "task_loss": 1.0883281230926514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5454007387161255, "epoch": 9.72, "learning_rate": 1.9567939887288665e-05, "loss": 0.6427, "step": 11495, "task_loss": 0.8469494581222534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32154542207717896, "epoch": 9.72, "learning_rate": 1.95648090169067e-05, "loss": 0.5438, "step": 11496, "task_loss": 0.5184254050254822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3503858745098114, "epoch": 9.72, "learning_rate": 1.9561678146524736e-05, "loss": 0.4417, "step": 11497, "task_loss": 0.7661807537078857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5276139378547668, "epoch": 9.72, "learning_rate": 1.9558547276142767e-05, "loss": 0.5982, "step": 11498, "task_loss": 0.4580424427986145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5940665006637573, "epoch": 9.72, "learning_rate": 1.9555416405760803e-05, "loss": 0.4639, "step": 11499, "task_loss": 0.17221124470233917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49801111221313477, "epoch": 9.72, "learning_rate": 1.9552285535378838e-05, "loss": 0.589, "step": 11500, "task_loss": 0.9272472858428955 }, { "epoch": 9.72, "eval_accuracy": 0.9050693069306931, "eval_loss": 0.3670479953289032, "eval_runtime": 208.4788, "eval_samples_per_second": 121.115, "eval_steps_per_second": 0.95, "step": 11500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5055623650550842, "epoch": 9.72, "learning_rate": 1.954915466499687e-05, "loss": 0.5683, "step": 11501, "task_loss": 0.3087790012359619 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5195136070251465, "epoch": 9.72, "learning_rate": 1.9546023794614905e-05, "loss": 0.4736, "step": 11502, "task_loss": 0.17636236548423767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29617321491241455, "epoch": 9.72, "learning_rate": 1.954289292423294e-05, "loss": 0.4949, "step": 11503, "task_loss": 0.7396499514579773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4905962646007538, "epoch": 9.72, "learning_rate": 1.953976205385097e-05, "loss": 0.6237, "step": 11504, "task_loss": 0.1277923434972763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44684380292892456, "epoch": 9.72, "learning_rate": 1.9536631183469007e-05, "loss": 0.454, "step": 11505, "task_loss": 0.9906193017959595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30436408519744873, "epoch": 9.73, "learning_rate": 1.953350031308704e-05, "loss": 0.3783, "step": 11506, "task_loss": 0.1257229745388031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6980328559875488, "epoch": 9.73, "learning_rate": 1.9530369442705074e-05, "loss": 0.4828, "step": 11507, "task_loss": 0.5750572085380554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6562527418136597, "epoch": 9.73, "learning_rate": 1.9527238572323106e-05, "loss": 0.5416, "step": 11508, "task_loss": 1.1739014387130737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34045255184173584, "epoch": 9.73, "learning_rate": 1.952410770194114e-05, "loss": 0.464, "step": 11509, "task_loss": 0.10824192315340042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5384789705276489, "epoch": 9.73, "learning_rate": 1.9520976831559172e-05, "loss": 0.6217, "step": 11510, "task_loss": 0.5669267177581787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.664076030254364, "epoch": 9.73, "learning_rate": 1.9517845961177208e-05, "loss": 0.794, "step": 11511, "task_loss": 1.052617073059082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5710515975952148, "epoch": 9.73, "learning_rate": 1.951471509079524e-05, "loss": 0.537, "step": 11512, "task_loss": 0.4157852232456207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6853793263435364, "epoch": 9.73, "learning_rate": 1.9511584220413275e-05, "loss": 0.7145, "step": 11513, "task_loss": 0.3109949231147766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5381698608398438, "epoch": 9.73, "learning_rate": 1.950845335003131e-05, "loss": 0.5616, "step": 11514, "task_loss": 0.23374778032302856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3592284023761749, "epoch": 9.73, "learning_rate": 1.9505322479649345e-05, "loss": 0.4228, "step": 11515, "task_loss": 0.4370739161968231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35530805587768555, "epoch": 9.73, "learning_rate": 1.950219160926738e-05, "loss": 0.4589, "step": 11516, "task_loss": 0.13837535679340363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.644140362739563, "epoch": 9.73, "learning_rate": 1.9499060738885412e-05, "loss": 0.5468, "step": 11517, "task_loss": 1.482623815536499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5352402329444885, "epoch": 9.74, "learning_rate": 1.9495929868503447e-05, "loss": 0.5958, "step": 11518, "task_loss": 0.8604409098625183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46655797958374023, "epoch": 9.74, "learning_rate": 1.949279899812148e-05, "loss": 0.5524, "step": 11519, "task_loss": 0.5307003855705261 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22586563229560852, "epoch": 9.74, "learning_rate": 1.9489668127739514e-05, "loss": 0.4928, "step": 11520, "task_loss": 0.11024802923202515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4278532862663269, "epoch": 9.74, "learning_rate": 1.9486537257357546e-05, "loss": 0.515, "step": 11521, "task_loss": 0.27911853790283203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7995514869689941, "epoch": 9.74, "learning_rate": 1.948340638697558e-05, "loss": 0.6132, "step": 11522, "task_loss": 0.6405843496322632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3479260206222534, "epoch": 9.74, "learning_rate": 1.9480275516593613e-05, "loss": 0.709, "step": 11523, "task_loss": 0.35373392701148987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36721569299697876, "epoch": 9.74, "learning_rate": 1.9477144646211648e-05, "loss": 0.4897, "step": 11524, "task_loss": 0.4563276469707489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5381394028663635, "epoch": 9.74, "learning_rate": 1.947401377582968e-05, "loss": 0.5057, "step": 11525, "task_loss": 0.7897980809211731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5762666463851929, "epoch": 9.74, "learning_rate": 1.9470882905447715e-05, "loss": 0.5534, "step": 11526, "task_loss": 0.8411628007888794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7679926753044128, "epoch": 9.74, "learning_rate": 1.946775203506575e-05, "loss": 0.5556, "step": 11527, "task_loss": 0.5225948095321655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5415346026420593, "epoch": 9.74, "learning_rate": 1.9464621164683782e-05, "loss": 0.5479, "step": 11528, "task_loss": 1.3079636096954346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4301774203777313, "epoch": 9.75, "learning_rate": 1.9461490294301817e-05, "loss": 0.4198, "step": 11529, "task_loss": 0.5403162837028503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6916524767875671, "epoch": 9.75, "learning_rate": 1.945835942391985e-05, "loss": 0.4477, "step": 11530, "task_loss": 0.3158518075942993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4151741862297058, "epoch": 9.75, "learning_rate": 1.9455228553537884e-05, "loss": 0.5158, "step": 11531, "task_loss": 0.7024104595184326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3743663430213928, "epoch": 9.75, "learning_rate": 1.945209768315592e-05, "loss": 0.4684, "step": 11532, "task_loss": 0.46835991740226746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9407538175582886, "epoch": 9.75, "learning_rate": 1.9448966812773954e-05, "loss": 0.6876, "step": 11533, "task_loss": 1.1040998697280884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3550580143928528, "epoch": 9.75, "learning_rate": 1.9445835942391986e-05, "loss": 0.5067, "step": 11534, "task_loss": 0.19404403865337372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4991888105869293, "epoch": 9.75, "learning_rate": 1.944270507201002e-05, "loss": 0.4778, "step": 11535, "task_loss": 1.1981998682022095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33815085887908936, "epoch": 9.75, "learning_rate": 1.9439574201628053e-05, "loss": 0.5725, "step": 11536, "task_loss": 0.4922480881214142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6942065954208374, "epoch": 9.75, "learning_rate": 1.9436443331246088e-05, "loss": 0.6307, "step": 11537, "task_loss": 1.0228185653686523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.232534408569336, "epoch": 9.75, "learning_rate": 1.943331246086412e-05, "loss": 0.7882, "step": 11538, "task_loss": 1.3487833738327026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49025022983551025, "epoch": 9.75, "learning_rate": 1.9430181590482155e-05, "loss": 0.4942, "step": 11539, "task_loss": 0.25904083251953125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43615204095840454, "epoch": 9.75, "learning_rate": 1.942705072010019e-05, "loss": 0.563, "step": 11540, "task_loss": 0.801078200340271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37649333477020264, "epoch": 9.76, "learning_rate": 1.9423919849718222e-05, "loss": 0.5688, "step": 11541, "task_loss": 0.18831807374954224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5912031531333923, "epoch": 9.76, "learning_rate": 1.9420788979336257e-05, "loss": 0.6711, "step": 11542, "task_loss": 1.584781527519226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6591867804527283, "epoch": 9.76, "learning_rate": 1.941765810895429e-05, "loss": 0.5197, "step": 11543, "task_loss": 0.7277752161026001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.648594319820404, "epoch": 9.76, "learning_rate": 1.9414527238572324e-05, "loss": 0.518, "step": 11544, "task_loss": 1.1423958539962769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5593070983886719, "epoch": 9.76, "learning_rate": 1.9411396368190356e-05, "loss": 0.4375, "step": 11545, "task_loss": 0.71848464012146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5305392742156982, "epoch": 9.76, "learning_rate": 1.940826549780839e-05, "loss": 0.5749, "step": 11546, "task_loss": 0.8690393567085266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49656644463539124, "epoch": 9.76, "learning_rate": 1.9405134627426423e-05, "loss": 0.5961, "step": 11547, "task_loss": 0.4500821828842163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8034920692443848, "epoch": 9.76, "learning_rate": 1.9402003757044458e-05, "loss": 0.5929, "step": 11548, "task_loss": 0.47071853280067444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5188835859298706, "epoch": 9.76, "learning_rate": 1.9398872886662493e-05, "loss": 0.4872, "step": 11549, "task_loss": 0.9092277884483337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7299895882606506, "epoch": 9.76, "learning_rate": 1.939574201628053e-05, "loss": 0.6106, "step": 11550, "task_loss": 0.989351212978363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5251337289810181, "epoch": 9.76, "learning_rate": 1.9392611145898564e-05, "loss": 0.5233, "step": 11551, "task_loss": 0.5385008454322815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7040932178497314, "epoch": 9.76, "learning_rate": 1.9389480275516595e-05, "loss": 0.7615, "step": 11552, "task_loss": 1.3508975505828857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5177218317985535, "epoch": 9.77, "learning_rate": 1.938634940513463e-05, "loss": 0.5612, "step": 11553, "task_loss": 1.1349859237670898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45916664600372314, "epoch": 9.77, "learning_rate": 1.9383218534752662e-05, "loss": 0.3912, "step": 11554, "task_loss": 0.8141466975212097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5239891409873962, "epoch": 9.77, "learning_rate": 1.9380087664370698e-05, "loss": 0.5367, "step": 11555, "task_loss": 0.5013299584388733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4903170168399811, "epoch": 9.77, "learning_rate": 1.937695679398873e-05, "loss": 0.4922, "step": 11556, "task_loss": 0.3154742419719696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5253037810325623, "epoch": 9.77, "learning_rate": 1.9373825923606764e-05, "loss": 0.5804, "step": 11557, "task_loss": 0.7890669107437134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8684146404266357, "epoch": 9.77, "learning_rate": 1.9370695053224796e-05, "loss": 0.716, "step": 11558, "task_loss": 0.4173005521297455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6388171911239624, "epoch": 9.77, "learning_rate": 1.936756418284283e-05, "loss": 0.6312, "step": 11559, "task_loss": 0.9064346551895142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5461736917495728, "epoch": 9.77, "learning_rate": 1.9364433312460863e-05, "loss": 0.6342, "step": 11560, "task_loss": 0.6102699041366577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6301363706588745, "epoch": 9.77, "learning_rate": 1.93613024420789e-05, "loss": 0.457, "step": 11561, "task_loss": 0.7611414194107056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5485497117042542, "epoch": 9.77, "learning_rate": 1.935817157169693e-05, "loss": 0.8057, "step": 11562, "task_loss": 0.38453030586242676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3792394697666168, "epoch": 9.77, "learning_rate": 1.9355040701314965e-05, "loss": 0.4725, "step": 11563, "task_loss": 0.7644608020782471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40754276514053345, "epoch": 9.77, "learning_rate": 1.9351909830933e-05, "loss": 0.5365, "step": 11564, "task_loss": 0.23216968774795532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5285704731941223, "epoch": 9.78, "learning_rate": 1.9348778960551032e-05, "loss": 0.5262, "step": 11565, "task_loss": 0.6981925964355469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33965396881103516, "epoch": 9.78, "learning_rate": 1.9345648090169067e-05, "loss": 0.387, "step": 11566, "task_loss": 0.6116869449615479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5505032539367676, "epoch": 9.78, "learning_rate": 1.9342517219787103e-05, "loss": 0.5972, "step": 11567, "task_loss": 0.41497817635536194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.462602436542511, "epoch": 9.78, "learning_rate": 1.9339386349405138e-05, "loss": 0.5303, "step": 11568, "task_loss": 1.3448935747146606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5178964138031006, "epoch": 9.78, "learning_rate": 1.933625547902317e-05, "loss": 0.544, "step": 11569, "task_loss": 1.0277327299118042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7191877961158752, "epoch": 9.78, "learning_rate": 1.9333124608641205e-05, "loss": 0.7025, "step": 11570, "task_loss": 1.7156333923339844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7906275987625122, "epoch": 9.78, "learning_rate": 1.9329993738259236e-05, "loss": 0.537, "step": 11571, "task_loss": 0.6520065069198608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9213466048240662, "epoch": 9.78, "learning_rate": 1.932686286787727e-05, "loss": 0.7303, "step": 11572, "task_loss": 1.2434160709381104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3979056775569916, "epoch": 9.78, "learning_rate": 1.9323731997495303e-05, "loss": 0.4742, "step": 11573, "task_loss": 0.40355876088142395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3085879981517792, "epoch": 9.78, "learning_rate": 1.932060112711334e-05, "loss": 0.4179, "step": 11574, "task_loss": 0.8947088718414307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5550545454025269, "epoch": 9.78, "learning_rate": 1.931747025673137e-05, "loss": 0.6216, "step": 11575, "task_loss": 0.5907147526741028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49374741315841675, "epoch": 9.78, "learning_rate": 1.9314339386349406e-05, "loss": 0.5298, "step": 11576, "task_loss": 0.8309386968612671 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23060837388038635, "epoch": 9.79, "learning_rate": 1.931120851596744e-05, "loss": 0.4744, "step": 11577, "task_loss": 0.4298765957355499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7837389707565308, "epoch": 9.79, "learning_rate": 1.9308077645585473e-05, "loss": 0.6037, "step": 11578, "task_loss": 0.4761753976345062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8629990816116333, "epoch": 9.79, "learning_rate": 1.9304946775203508e-05, "loss": 0.7452, "step": 11579, "task_loss": 0.7306995391845703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5794805884361267, "epoch": 9.79, "learning_rate": 1.930181590482154e-05, "loss": 0.4533, "step": 11580, "task_loss": 1.1658952236175537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5774139165878296, "epoch": 9.79, "learning_rate": 1.9298685034439575e-05, "loss": 0.5396, "step": 11581, "task_loss": 0.3340483605861664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49272406101226807, "epoch": 9.79, "learning_rate": 1.929555416405761e-05, "loss": 0.4421, "step": 11582, "task_loss": 0.6467495560646057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5440459251403809, "epoch": 9.79, "learning_rate": 1.929242329367564e-05, "loss": 0.6028, "step": 11583, "task_loss": 0.11604884266853333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5874073505401611, "epoch": 9.79, "learning_rate": 1.9289292423293677e-05, "loss": 0.532, "step": 11584, "task_loss": 0.9532662630081177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5955988168716431, "epoch": 9.79, "learning_rate": 1.9286161552911712e-05, "loss": 0.5659, "step": 11585, "task_loss": 0.6402496099472046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7460747361183167, "epoch": 9.79, "learning_rate": 1.9283030682529744e-05, "loss": 0.6155, "step": 11586, "task_loss": 0.7455217838287354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42250698804855347, "epoch": 9.79, "learning_rate": 1.927989981214778e-05, "loss": 0.5438, "step": 11587, "task_loss": 1.042486548423767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6501675248146057, "epoch": 9.79, "learning_rate": 1.9276768941765814e-05, "loss": 0.5539, "step": 11588, "task_loss": 0.29313334822654724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7363775968551636, "epoch": 9.8, "learning_rate": 1.9273638071383846e-05, "loss": 0.6043, "step": 11589, "task_loss": 0.6457009315490723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5330055952072144, "epoch": 9.8, "learning_rate": 1.927050720100188e-05, "loss": 0.5794, "step": 11590, "task_loss": 0.6738833785057068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6053811311721802, "epoch": 9.8, "learning_rate": 1.9267376330619913e-05, "loss": 0.6913, "step": 11591, "task_loss": 0.31903544068336487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31426000595092773, "epoch": 9.8, "learning_rate": 1.9264245460237948e-05, "loss": 0.6199, "step": 11592, "task_loss": 0.4453038275241852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4127032458782196, "epoch": 9.8, "learning_rate": 1.926111458985598e-05, "loss": 0.4937, "step": 11593, "task_loss": 0.2862144708633423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45055079460144043, "epoch": 9.8, "learning_rate": 1.9257983719474015e-05, "loss": 0.5383, "step": 11594, "task_loss": 0.9530044198036194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48205122351646423, "epoch": 9.8, "learning_rate": 1.9254852849092047e-05, "loss": 0.472, "step": 11595, "task_loss": 0.2901204526424408 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29041963815689087, "epoch": 9.8, "learning_rate": 1.9251721978710082e-05, "loss": 0.4681, "step": 11596, "task_loss": 1.7053228616714478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43605008721351624, "epoch": 9.8, "learning_rate": 1.9248591108328114e-05, "loss": 0.4204, "step": 11597, "task_loss": 0.6596958637237549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3363676071166992, "epoch": 9.8, "learning_rate": 1.924546023794615e-05, "loss": 0.6752, "step": 11598, "task_loss": 0.8508536219596863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34936535358428955, "epoch": 9.8, "learning_rate": 1.9242329367564184e-05, "loss": 0.4866, "step": 11599, "task_loss": 0.43759697675704956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29865720868110657, "epoch": 9.81, "learning_rate": 1.923919849718222e-05, "loss": 0.4272, "step": 11600, "task_loss": 0.25117528438568115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37694886326789856, "epoch": 9.81, "learning_rate": 1.9236067626800254e-05, "loss": 0.5007, "step": 11601, "task_loss": 0.6251943707466125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6062031388282776, "epoch": 9.81, "learning_rate": 1.9232936756418286e-05, "loss": 0.5204, "step": 11602, "task_loss": 0.16507107019424438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9894158840179443, "epoch": 9.81, "learning_rate": 1.922980588603632e-05, "loss": 0.6728, "step": 11603, "task_loss": 0.3643621802330017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5071310997009277, "epoch": 9.81, "learning_rate": 1.9226675015654353e-05, "loss": 0.5007, "step": 11604, "task_loss": 0.4692237377166748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42064419388771057, "epoch": 9.81, "learning_rate": 1.9223544145272388e-05, "loss": 0.4918, "step": 11605, "task_loss": 0.41767436265945435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3294799327850342, "epoch": 9.81, "learning_rate": 1.922041327489042e-05, "loss": 0.5259, "step": 11606, "task_loss": 0.6014008522033691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36913594603538513, "epoch": 9.81, "learning_rate": 1.9217282404508455e-05, "loss": 0.5324, "step": 11607, "task_loss": 0.6401878595352173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6220061779022217, "epoch": 9.81, "learning_rate": 1.9214151534126487e-05, "loss": 0.6252, "step": 11608, "task_loss": 0.5668118596076965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33942121267318726, "epoch": 9.81, "learning_rate": 1.9211020663744522e-05, "loss": 0.6232, "step": 11609, "task_loss": 0.33705922961235046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7575415968894958, "epoch": 9.81, "learning_rate": 1.9207889793362554e-05, "loss": 0.6018, "step": 11610, "task_loss": 0.8695986866950989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2809826731681824, "epoch": 9.81, "learning_rate": 1.920475892298059e-05, "loss": 0.4007, "step": 11611, "task_loss": 0.5300547480583191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3313601315021515, "epoch": 9.82, "learning_rate": 1.920162805259862e-05, "loss": 0.5764, "step": 11612, "task_loss": 0.2890830934047699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49127912521362305, "epoch": 9.82, "learning_rate": 1.9198497182216656e-05, "loss": 0.533, "step": 11613, "task_loss": 0.400774210691452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40472733974456787, "epoch": 9.82, "learning_rate": 1.919536631183469e-05, "loss": 0.5049, "step": 11614, "task_loss": 0.82468581199646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4780983328819275, "epoch": 9.82, "learning_rate": 1.9192235441452723e-05, "loss": 0.4921, "step": 11615, "task_loss": 0.9737722277641296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2713300585746765, "epoch": 9.82, "learning_rate": 1.9189104571070758e-05, "loss": 0.4388, "step": 11616, "task_loss": 0.7067311406135559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25810959935188293, "epoch": 9.82, "learning_rate": 1.9185973700688793e-05, "loss": 0.5488, "step": 11617, "task_loss": 0.06526412814855576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.653371274471283, "epoch": 9.82, "learning_rate": 1.918284283030683e-05, "loss": 0.3965, "step": 11618, "task_loss": 0.7719858884811401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6831097602844238, "epoch": 9.82, "learning_rate": 1.917971195992486e-05, "loss": 0.4826, "step": 11619, "task_loss": 0.6945610046386719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40035974979400635, "epoch": 9.82, "learning_rate": 1.9176581089542895e-05, "loss": 0.4684, "step": 11620, "task_loss": 1.38874351978302 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4912932813167572, "epoch": 9.82, "learning_rate": 1.9173450219160927e-05, "loss": 0.7159, "step": 11621, "task_loss": 0.48841750621795654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6443482637405396, "epoch": 9.82, "learning_rate": 1.9170319348778962e-05, "loss": 0.6035, "step": 11622, "task_loss": 0.3873414695262909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4334559142589569, "epoch": 9.82, "learning_rate": 1.9167188478396994e-05, "loss": 0.4711, "step": 11623, "task_loss": 0.3239089846611023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6895104646682739, "epoch": 9.83, "learning_rate": 1.916405760801503e-05, "loss": 0.5415, "step": 11624, "task_loss": 1.085261344909668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6037175059318542, "epoch": 9.83, "learning_rate": 1.9160926737633064e-05, "loss": 0.5711, "step": 11625, "task_loss": 0.40324530005455017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8174339532852173, "epoch": 9.83, "learning_rate": 1.9157795867251096e-05, "loss": 0.6082, "step": 11626, "task_loss": 0.9434210062026978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35624364018440247, "epoch": 9.83, "learning_rate": 1.915466499686913e-05, "loss": 0.4692, "step": 11627, "task_loss": 0.5591573119163513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3987756669521332, "epoch": 9.83, "learning_rate": 1.9151534126487163e-05, "loss": 0.5968, "step": 11628, "task_loss": 1.1687684059143066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4443475008010864, "epoch": 9.83, "learning_rate": 1.91484032561052e-05, "loss": 0.5354, "step": 11629, "task_loss": 1.0343165397644043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6272701621055603, "epoch": 9.83, "learning_rate": 1.914527238572323e-05, "loss": 0.6106, "step": 11630, "task_loss": 0.6612398028373718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31788021326065063, "epoch": 9.83, "learning_rate": 1.9142141515341265e-05, "loss": 0.5084, "step": 11631, "task_loss": 0.17871107161045074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5376238226890564, "epoch": 9.83, "learning_rate": 1.9139010644959297e-05, "loss": 0.4395, "step": 11632, "task_loss": 0.4230214059352875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4418473243713379, "epoch": 9.83, "learning_rate": 1.9135879774577332e-05, "loss": 0.4025, "step": 11633, "task_loss": 0.2775084376335144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5347391963005066, "epoch": 9.83, "learning_rate": 1.9132748904195367e-05, "loss": 0.4937, "step": 11634, "task_loss": 0.35187169909477234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4136130213737488, "epoch": 9.83, "learning_rate": 1.9129618033813403e-05, "loss": 0.4501, "step": 11635, "task_loss": 0.3466648459434509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6766029000282288, "epoch": 9.84, "learning_rate": 1.9126487163431434e-05, "loss": 0.5412, "step": 11636, "task_loss": 0.7170143723487854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3744688332080841, "epoch": 9.84, "learning_rate": 1.912335629304947e-05, "loss": 0.4683, "step": 11637, "task_loss": 1.1521302461624146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8409370183944702, "epoch": 9.84, "learning_rate": 1.9120225422667505e-05, "loss": 0.5236, "step": 11638, "task_loss": 0.8573271632194519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.334531307220459, "epoch": 9.84, "learning_rate": 1.9117094552285537e-05, "loss": 0.4713, "step": 11639, "task_loss": 0.4135565757751465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6684764623641968, "epoch": 9.84, "learning_rate": 1.9113963681903572e-05, "loss": 0.4439, "step": 11640, "task_loss": 0.8059796094894409 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3719712197780609, "epoch": 9.84, "learning_rate": 1.9110832811521603e-05, "loss": 0.5351, "step": 11641, "task_loss": 0.583067774772644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4075293242931366, "epoch": 9.84, "learning_rate": 1.910770194113964e-05, "loss": 0.567, "step": 11642, "task_loss": 0.371187686920166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3296508193016052, "epoch": 9.84, "learning_rate": 1.910457107075767e-05, "loss": 0.4511, "step": 11643, "task_loss": 0.1307692527770996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6070318222045898, "epoch": 9.84, "learning_rate": 1.9101440200375706e-05, "loss": 0.5028, "step": 11644, "task_loss": 0.82576584815979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7037962675094604, "epoch": 9.84, "learning_rate": 1.9098309329993737e-05, "loss": 0.5359, "step": 11645, "task_loss": 1.251917839050293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6993818283081055, "epoch": 9.84, "learning_rate": 1.9095178459611773e-05, "loss": 0.7712, "step": 11646, "task_loss": 0.3885155916213989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46423831582069397, "epoch": 9.84, "learning_rate": 1.9092047589229804e-05, "loss": 0.4277, "step": 11647, "task_loss": 0.28110402822494507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3986390233039856, "epoch": 9.85, "learning_rate": 1.908891671884784e-05, "loss": 0.4214, "step": 11648, "task_loss": 0.4870525002479553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30507081747055054, "epoch": 9.85, "learning_rate": 1.9085785848465875e-05, "loss": 0.5265, "step": 11649, "task_loss": 0.11836572736501694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5322054028511047, "epoch": 9.85, "learning_rate": 1.9082654978083906e-05, "loss": 0.466, "step": 11650, "task_loss": 1.4828978776931763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5156811475753784, "epoch": 9.85, "learning_rate": 1.907952410770194e-05, "loss": 0.4594, "step": 11651, "task_loss": 0.7437028288841248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44915640354156494, "epoch": 9.85, "learning_rate": 1.9076393237319977e-05, "loss": 0.4448, "step": 11652, "task_loss": 1.2757439613342285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5807198286056519, "epoch": 9.85, "learning_rate": 1.9073262366938012e-05, "loss": 0.5992, "step": 11653, "task_loss": 0.7748962640762329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3512771427631378, "epoch": 9.85, "learning_rate": 1.9070131496556044e-05, "loss": 0.4748, "step": 11654, "task_loss": 1.020880937576294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5941442251205444, "epoch": 9.85, "learning_rate": 1.906700062617408e-05, "loss": 0.5484, "step": 11655, "task_loss": 0.582425057888031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7293955087661743, "epoch": 9.85, "learning_rate": 1.906386975579211e-05, "loss": 0.5198, "step": 11656, "task_loss": 1.1509495973587036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6824651956558228, "epoch": 9.85, "learning_rate": 1.9060738885410146e-05, "loss": 0.6582, "step": 11657, "task_loss": 0.3971060812473297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5371817350387573, "epoch": 9.85, "learning_rate": 1.9057608015028178e-05, "loss": 0.5574, "step": 11658, "task_loss": 0.7185880541801453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41720718145370483, "epoch": 9.85, "learning_rate": 1.9054477144646213e-05, "loss": 0.4911, "step": 11659, "task_loss": 0.5411965847015381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8634131550788879, "epoch": 9.86, "learning_rate": 1.9051346274264245e-05, "loss": 0.5785, "step": 11660, "task_loss": 0.8641926050186157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.70477294921875, "epoch": 9.86, "learning_rate": 1.904821540388228e-05, "loss": 0.4723, "step": 11661, "task_loss": 1.615020513534546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4454159736633301, "epoch": 9.86, "learning_rate": 1.9045084533500315e-05, "loss": 0.4174, "step": 11662, "task_loss": 0.5658724308013916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7771435976028442, "epoch": 9.86, "learning_rate": 1.9041953663118347e-05, "loss": 0.5999, "step": 11663, "task_loss": 1.2596251964569092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47270432114601135, "epoch": 9.86, "learning_rate": 1.9038822792736382e-05, "loss": 0.4755, "step": 11664, "task_loss": 0.8126937747001648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5478535890579224, "epoch": 9.86, "learning_rate": 1.9035691922354414e-05, "loss": 0.5346, "step": 11665, "task_loss": 0.2695099413394928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47677361965179443, "epoch": 9.86, "learning_rate": 1.903256105197245e-05, "loss": 0.4375, "step": 11666, "task_loss": 0.9839802384376526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1930397748947144, "epoch": 9.86, "learning_rate": 1.9029430181590484e-05, "loss": 0.6807, "step": 11667, "task_loss": 1.2367212772369385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4396187663078308, "epoch": 9.86, "learning_rate": 1.902629931120852e-05, "loss": 0.6117, "step": 11668, "task_loss": 0.8203877210617065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6872769594192505, "epoch": 9.86, "learning_rate": 1.902316844082655e-05, "loss": 0.5027, "step": 11669, "task_loss": 0.7643510699272156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5503201484680176, "epoch": 9.86, "learning_rate": 1.9020037570444586e-05, "loss": 0.5617, "step": 11670, "task_loss": 0.4945705533027649 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3918520510196686, "epoch": 9.87, "learning_rate": 1.9016906700062618e-05, "loss": 0.6168, "step": 11671, "task_loss": 0.07031655311584473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32385683059692383, "epoch": 9.87, "learning_rate": 1.9013775829680653e-05, "loss": 0.4147, "step": 11672, "task_loss": 0.5764465928077698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5020001530647278, "epoch": 9.87, "learning_rate": 1.9010644959298685e-05, "loss": 0.4609, "step": 11673, "task_loss": 0.9338234663009644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7054738998413086, "epoch": 9.87, "learning_rate": 1.900751408891672e-05, "loss": 0.6544, "step": 11674, "task_loss": 1.5539976358413696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9579525589942932, "epoch": 9.87, "learning_rate": 1.9004383218534755e-05, "loss": 0.6606, "step": 11675, "task_loss": 1.654561161994934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4573095440864563, "epoch": 9.87, "learning_rate": 1.9001252348152787e-05, "loss": 0.6438, "step": 11676, "task_loss": 0.970325767993927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3295503258705139, "epoch": 9.87, "learning_rate": 1.8998121477770822e-05, "loss": 0.5734, "step": 11677, "task_loss": 0.06749876588582993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3778839111328125, "epoch": 9.87, "learning_rate": 1.8994990607388854e-05, "loss": 0.3907, "step": 11678, "task_loss": 0.5506734251976013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4969003200531006, "epoch": 9.87, "learning_rate": 1.899185973700689e-05, "loss": 0.55, "step": 11679, "task_loss": 0.9779006838798523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3716694712638855, "epoch": 9.87, "learning_rate": 1.898872886662492e-05, "loss": 0.5839, "step": 11680, "task_loss": 0.47546854615211487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5105975270271301, "epoch": 9.87, "learning_rate": 1.8985597996242956e-05, "loss": 0.5143, "step": 11681, "task_loss": 0.28621628880500793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3418923020362854, "epoch": 9.87, "learning_rate": 1.8982467125860988e-05, "loss": 0.4533, "step": 11682, "task_loss": 0.19986537098884583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5049116015434265, "epoch": 9.88, "learning_rate": 1.8979336255479023e-05, "loss": 0.5808, "step": 11683, "task_loss": 0.534247100353241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7671142816543579, "epoch": 9.88, "learning_rate": 1.8976205385097058e-05, "loss": 0.5522, "step": 11684, "task_loss": 0.9490468502044678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5991873741149902, "epoch": 9.88, "learning_rate": 1.8973074514715093e-05, "loss": 0.7618, "step": 11685, "task_loss": 1.162739634513855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.75941401720047, "epoch": 9.88, "learning_rate": 1.896994364433313e-05, "loss": 0.5908, "step": 11686, "task_loss": 0.27749398350715637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4548431634902954, "epoch": 9.88, "learning_rate": 1.896681277395116e-05, "loss": 0.4404, "step": 11687, "task_loss": 0.10448364168405533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5203955769538879, "epoch": 9.88, "learning_rate": 1.8963681903569195e-05, "loss": 0.5176, "step": 11688, "task_loss": 1.1743535995483398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7890534400939941, "epoch": 9.88, "learning_rate": 1.8960551033187227e-05, "loss": 0.6389, "step": 11689, "task_loss": 1.1439086198806763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19163915514945984, "epoch": 9.88, "learning_rate": 1.8957420162805262e-05, "loss": 0.4439, "step": 11690, "task_loss": 0.0605764165520668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4326263666152954, "epoch": 9.88, "learning_rate": 1.8954289292423294e-05, "loss": 0.5058, "step": 11691, "task_loss": 0.9461544752120972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3897644877433777, "epoch": 9.88, "learning_rate": 1.895115842204133e-05, "loss": 0.6968, "step": 11692, "task_loss": 0.9366523027420044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5151474475860596, "epoch": 9.88, "learning_rate": 1.894802755165936e-05, "loss": 0.6051, "step": 11693, "task_loss": 0.5469402074813843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5115368962287903, "epoch": 9.88, "learning_rate": 1.8944896681277396e-05, "loss": 0.4773, "step": 11694, "task_loss": 0.8160803318023682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7056788206100464, "epoch": 9.89, "learning_rate": 1.8941765810895428e-05, "loss": 0.594, "step": 11695, "task_loss": 0.3815004825592041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6320429444313049, "epoch": 9.89, "learning_rate": 1.8938634940513463e-05, "loss": 0.6134, "step": 11696, "task_loss": 0.41797196865081787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6054628491401672, "epoch": 9.89, "learning_rate": 1.8935504070131495e-05, "loss": 0.4766, "step": 11697, "task_loss": 0.2623986005783081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6233913898468018, "epoch": 9.89, "learning_rate": 1.893237319974953e-05, "loss": 0.6951, "step": 11698, "task_loss": 1.2117607593536377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6299286484718323, "epoch": 9.89, "learning_rate": 1.8929242329367565e-05, "loss": 0.5079, "step": 11699, "task_loss": 0.5708884596824646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48938143253326416, "epoch": 9.89, "learning_rate": 1.8926111458985597e-05, "loss": 0.5166, "step": 11700, "task_loss": 0.549912691116333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27777087688446045, "epoch": 9.89, "learning_rate": 1.8922980588603632e-05, "loss": 0.3503, "step": 11701, "task_loss": 1.0480424165725708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2659012973308563, "epoch": 9.89, "learning_rate": 1.8919849718221668e-05, "loss": 0.4421, "step": 11702, "task_loss": 0.4578627347946167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5705265998840332, "epoch": 9.89, "learning_rate": 1.8916718847839703e-05, "loss": 0.5251, "step": 11703, "task_loss": 0.7417893409729004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9203222393989563, "epoch": 9.89, "learning_rate": 1.8913587977457734e-05, "loss": 0.5766, "step": 11704, "task_loss": 1.2614158391952515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6709079146385193, "epoch": 9.89, "learning_rate": 1.891045710707577e-05, "loss": 0.6761, "step": 11705, "task_loss": 0.5382978320121765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3294762969017029, "epoch": 9.89, "learning_rate": 1.89073262366938e-05, "loss": 0.5423, "step": 11706, "task_loss": 0.09254828840494156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6487944722175598, "epoch": 9.9, "learning_rate": 1.8904195366311837e-05, "loss": 0.6204, "step": 11707, "task_loss": 0.739871621131897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6824483871459961, "epoch": 9.9, "learning_rate": 1.890106449592987e-05, "loss": 0.6486, "step": 11708, "task_loss": 1.5229136943817139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5428893566131592, "epoch": 9.9, "learning_rate": 1.8897933625547904e-05, "loss": 0.7813, "step": 11709, "task_loss": 0.5793831944465637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6293118000030518, "epoch": 9.9, "learning_rate": 1.8894802755165935e-05, "loss": 0.6736, "step": 11710, "task_loss": 0.8805944919586182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5093080997467041, "epoch": 9.9, "learning_rate": 1.889167188478397e-05, "loss": 0.5526, "step": 11711, "task_loss": 1.0464494228363037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.264046847820282, "epoch": 9.9, "learning_rate": 1.8888541014402006e-05, "loss": 0.4012, "step": 11712, "task_loss": 0.43050819635391235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6901537179946899, "epoch": 9.9, "learning_rate": 1.8885410144020037e-05, "loss": 0.7145, "step": 11713, "task_loss": 0.4162661135196686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6292136907577515, "epoch": 9.9, "learning_rate": 1.8882279273638073e-05, "loss": 0.5291, "step": 11714, "task_loss": 0.871176540851593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4616739749908447, "epoch": 9.9, "learning_rate": 1.8879148403256104e-05, "loss": 0.5026, "step": 11715, "task_loss": 0.728681743144989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5086719989776611, "epoch": 9.9, "learning_rate": 1.887601753287414e-05, "loss": 0.4575, "step": 11716, "task_loss": 0.6137264370918274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6728965044021606, "epoch": 9.9, "learning_rate": 1.887288666249217e-05, "loss": 0.7244, "step": 11717, "task_loss": 0.782421886920929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5934192538261414, "epoch": 9.9, "learning_rate": 1.8869755792110206e-05, "loss": 0.4614, "step": 11718, "task_loss": 1.1765084266662598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4801115393638611, "epoch": 9.91, "learning_rate": 1.886662492172824e-05, "loss": 0.6468, "step": 11719, "task_loss": 0.38565534353256226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8073398470878601, "epoch": 9.91, "learning_rate": 1.8863494051346277e-05, "loss": 0.5934, "step": 11720, "task_loss": 1.2019716501235962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5098705291748047, "epoch": 9.91, "learning_rate": 1.886036318096431e-05, "loss": 0.4972, "step": 11721, "task_loss": 1.0211752653121948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3643978536128998, "epoch": 9.91, "learning_rate": 1.8857232310582344e-05, "loss": 0.5838, "step": 11722, "task_loss": 1.4557913541793823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34722352027893066, "epoch": 9.91, "learning_rate": 1.885410144020038e-05, "loss": 0.6114, "step": 11723, "task_loss": 0.25257423520088196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3816722631454468, "epoch": 9.91, "learning_rate": 1.885097056981841e-05, "loss": 0.5181, "step": 11724, "task_loss": 0.8350476026535034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9122496843338013, "epoch": 9.91, "learning_rate": 1.8847839699436446e-05, "loss": 0.6005, "step": 11725, "task_loss": 0.45714980363845825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34653598070144653, "epoch": 9.91, "learning_rate": 1.8844708829054478e-05, "loss": 0.4141, "step": 11726, "task_loss": 0.2567834258079529 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8041822910308838, "epoch": 9.91, "learning_rate": 1.8841577958672513e-05, "loss": 0.5451, "step": 11727, "task_loss": 0.7415429949760437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5322411060333252, "epoch": 9.91, "learning_rate": 1.8838447088290545e-05, "loss": 0.612, "step": 11728, "task_loss": 1.3931331634521484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8586335778236389, "epoch": 9.91, "learning_rate": 1.883531621790858e-05, "loss": 0.6408, "step": 11729, "task_loss": 0.9485067129135132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5577413439750671, "epoch": 9.91, "learning_rate": 1.883218534752661e-05, "loss": 0.5159, "step": 11730, "task_loss": 1.0154595375061035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5434437394142151, "epoch": 9.92, "learning_rate": 1.8829054477144647e-05, "loss": 0.6714, "step": 11731, "task_loss": 0.738750159740448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6281569004058838, "epoch": 9.92, "learning_rate": 1.882592360676268e-05, "loss": 0.488, "step": 11732, "task_loss": 1.6868007183074951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7241640686988831, "epoch": 9.92, "learning_rate": 1.8822792736380714e-05, "loss": 0.6329, "step": 11733, "task_loss": 0.7312030792236328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.333202064037323, "epoch": 9.92, "learning_rate": 1.881966186599875e-05, "loss": 0.4861, "step": 11734, "task_loss": 0.4666426479816437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6310888528823853, "epoch": 9.92, "learning_rate": 1.881653099561678e-05, "loss": 0.549, "step": 11735, "task_loss": 0.4868190586566925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3222123086452484, "epoch": 9.92, "learning_rate": 1.8813400125234816e-05, "loss": 0.4741, "step": 11736, "task_loss": 0.07343541085720062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6123582720756531, "epoch": 9.92, "learning_rate": 1.881026925485285e-05, "loss": 0.6344, "step": 11737, "task_loss": 0.65635085105896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6079263687133789, "epoch": 9.92, "learning_rate": 1.8807138384470886e-05, "loss": 0.5723, "step": 11738, "task_loss": 0.3748495280742645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6696275472640991, "epoch": 9.92, "learning_rate": 1.8804007514088918e-05, "loss": 0.6847, "step": 11739, "task_loss": 0.8432466983795166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.363864541053772, "epoch": 9.92, "learning_rate": 1.8800876643706953e-05, "loss": 0.5845, "step": 11740, "task_loss": 0.46451112627983093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46929818391799927, "epoch": 9.92, "learning_rate": 1.8797745773324985e-05, "loss": 0.4333, "step": 11741, "task_loss": 0.5042869448661804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3879769444465637, "epoch": 9.93, "learning_rate": 1.879461490294302e-05, "loss": 0.4926, "step": 11742, "task_loss": 0.7122860550880432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3491581380367279, "epoch": 9.93, "learning_rate": 1.8791484032561052e-05, "loss": 0.5514, "step": 11743, "task_loss": 0.5618366003036499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6292937994003296, "epoch": 9.93, "learning_rate": 1.8788353162179087e-05, "loss": 0.5626, "step": 11744, "task_loss": 0.3628067970275879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9676641821861267, "epoch": 9.93, "learning_rate": 1.878522229179712e-05, "loss": 0.6863, "step": 11745, "task_loss": 1.1600873470306396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5187219381332397, "epoch": 9.93, "learning_rate": 1.8782091421415154e-05, "loss": 0.6204, "step": 11746, "task_loss": 1.3426660299301147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9243628978729248, "epoch": 9.93, "learning_rate": 1.8778960551033186e-05, "loss": 0.7063, "step": 11747, "task_loss": 0.722308874130249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27294445037841797, "epoch": 9.93, "learning_rate": 1.877582968065122e-05, "loss": 0.672, "step": 11748, "task_loss": 1.0869640111923218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7464796304702759, "epoch": 9.93, "learning_rate": 1.8772698810269256e-05, "loss": 0.4275, "step": 11749, "task_loss": 1.504287600517273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6274821162223816, "epoch": 9.93, "learning_rate": 1.8769567939887288e-05, "loss": 0.6056, "step": 11750, "task_loss": 1.0555065870285034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34639567136764526, "epoch": 9.93, "learning_rate": 1.8766437069505323e-05, "loss": 0.4342, "step": 11751, "task_loss": 1.2775434255599976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2621229588985443, "epoch": 9.93, "learning_rate": 1.8763306199123358e-05, "loss": 0.5517, "step": 11752, "task_loss": 0.031082237139344215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7995820045471191, "epoch": 9.93, "learning_rate": 1.8760175328741393e-05, "loss": 0.4929, "step": 11753, "task_loss": 0.8363316059112549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43546241521835327, "epoch": 9.94, "learning_rate": 1.8757044458359425e-05, "loss": 0.4623, "step": 11754, "task_loss": 0.23944571614265442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47052475810050964, "epoch": 9.94, "learning_rate": 1.875391358797746e-05, "loss": 0.4049, "step": 11755, "task_loss": 0.187409445643425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4224753677845001, "epoch": 9.94, "learning_rate": 1.8750782717595492e-05, "loss": 0.5641, "step": 11756, "task_loss": 0.5005813241004944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5649669766426086, "epoch": 9.94, "learning_rate": 1.8747651847213527e-05, "loss": 0.5373, "step": 11757, "task_loss": 0.2670878767967224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.554460346698761, "epoch": 9.94, "learning_rate": 1.874452097683156e-05, "loss": 0.6564, "step": 11758, "task_loss": 0.37876802682876587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5926956534385681, "epoch": 9.94, "learning_rate": 1.8741390106449594e-05, "loss": 0.4811, "step": 11759, "task_loss": 0.6162615418434143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8007664084434509, "epoch": 9.94, "learning_rate": 1.873825923606763e-05, "loss": 0.488, "step": 11760, "task_loss": 0.27505552768707275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.626805305480957, "epoch": 9.94, "learning_rate": 1.873512836568566e-05, "loss": 0.5527, "step": 11761, "task_loss": 0.32076558470726013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3736714720726013, "epoch": 9.94, "learning_rate": 1.8731997495303696e-05, "loss": 0.4673, "step": 11762, "task_loss": 0.22960704565048218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5448825359344482, "epoch": 9.94, "learning_rate": 1.8728866624921728e-05, "loss": 0.5179, "step": 11763, "task_loss": 0.4545641541481018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4515939950942993, "epoch": 9.94, "learning_rate": 1.8725735754539763e-05, "loss": 0.6344, "step": 11764, "task_loss": 0.3224949836730957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3991573452949524, "epoch": 9.94, "learning_rate": 1.8722604884157795e-05, "loss": 0.4559, "step": 11765, "task_loss": 0.45076531171798706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4022199511528015, "epoch": 9.95, "learning_rate": 1.871947401377583e-05, "loss": 0.4489, "step": 11766, "task_loss": 0.33653396368026733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4383222460746765, "epoch": 9.95, "learning_rate": 1.8716343143393862e-05, "loss": 0.4107, "step": 11767, "task_loss": 0.08910595625638962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4132869243621826, "epoch": 9.95, "learning_rate": 1.8713212273011897e-05, "loss": 0.5718, "step": 11768, "task_loss": 0.2924239933490753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3181550204753876, "epoch": 9.95, "learning_rate": 1.8710081402629932e-05, "loss": 0.4286, "step": 11769, "task_loss": 0.3659997880458832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7958353757858276, "epoch": 9.95, "learning_rate": 1.8706950532247968e-05, "loss": 0.5401, "step": 11770, "task_loss": 0.428263783454895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8368152379989624, "epoch": 9.95, "learning_rate": 1.8703819661866e-05, "loss": 0.5113, "step": 11771, "task_loss": 1.0283136367797852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40531042218208313, "epoch": 9.95, "learning_rate": 1.8700688791484034e-05, "loss": 0.4952, "step": 11772, "task_loss": 1.3702152967453003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28863608837127686, "epoch": 9.95, "learning_rate": 1.869755792110207e-05, "loss": 0.4825, "step": 11773, "task_loss": 0.1701551228761673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8097884058952332, "epoch": 9.95, "learning_rate": 1.86944270507201e-05, "loss": 0.4874, "step": 11774, "task_loss": 1.4247848987579346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37122994661331177, "epoch": 9.95, "learning_rate": 1.8691296180338137e-05, "loss": 0.5906, "step": 11775, "task_loss": 1.2842862606048584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2793921232223511, "epoch": 9.95, "learning_rate": 1.868816530995617e-05, "loss": 0.5722, "step": 11776, "task_loss": 0.9509588479995728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3793666958808899, "epoch": 9.95, "learning_rate": 1.8685034439574204e-05, "loss": 0.4974, "step": 11777, "task_loss": 0.9551975727081299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4627087116241455, "epoch": 9.96, "learning_rate": 1.8681903569192235e-05, "loss": 0.5431, "step": 11778, "task_loss": 0.7242524027824402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3095799684524536, "epoch": 9.96, "learning_rate": 1.867877269881027e-05, "loss": 0.4391, "step": 11779, "task_loss": 0.697620153427124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.568248987197876, "epoch": 9.96, "learning_rate": 1.8675641828428302e-05, "loss": 0.7444, "step": 11780, "task_loss": 0.638322114944458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6804888248443604, "epoch": 9.96, "learning_rate": 1.8672510958046337e-05, "loss": 0.4696, "step": 11781, "task_loss": 0.44510912895202637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41400861740112305, "epoch": 9.96, "learning_rate": 1.866938008766437e-05, "loss": 0.4956, "step": 11782, "task_loss": 0.5615767240524292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49559861421585083, "epoch": 9.96, "learning_rate": 1.8666249217282404e-05, "loss": 0.4938, "step": 11783, "task_loss": 0.7224570512771606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30016028881073, "epoch": 9.96, "learning_rate": 1.866311834690044e-05, "loss": 0.4353, "step": 11784, "task_loss": 0.392351359128952 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6252831220626831, "epoch": 9.96, "learning_rate": 1.865998747651847e-05, "loss": 0.8383, "step": 11785, "task_loss": 0.7282399535179138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44238409399986267, "epoch": 9.96, "learning_rate": 1.8656856606136507e-05, "loss": 0.4082, "step": 11786, "task_loss": 0.07588744163513184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45960357785224915, "epoch": 9.96, "learning_rate": 1.8653725735754542e-05, "loss": 0.4903, "step": 11787, "task_loss": 0.26694414019584656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22507253289222717, "epoch": 9.96, "learning_rate": 1.8650594865372577e-05, "loss": 0.4311, "step": 11788, "task_loss": 0.6326087117195129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44818174839019775, "epoch": 9.96, "learning_rate": 1.864746399499061e-05, "loss": 0.4934, "step": 11789, "task_loss": 0.11435481905937195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38110971450805664, "epoch": 9.97, "learning_rate": 1.8644333124608644e-05, "loss": 0.5092, "step": 11790, "task_loss": 0.1679525524377823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48966848850250244, "epoch": 9.97, "learning_rate": 1.8641202254226676e-05, "loss": 0.4204, "step": 11791, "task_loss": 0.5117335915565491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28338634967803955, "epoch": 9.97, "learning_rate": 1.863807138384471e-05, "loss": 0.3713, "step": 11792, "task_loss": 0.09126608073711395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6019378304481506, "epoch": 9.97, "learning_rate": 1.8634940513462743e-05, "loss": 0.4705, "step": 11793, "task_loss": 0.29883450269699097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35695675015449524, "epoch": 9.97, "learning_rate": 1.8631809643080778e-05, "loss": 0.4897, "step": 11794, "task_loss": 0.5929141044616699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6013560891151428, "epoch": 9.97, "learning_rate": 1.862867877269881e-05, "loss": 0.6065, "step": 11795, "task_loss": 1.1748204231262207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4136088490486145, "epoch": 9.97, "learning_rate": 1.8625547902316845e-05, "loss": 0.5201, "step": 11796, "task_loss": 0.7465836405754089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23197601735591888, "epoch": 9.97, "learning_rate": 1.862241703193488e-05, "loss": 0.4645, "step": 11797, "task_loss": 0.6430725455284119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.713988184928894, "epoch": 9.97, "learning_rate": 1.861928616155291e-05, "loss": 0.5508, "step": 11798, "task_loss": 0.6547066569328308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196396708488464, "epoch": 9.97, "learning_rate": 1.8616155291170947e-05, "loss": 0.5612, "step": 11799, "task_loss": 1.2492334842681885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3766554296016693, "epoch": 9.97, "learning_rate": 1.861302442078898e-05, "loss": 0.4772, "step": 11800, "task_loss": 0.1991298943758011 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5182839035987854, "epoch": 9.97, "learning_rate": 1.8609893550407014e-05, "loss": 0.4523, "step": 11801, "task_loss": 0.4825158417224884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3331873118877411, "epoch": 9.98, "learning_rate": 1.8606762680025046e-05, "loss": 0.3847, "step": 11802, "task_loss": 0.40961912274360657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5574594736099243, "epoch": 9.98, "learning_rate": 1.860363180964308e-05, "loss": 0.6035, "step": 11803, "task_loss": 0.8215177655220032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5494710803031921, "epoch": 9.98, "learning_rate": 1.8600500939261116e-05, "loss": 0.3717, "step": 11804, "task_loss": 0.5676952600479126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42646145820617676, "epoch": 9.98, "learning_rate": 1.859737006887915e-05, "loss": 0.6016, "step": 11805, "task_loss": 0.5025997757911682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6166901588439941, "epoch": 9.98, "learning_rate": 1.8594239198497183e-05, "loss": 0.456, "step": 11806, "task_loss": 0.3977431654930115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4038289785385132, "epoch": 9.98, "learning_rate": 1.8591108328115218e-05, "loss": 0.4677, "step": 11807, "task_loss": 0.40876325964927673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8574131727218628, "epoch": 9.98, "learning_rate": 1.858797745773325e-05, "loss": 0.5171, "step": 11808, "task_loss": 0.36882442235946655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4938904643058777, "epoch": 9.98, "learning_rate": 1.8584846587351285e-05, "loss": 0.4648, "step": 11809, "task_loss": 0.4225209653377533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.717519998550415, "epoch": 9.98, "learning_rate": 1.858171571696932e-05, "loss": 0.6975, "step": 11810, "task_loss": 0.996735692024231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.724993109703064, "epoch": 9.98, "learning_rate": 1.8578584846587352e-05, "loss": 0.5377, "step": 11811, "task_loss": 1.1715377569198608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.62744140625, "epoch": 9.98, "learning_rate": 1.8575453976205387e-05, "loss": 0.5704, "step": 11812, "task_loss": 0.4236169159412384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21306070685386658, "epoch": 9.99, "learning_rate": 1.857232310582342e-05, "loss": 0.4849, "step": 11813, "task_loss": 0.43288013339042664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5285851955413818, "epoch": 9.99, "learning_rate": 1.8569192235441454e-05, "loss": 0.4933, "step": 11814, "task_loss": 0.2645236551761627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33192771673202515, "epoch": 9.99, "learning_rate": 1.8566061365059486e-05, "loss": 0.5465, "step": 11815, "task_loss": 0.27421340346336365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6576056480407715, "epoch": 9.99, "learning_rate": 1.856293049467752e-05, "loss": 0.6887, "step": 11816, "task_loss": 1.8824236392974854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35208582878112793, "epoch": 9.99, "learning_rate": 1.8559799624295553e-05, "loss": 0.4135, "step": 11817, "task_loss": 0.6364741325378418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4812324047088623, "epoch": 9.99, "learning_rate": 1.8556668753913588e-05, "loss": 0.4902, "step": 11818, "task_loss": 0.6208165884017944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28622639179229736, "epoch": 9.99, "learning_rate": 1.8553537883531623e-05, "loss": 0.3529, "step": 11819, "task_loss": 0.023107778280973434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4559904932975769, "epoch": 9.99, "learning_rate": 1.8550407013149658e-05, "loss": 0.5327, "step": 11820, "task_loss": 1.1600409746170044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5496419668197632, "epoch": 9.99, "learning_rate": 1.854727614276769e-05, "loss": 0.4649, "step": 11821, "task_loss": 0.4436890184879303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5788514614105225, "epoch": 9.99, "learning_rate": 1.8544145272385725e-05, "loss": 0.4754, "step": 11822, "task_loss": 0.8245562314987183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.242612823843956, "epoch": 9.99, "learning_rate": 1.854101440200376e-05, "loss": 0.5781, "step": 11823, "task_loss": 0.061053112149238586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6755926012992859, "epoch": 9.99, "learning_rate": 1.8537883531621792e-05, "loss": 0.5082, "step": 11824, "task_loss": 1.1956509351730347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2775809168815613, "epoch": 10.0, "learning_rate": 1.8534752661239827e-05, "loss": 0.4623, "step": 11825, "task_loss": 0.0929676815867424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6108213067054749, "epoch": 10.0, "learning_rate": 1.853162179085786e-05, "loss": 0.5291, "step": 11826, "task_loss": 0.6461613178253174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39011693000793457, "epoch": 10.0, "learning_rate": 1.8528490920475894e-05, "loss": 0.5367, "step": 11827, "task_loss": 1.197256088256836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7376540303230286, "epoch": 10.0, "learning_rate": 1.8525360050093926e-05, "loss": 0.513, "step": 11828, "task_loss": 0.9222686886787415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5071805715560913, "epoch": 10.0, "learning_rate": 1.852222917971196e-05, "loss": 0.5587, "step": 11829, "task_loss": 0.86247318983078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39048558473587036, "epoch": 10.0, "learning_rate": 1.8519098309329993e-05, "loss": 0.5477, "step": 11830, "task_loss": 1.0019280910491943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.573614239692688, "epoch": 10.0, "learning_rate": 1.8515967438948028e-05, "loss": 1.0209, "step": 11831, "task_loss": 0.5039496421813965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0122331380844116, "epoch": 10.0, "learning_rate": 1.851283656856606e-05, "loss": 0.545, "step": 11832, "task_loss": 1.4548113346099854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6015266180038452, "epoch": 10.0, "learning_rate": 1.8509705698184095e-05, "loss": 0.5483, "step": 11833, "task_loss": 0.9096762537956238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.519451379776001, "epoch": 10.0, "learning_rate": 1.850657482780213e-05, "loss": 0.6555, "step": 11834, "task_loss": 0.8040266633033752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6428947448730469, "epoch": 10.0, "learning_rate": 1.8503443957420162e-05, "loss": 0.5459, "step": 11835, "task_loss": 1.3019704818725586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4378812611103058, "epoch": 10.01, "learning_rate": 1.8500313087038197e-05, "loss": 0.572, "step": 11836, "task_loss": 0.2593410611152649 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4471648931503296, "epoch": 10.01, "learning_rate": 1.8497182216656232e-05, "loss": 0.4662, "step": 11837, "task_loss": 0.3572514057159424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35012274980545044, "epoch": 10.01, "learning_rate": 1.8494051346274268e-05, "loss": 0.51, "step": 11838, "task_loss": 0.7921426892280579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3982032537460327, "epoch": 10.01, "learning_rate": 1.84909204758923e-05, "loss": 0.6048, "step": 11839, "task_loss": 0.4389876425266266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2256946861743927, "epoch": 10.01, "learning_rate": 1.8487789605510335e-05, "loss": 0.4633, "step": 11840, "task_loss": 0.39759141206741333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3144148290157318, "epoch": 10.01, "learning_rate": 1.8484658735128366e-05, "loss": 0.5446, "step": 11841, "task_loss": 0.3085957467556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4704146385192871, "epoch": 10.01, "learning_rate": 1.84815278647464e-05, "loss": 0.4586, "step": 11842, "task_loss": 0.20664894580841064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34629982709884644, "epoch": 10.01, "learning_rate": 1.8478396994364433e-05, "loss": 0.4897, "step": 11843, "task_loss": 0.6387989521026611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7815293669700623, "epoch": 10.01, "learning_rate": 1.847526612398247e-05, "loss": 0.6093, "step": 11844, "task_loss": 1.727777361869812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4518944025039673, "epoch": 10.01, "learning_rate": 1.84721352536005e-05, "loss": 0.4962, "step": 11845, "task_loss": 0.31168121099472046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8797907829284668, "epoch": 10.01, "learning_rate": 1.8469004383218535e-05, "loss": 0.6421, "step": 11846, "task_loss": 1.1763713359832764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36543023586273193, "epoch": 10.01, "learning_rate": 1.846587351283657e-05, "loss": 0.6842, "step": 11847, "task_loss": 0.04626606032252312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.943684458732605, "epoch": 10.02, "learning_rate": 1.8462742642454602e-05, "loss": 0.628, "step": 11848, "task_loss": 1.1483348608016968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.348634272813797, "epoch": 10.02, "learning_rate": 1.8459611772072637e-05, "loss": 0.5683, "step": 11849, "task_loss": 0.7357624173164368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4989299178123474, "epoch": 10.02, "learning_rate": 1.845648090169067e-05, "loss": 0.5808, "step": 11850, "task_loss": 1.176228642463684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5001161098480225, "epoch": 10.02, "learning_rate": 1.8453350031308704e-05, "loss": 0.4869, "step": 11851, "task_loss": 1.078476905822754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4927525222301483, "epoch": 10.02, "learning_rate": 1.8450219160926736e-05, "loss": 0.4741, "step": 11852, "task_loss": 0.7335951924324036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4639641046524048, "epoch": 10.02, "learning_rate": 1.844708829054477e-05, "loss": 0.5683, "step": 11853, "task_loss": 0.7101012468338013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27075132727622986, "epoch": 10.02, "learning_rate": 1.8443957420162807e-05, "loss": 0.4864, "step": 11854, "task_loss": 0.6216970086097717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2400934100151062, "epoch": 10.02, "learning_rate": 1.8440826549780842e-05, "loss": 0.3966, "step": 11855, "task_loss": 0.5425400137901306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5233400464057922, "epoch": 10.02, "learning_rate": 1.8437695679398874e-05, "loss": 0.4811, "step": 11856, "task_loss": 0.8149230480194092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2610895335674286, "epoch": 10.02, "learning_rate": 1.843456480901691e-05, "loss": 0.3849, "step": 11857, "task_loss": 0.24298550188541412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4431195557117462, "epoch": 10.02, "learning_rate": 1.8431433938634944e-05, "loss": 0.4412, "step": 11858, "task_loss": 1.0330628156661987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6324293613433838, "epoch": 10.02, "learning_rate": 1.8428303068252976e-05, "loss": 0.5947, "step": 11859, "task_loss": 0.7978057265281677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2642650604248047, "epoch": 10.03, "learning_rate": 1.842517219787101e-05, "loss": 0.4966, "step": 11860, "task_loss": 0.15288308262825012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5137324333190918, "epoch": 10.03, "learning_rate": 1.8422041327489043e-05, "loss": 0.3964, "step": 11861, "task_loss": 0.2097819596529007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5943300724029541, "epoch": 10.03, "learning_rate": 1.8418910457107078e-05, "loss": 0.5667, "step": 11862, "task_loss": 0.66886967420578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4728262424468994, "epoch": 10.03, "learning_rate": 1.841577958672511e-05, "loss": 0.6727, "step": 11863, "task_loss": 0.33133354783058167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7342443466186523, "epoch": 10.03, "learning_rate": 1.8412648716343145e-05, "loss": 0.4155, "step": 11864, "task_loss": 0.9448426365852356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38220518827438354, "epoch": 10.03, "learning_rate": 1.8409517845961176e-05, "loss": 0.4854, "step": 11865, "task_loss": 0.14742380380630493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.522541344165802, "epoch": 10.03, "learning_rate": 1.840638697557921e-05, "loss": 0.4112, "step": 11866, "task_loss": 0.6054068803787231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3673941195011139, "epoch": 10.03, "learning_rate": 1.8403256105197243e-05, "loss": 0.4267, "step": 11867, "task_loss": 1.269416332244873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8133429884910583, "epoch": 10.03, "learning_rate": 1.840012523481528e-05, "loss": 0.6918, "step": 11868, "task_loss": 0.9615734815597534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4054149389266968, "epoch": 10.03, "learning_rate": 1.839699436443331e-05, "loss": 0.4636, "step": 11869, "task_loss": 0.043320443481206894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5212281346321106, "epoch": 10.03, "learning_rate": 1.8393863494051346e-05, "loss": 0.5379, "step": 11870, "task_loss": 1.0919591188430786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3338669240474701, "epoch": 10.03, "learning_rate": 1.839073262366938e-05, "loss": 0.5319, "step": 11871, "task_loss": 1.1334413290023804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4621918797492981, "epoch": 10.04, "learning_rate": 1.8387601753287416e-05, "loss": 0.5888, "step": 11872, "task_loss": 1.0532444715499878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5453698635101318, "epoch": 10.04, "learning_rate": 1.838447088290545e-05, "loss": 0.4478, "step": 11873, "task_loss": 0.7809507846832275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4258708953857422, "epoch": 10.04, "learning_rate": 1.8381340012523483e-05, "loss": 0.5177, "step": 11874, "task_loss": 0.13625706732273102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6970002055168152, "epoch": 10.04, "learning_rate": 1.8378209142141518e-05, "loss": 0.6418, "step": 11875, "task_loss": 1.2745261192321777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4338301420211792, "epoch": 10.04, "learning_rate": 1.837507827175955e-05, "loss": 0.5058, "step": 11876, "task_loss": 0.4660428464412689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3863411247730255, "epoch": 10.04, "learning_rate": 1.8371947401377585e-05, "loss": 0.5034, "step": 11877, "task_loss": 0.46990084648132324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6842662692070007, "epoch": 10.04, "learning_rate": 1.8368816530995617e-05, "loss": 0.6167, "step": 11878, "task_loss": 1.072363257408142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6997926235198975, "epoch": 10.04, "learning_rate": 1.8365685660613652e-05, "loss": 0.6171, "step": 11879, "task_loss": 1.1311322450637817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5067988038063049, "epoch": 10.04, "learning_rate": 1.8362554790231684e-05, "loss": 0.4079, "step": 11880, "task_loss": 0.27674242854118347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34035414457321167, "epoch": 10.04, "learning_rate": 1.835942391984972e-05, "loss": 0.4632, "step": 11881, "task_loss": 0.18859435617923737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5251001119613647, "epoch": 10.04, "learning_rate": 1.8356293049467754e-05, "loss": 0.4955, "step": 11882, "task_loss": 0.5429094433784485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38952159881591797, "epoch": 10.04, "learning_rate": 1.8353162179085786e-05, "loss": 0.5652, "step": 11883, "task_loss": 0.38006290793418884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5447397232055664, "epoch": 10.05, "learning_rate": 1.835003130870382e-05, "loss": 0.5162, "step": 11884, "task_loss": 0.9888135194778442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4869982600212097, "epoch": 10.05, "learning_rate": 1.8346900438321853e-05, "loss": 0.5098, "step": 11885, "task_loss": 0.6297141909599304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4317600131034851, "epoch": 10.05, "learning_rate": 1.8343769567939888e-05, "loss": 0.5166, "step": 11886, "task_loss": 0.3621313273906708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38352957367897034, "epoch": 10.05, "learning_rate": 1.834063869755792e-05, "loss": 0.434, "step": 11887, "task_loss": 0.33340156078338623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5041660070419312, "epoch": 10.05, "learning_rate": 1.8337507827175955e-05, "loss": 0.5573, "step": 11888, "task_loss": 0.3025062680244446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5290478467941284, "epoch": 10.05, "learning_rate": 1.833437695679399e-05, "loss": 0.5301, "step": 11889, "task_loss": 0.6290261149406433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43631699681282043, "epoch": 10.05, "learning_rate": 1.8331246086412025e-05, "loss": 0.4709, "step": 11890, "task_loss": 0.9419873952865601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37002548575401306, "epoch": 10.05, "learning_rate": 1.8328115216030057e-05, "loss": 0.5224, "step": 11891, "task_loss": 1.400595784187317 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35538381338119507, "epoch": 10.05, "learning_rate": 1.8324984345648092e-05, "loss": 0.5548, "step": 11892, "task_loss": 0.5066011548042297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.330211877822876, "epoch": 10.05, "learning_rate": 1.8321853475266124e-05, "loss": 0.4801, "step": 11893, "task_loss": 0.6809589266777039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7314883470535278, "epoch": 10.05, "learning_rate": 1.831872260488416e-05, "loss": 0.6155, "step": 11894, "task_loss": 1.7068151235580444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4618019461631775, "epoch": 10.05, "learning_rate": 1.8315591734502194e-05, "loss": 0.5994, "step": 11895, "task_loss": 0.49507543444633484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5336235761642456, "epoch": 10.06, "learning_rate": 1.8312460864120226e-05, "loss": 0.4046, "step": 11896, "task_loss": 1.6036204099655151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8582067489624023, "epoch": 10.06, "learning_rate": 1.830932999373826e-05, "loss": 0.518, "step": 11897, "task_loss": 0.7478420734405518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7184785604476929, "epoch": 10.06, "learning_rate": 1.8306199123356293e-05, "loss": 0.6585, "step": 11898, "task_loss": 0.6461512446403503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2741459310054779, "epoch": 10.06, "learning_rate": 1.8303068252974328e-05, "loss": 0.4307, "step": 11899, "task_loss": 0.22900283336639404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42498379945755005, "epoch": 10.06, "learning_rate": 1.829993738259236e-05, "loss": 0.3986, "step": 11900, "task_loss": 0.29451608657836914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6632781028747559, "epoch": 10.06, "learning_rate": 1.8296806512210395e-05, "loss": 0.5895, "step": 11901, "task_loss": 1.779968500137329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5856399536132812, "epoch": 10.06, "learning_rate": 1.8293675641828427e-05, "loss": 0.5981, "step": 11902, "task_loss": 1.1742357015609741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4982651472091675, "epoch": 10.06, "learning_rate": 1.8290544771446462e-05, "loss": 0.4392, "step": 11903, "task_loss": 0.7314689755439758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43142369389533997, "epoch": 10.06, "learning_rate": 1.8287413901064497e-05, "loss": 0.5093, "step": 11904, "task_loss": 0.2720101773738861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5521692037582397, "epoch": 10.06, "learning_rate": 1.8284283030682532e-05, "loss": 0.4817, "step": 11905, "task_loss": 0.2980252802371979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8019011616706848, "epoch": 10.06, "learning_rate": 1.8281152160300564e-05, "loss": 0.5841, "step": 11906, "task_loss": 1.507855772972107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6010299324989319, "epoch": 10.07, "learning_rate": 1.82780212899186e-05, "loss": 0.6771, "step": 11907, "task_loss": 0.8570787310600281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5556377172470093, "epoch": 10.07, "learning_rate": 1.8274890419536635e-05, "loss": 0.5176, "step": 11908, "task_loss": 1.381196141242981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38272637128829956, "epoch": 10.07, "learning_rate": 1.8271759549154666e-05, "loss": 0.4155, "step": 11909, "task_loss": 0.8253744840621948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6307418346405029, "epoch": 10.07, "learning_rate": 1.82686286787727e-05, "loss": 0.4544, "step": 11910, "task_loss": 0.6507962942123413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7920186519622803, "epoch": 10.07, "learning_rate": 1.8265497808390733e-05, "loss": 0.5492, "step": 11911, "task_loss": 0.17453062534332275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42062264680862427, "epoch": 10.07, "learning_rate": 1.826236693800877e-05, "loss": 0.5439, "step": 11912, "task_loss": 0.43165865540504456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45185399055480957, "epoch": 10.07, "learning_rate": 1.82592360676268e-05, "loss": 0.5481, "step": 11913, "task_loss": 0.866303563117981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2518584430217743, "epoch": 10.07, "learning_rate": 1.8256105197244835e-05, "loss": 0.4661, "step": 11914, "task_loss": 0.42644309997558594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39169013500213623, "epoch": 10.07, "learning_rate": 1.8252974326862867e-05, "loss": 0.6316, "step": 11915, "task_loss": 0.3899060785770416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5305859446525574, "epoch": 10.07, "learning_rate": 1.8249843456480902e-05, "loss": 0.4936, "step": 11916, "task_loss": 1.2438266277313232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5410599112510681, "epoch": 10.07, "learning_rate": 1.8246712586098934e-05, "loss": 0.472, "step": 11917, "task_loss": 0.7229633331298828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3879723846912384, "epoch": 10.07, "learning_rate": 1.824358171571697e-05, "loss": 0.4581, "step": 11918, "task_loss": 0.7790693044662476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3686602711677551, "epoch": 10.08, "learning_rate": 1.8240450845335004e-05, "loss": 0.4496, "step": 11919, "task_loss": 0.3374233543872833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5402283668518066, "epoch": 10.08, "learning_rate": 1.8237319974953036e-05, "loss": 0.5897, "step": 11920, "task_loss": 0.7575618028640747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6014347076416016, "epoch": 10.08, "learning_rate": 1.823418910457107e-05, "loss": 0.5681, "step": 11921, "task_loss": 0.5591609477996826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36061975359916687, "epoch": 10.08, "learning_rate": 1.8231058234189107e-05, "loss": 0.4673, "step": 11922, "task_loss": 0.61372309923172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33913707733154297, "epoch": 10.08, "learning_rate": 1.8227927363807142e-05, "loss": 0.4101, "step": 11923, "task_loss": 0.7801917791366577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4929392337799072, "epoch": 10.08, "learning_rate": 1.8224796493425174e-05, "loss": 0.4061, "step": 11924, "task_loss": 0.2593909502029419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49953585863113403, "epoch": 10.08, "learning_rate": 1.822166562304321e-05, "loss": 0.522, "step": 11925, "task_loss": 0.22385400533676147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3363998830318451, "epoch": 10.08, "learning_rate": 1.821853475266124e-05, "loss": 0.4972, "step": 11926, "task_loss": 0.4675111472606659 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5529756546020508, "epoch": 10.08, "learning_rate": 1.8215403882279276e-05, "loss": 0.6389, "step": 11927, "task_loss": 0.6886316537857056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25953251123428345, "epoch": 10.08, "learning_rate": 1.8212273011897307e-05, "loss": 0.5668, "step": 11928, "task_loss": 0.3315168023109436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2634521424770355, "epoch": 10.08, "learning_rate": 1.8209142141515343e-05, "loss": 0.3497, "step": 11929, "task_loss": 0.3371821939945221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.480176717042923, "epoch": 10.08, "learning_rate": 1.8206011271133374e-05, "loss": 0.4868, "step": 11930, "task_loss": 0.6029389500617981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3795178532600403, "epoch": 10.09, "learning_rate": 1.820288040075141e-05, "loss": 0.4306, "step": 11931, "task_loss": 1.2051299810409546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7666014432907104, "epoch": 10.09, "learning_rate": 1.8199749530369445e-05, "loss": 0.5767, "step": 11932, "task_loss": 0.8345617651939392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196911096572876, "epoch": 10.09, "learning_rate": 1.8196618659987477e-05, "loss": 0.5714, "step": 11933, "task_loss": 0.5302334427833557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.434179425239563, "epoch": 10.09, "learning_rate": 1.819348778960551e-05, "loss": 0.5751, "step": 11934, "task_loss": 1.4489260911941528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6102753281593323, "epoch": 10.09, "learning_rate": 1.8190356919223543e-05, "loss": 0.5324, "step": 11935, "task_loss": 0.3866889178752899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5891766548156738, "epoch": 10.09, "learning_rate": 1.818722604884158e-05, "loss": 0.5013, "step": 11936, "task_loss": 1.1297861337661743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6296302080154419, "epoch": 10.09, "learning_rate": 1.818409517845961e-05, "loss": 0.4744, "step": 11937, "task_loss": 1.045244812965393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20342552661895752, "epoch": 10.09, "learning_rate": 1.8180964308077646e-05, "loss": 0.4396, "step": 11938, "task_loss": 0.09281840175390244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26525214314460754, "epoch": 10.09, "learning_rate": 1.817783343769568e-05, "loss": 0.3553, "step": 11939, "task_loss": 0.20385567843914032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48645997047424316, "epoch": 10.09, "learning_rate": 1.8174702567313716e-05, "loss": 0.5393, "step": 11940, "task_loss": 0.9254416823387146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49700677394866943, "epoch": 10.09, "learning_rate": 1.8171571696931748e-05, "loss": 0.5649, "step": 11941, "task_loss": 1.1869666576385498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3799194395542145, "epoch": 10.09, "learning_rate": 1.8168440826549783e-05, "loss": 0.5607, "step": 11942, "task_loss": 0.8608298301696777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39190685749053955, "epoch": 10.1, "learning_rate": 1.8165309956167815e-05, "loss": 0.544, "step": 11943, "task_loss": 0.19806435704231262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3603771924972534, "epoch": 10.1, "learning_rate": 1.816217908578585e-05, "loss": 0.616, "step": 11944, "task_loss": 0.13370120525360107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4323745369911194, "epoch": 10.1, "learning_rate": 1.8159048215403885e-05, "loss": 0.4827, "step": 11945, "task_loss": 0.5119966268539429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.714235782623291, "epoch": 10.1, "learning_rate": 1.8155917345021917e-05, "loss": 0.6007, "step": 11946, "task_loss": 1.5682661533355713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5041913986206055, "epoch": 10.1, "learning_rate": 1.8152786474639952e-05, "loss": 0.5199, "step": 11947, "task_loss": 0.6148562431335449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5477074384689331, "epoch": 10.1, "learning_rate": 1.8149655604257984e-05, "loss": 0.5768, "step": 11948, "task_loss": 0.3812944293022156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.553023099899292, "epoch": 10.1, "learning_rate": 1.814652473387602e-05, "loss": 0.5344, "step": 11949, "task_loss": 0.6007188558578491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7481228113174438, "epoch": 10.1, "learning_rate": 1.814339386349405e-05, "loss": 0.5926, "step": 11950, "task_loss": 1.2068064212799072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6951996088027954, "epoch": 10.1, "learning_rate": 1.8140262993112086e-05, "loss": 0.5857, "step": 11951, "task_loss": 1.8991118669509888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4695318639278412, "epoch": 10.1, "learning_rate": 1.8137132122730118e-05, "loss": 0.5046, "step": 11952, "task_loss": 0.2603614628314972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21509301662445068, "epoch": 10.1, "learning_rate": 1.8134001252348153e-05, "loss": 0.4821, "step": 11953, "task_loss": 0.27137383818626404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8246045708656311, "epoch": 10.1, "learning_rate": 1.8130870381966185e-05, "loss": 0.6164, "step": 11954, "task_loss": 0.9676323533058167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6132639646530151, "epoch": 10.11, "learning_rate": 1.812773951158422e-05, "loss": 0.4439, "step": 11955, "task_loss": 1.0401791334152222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8934365510940552, "epoch": 10.11, "learning_rate": 1.8124608641202255e-05, "loss": 0.6176, "step": 11956, "task_loss": 0.4012407958507538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33261188864707947, "epoch": 10.11, "learning_rate": 1.812147777082029e-05, "loss": 0.4346, "step": 11957, "task_loss": 0.345401406288147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6085391044616699, "epoch": 10.11, "learning_rate": 1.8118346900438325e-05, "loss": 0.6793, "step": 11958, "task_loss": 2.874612808227539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7885323762893677, "epoch": 10.11, "learning_rate": 1.8115216030056357e-05, "loss": 0.5288, "step": 11959, "task_loss": 0.784598708152771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6752193570137024, "epoch": 10.11, "learning_rate": 1.8112085159674392e-05, "loss": 0.4734, "step": 11960, "task_loss": 0.5668706893920898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5953697562217712, "epoch": 10.11, "learning_rate": 1.8108954289292424e-05, "loss": 0.6645, "step": 11961, "task_loss": 0.49043387174606323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5394877195358276, "epoch": 10.11, "learning_rate": 1.810582341891046e-05, "loss": 0.4972, "step": 11962, "task_loss": 1.133232593536377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4368830621242523, "epoch": 10.11, "learning_rate": 1.810269254852849e-05, "loss": 0.4705, "step": 11963, "task_loss": 0.8905189037322998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28562018275260925, "epoch": 10.11, "learning_rate": 1.8099561678146526e-05, "loss": 0.4671, "step": 11964, "task_loss": 0.6866825819015503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4449443221092224, "epoch": 10.11, "learning_rate": 1.8096430807764558e-05, "loss": 0.4076, "step": 11965, "task_loss": 0.3457188308238983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7404114603996277, "epoch": 10.11, "learning_rate": 1.8093299937382593e-05, "loss": 0.4487, "step": 11966, "task_loss": 0.09735433012247086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44974032044410706, "epoch": 10.12, "learning_rate": 1.8090169067000625e-05, "loss": 0.4208, "step": 11967, "task_loss": 1.4680790901184082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2576064467430115, "epoch": 10.12, "learning_rate": 1.808703819661866e-05, "loss": 0.3902, "step": 11968, "task_loss": 0.49863123893737793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5658372640609741, "epoch": 10.12, "learning_rate": 1.8083907326236695e-05, "loss": 0.5053, "step": 11969, "task_loss": 0.07551191747188568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3111751973628998, "epoch": 10.12, "learning_rate": 1.8080776455854727e-05, "loss": 0.552, "step": 11970, "task_loss": 0.37528494000434875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2674192786216736, "epoch": 10.12, "learning_rate": 1.8077645585472762e-05, "loss": 0.4566, "step": 11971, "task_loss": 0.4724051058292389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32935386896133423, "epoch": 10.12, "learning_rate": 1.8074514715090797e-05, "loss": 0.3986, "step": 11972, "task_loss": 0.5914874076843262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3320724368095398, "epoch": 10.12, "learning_rate": 1.807138384470883e-05, "loss": 0.5075, "step": 11973, "task_loss": 0.6343813538551331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6177480220794678, "epoch": 10.12, "learning_rate": 1.8068252974326864e-05, "loss": 0.5055, "step": 11974, "task_loss": 0.6024479269981384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5748840570449829, "epoch": 10.12, "learning_rate": 1.80651221039449e-05, "loss": 0.5444, "step": 11975, "task_loss": 0.3186684548854828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32908183336257935, "epoch": 10.12, "learning_rate": 1.806199123356293e-05, "loss": 0.418, "step": 11976, "task_loss": 0.2868935167789459 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7213194370269775, "epoch": 10.12, "learning_rate": 1.8058860363180966e-05, "loss": 0.6759, "step": 11977, "task_loss": 0.5693639516830444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4019198417663574, "epoch": 10.13, "learning_rate": 1.8055729492798998e-05, "loss": 0.5637, "step": 11978, "task_loss": 0.4135168194770813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5542411804199219, "epoch": 10.13, "learning_rate": 1.8052598622417033e-05, "loss": 0.5137, "step": 11979, "task_loss": 0.2759263813495636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5234736204147339, "epoch": 10.13, "learning_rate": 1.8049467752035065e-05, "loss": 0.5621, "step": 11980, "task_loss": 0.30167049169540405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29652249813079834, "epoch": 10.13, "learning_rate": 1.80463368816531e-05, "loss": 0.4804, "step": 11981, "task_loss": 0.3228597640991211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4809386134147644, "epoch": 10.13, "learning_rate": 1.8043206011271135e-05, "loss": 0.5172, "step": 11982, "task_loss": 0.7798720002174377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22345766425132751, "epoch": 10.13, "learning_rate": 1.8040075140889167e-05, "loss": 0.4174, "step": 11983, "task_loss": 0.2587651312351227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45418813824653625, "epoch": 10.13, "learning_rate": 1.8036944270507202e-05, "loss": 0.4643, "step": 11984, "task_loss": 0.32602351903915405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7528649568557739, "epoch": 10.13, "learning_rate": 1.8033813400125234e-05, "loss": 0.6221, "step": 11985, "task_loss": 0.9729608297348022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.519275426864624, "epoch": 10.13, "learning_rate": 1.803068252974327e-05, "loss": 0.5816, "step": 11986, "task_loss": 0.8167797923088074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6711570024490356, "epoch": 10.13, "learning_rate": 1.80275516593613e-05, "loss": 0.5881, "step": 11987, "task_loss": 0.586622416973114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47273746132850647, "epoch": 10.13, "learning_rate": 1.8024420788979336e-05, "loss": 0.5917, "step": 11988, "task_loss": 0.7469164729118347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49289005994796753, "epoch": 10.13, "learning_rate": 1.802128991859737e-05, "loss": 0.5231, "step": 11989, "task_loss": 0.2942104935646057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6479471325874329, "epoch": 10.14, "learning_rate": 1.8018159048215407e-05, "loss": 0.72, "step": 11990, "task_loss": 0.3644965887069702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6494221091270447, "epoch": 10.14, "learning_rate": 1.801502817783344e-05, "loss": 0.553, "step": 11991, "task_loss": 0.7293792366981506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40172338485717773, "epoch": 10.14, "learning_rate": 1.8011897307451474e-05, "loss": 0.5032, "step": 11992, "task_loss": 0.7648729681968689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7259783744812012, "epoch": 10.14, "learning_rate": 1.800876643706951e-05, "loss": 0.6539, "step": 11993, "task_loss": 0.645050048828125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.479883074760437, "epoch": 10.14, "learning_rate": 1.800563556668754e-05, "loss": 0.4979, "step": 11994, "task_loss": 1.3832471370697021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3185862898826599, "epoch": 10.14, "learning_rate": 1.8002504696305576e-05, "loss": 0.38, "step": 11995, "task_loss": 0.20617687702178955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7085285186767578, "epoch": 10.14, "learning_rate": 1.7999373825923607e-05, "loss": 0.5246, "step": 11996, "task_loss": 1.2935453653335571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34835192561149597, "epoch": 10.14, "learning_rate": 1.7996242955541643e-05, "loss": 0.4907, "step": 11997, "task_loss": 0.23376747965812683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37399110198020935, "epoch": 10.14, "learning_rate": 1.7993112085159674e-05, "loss": 0.3769, "step": 11998, "task_loss": 0.45206204056739807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6348373889923096, "epoch": 10.14, "learning_rate": 1.798998121477771e-05, "loss": 0.6645, "step": 11999, "task_loss": 0.6434426307678223 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4983244240283966, "epoch": 10.14, "learning_rate": 1.798685034439574e-05, "loss": 0.425, "step": 12000, "task_loss": 0.6595286726951599 }, { "epoch": 10.14, "eval_accuracy": 0.905940594059406, "eval_loss": 0.36149701476097107, "eval_runtime": 206.8852, "eval_samples_per_second": 122.048, "eval_steps_per_second": 0.957, "step": 12000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4854140281677246, "epoch": 10.14, "learning_rate": 1.7983719474013777e-05, "loss": 0.6094, "step": 12001, "task_loss": 1.6257245540618896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47899723052978516, "epoch": 10.15, "learning_rate": 1.798058860363181e-05, "loss": 0.5563, "step": 12002, "task_loss": 0.7581605315208435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3620864748954773, "epoch": 10.15, "learning_rate": 1.7977457733249844e-05, "loss": 0.4747, "step": 12003, "task_loss": 0.41931480169296265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3591528534889221, "epoch": 10.15, "learning_rate": 1.7974326862867875e-05, "loss": 0.3974, "step": 12004, "task_loss": 0.3311234414577484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41835927963256836, "epoch": 10.15, "learning_rate": 1.797119599248591e-05, "loss": 0.4301, "step": 12005, "task_loss": 0.20978368818759918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3658050298690796, "epoch": 10.15, "learning_rate": 1.7968065122103946e-05, "loss": 0.5229, "step": 12006, "task_loss": 0.30081966519355774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47107937932014465, "epoch": 10.15, "learning_rate": 1.796493425172198e-05, "loss": 0.5092, "step": 12007, "task_loss": 0.4290759861469269 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27061188220977783, "epoch": 10.15, "learning_rate": 1.7961803381340016e-05, "loss": 0.421, "step": 12008, "task_loss": 0.7357758283615112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5408056378364563, "epoch": 10.15, "learning_rate": 1.7958672510958048e-05, "loss": 0.5592, "step": 12009, "task_loss": 0.872681736946106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5029695630073547, "epoch": 10.15, "learning_rate": 1.7955541640576083e-05, "loss": 0.5193, "step": 12010, "task_loss": 0.6903013586997986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3501845896244049, "epoch": 10.15, "learning_rate": 1.7952410770194115e-05, "loss": 0.4613, "step": 12011, "task_loss": 1.014859914779663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2505016326904297, "epoch": 10.15, "learning_rate": 1.794927989981215e-05, "loss": 0.3909, "step": 12012, "task_loss": 0.8884826302528381 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.615286111831665, "epoch": 10.15, "learning_rate": 1.794614902943018e-05, "loss": 0.4775, "step": 12013, "task_loss": 0.40324491262435913 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6798600554466248, "epoch": 10.16, "learning_rate": 1.7943018159048217e-05, "loss": 0.5216, "step": 12014, "task_loss": 1.2024500370025635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5951483249664307, "epoch": 10.16, "learning_rate": 1.793988728866625e-05, "loss": 0.4995, "step": 12015, "task_loss": 0.8973203897476196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6922138333320618, "epoch": 10.16, "learning_rate": 1.7936756418284284e-05, "loss": 0.6149, "step": 12016, "task_loss": 1.4704807996749878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4563525319099426, "epoch": 10.16, "learning_rate": 1.793362554790232e-05, "loss": 0.5013, "step": 12017, "task_loss": 0.15977802872657776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31994450092315674, "epoch": 10.16, "learning_rate": 1.793049467752035e-05, "loss": 0.6545, "step": 12018, "task_loss": 0.2737879455089569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5105599164962769, "epoch": 10.16, "learning_rate": 1.7927363807138386e-05, "loss": 0.3714, "step": 12019, "task_loss": 0.7855756282806396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49507635831832886, "epoch": 10.16, "learning_rate": 1.7924232936756418e-05, "loss": 0.3515, "step": 12020, "task_loss": 0.16255690157413483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3792104125022888, "epoch": 10.16, "learning_rate": 1.7921102066374453e-05, "loss": 0.4951, "step": 12021, "task_loss": 0.4633488655090332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6868609189987183, "epoch": 10.16, "learning_rate": 1.7917971195992485e-05, "loss": 0.5467, "step": 12022, "task_loss": 0.35662999749183655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5744777321815491, "epoch": 10.16, "learning_rate": 1.791484032561052e-05, "loss": 0.6039, "step": 12023, "task_loss": 0.8216590881347656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7112334966659546, "epoch": 10.16, "learning_rate": 1.7911709455228555e-05, "loss": 0.5968, "step": 12024, "task_loss": 0.9066623449325562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8560420274734497, "epoch": 10.16, "learning_rate": 1.790857858484659e-05, "loss": 0.621, "step": 12025, "task_loss": 0.7734057903289795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7827113270759583, "epoch": 10.17, "learning_rate": 1.7905447714464622e-05, "loss": 0.6318, "step": 12026, "task_loss": 1.2292592525482178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4711020886898041, "epoch": 10.17, "learning_rate": 1.7902316844082657e-05, "loss": 0.6176, "step": 12027, "task_loss": 0.6844468712806702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22153571248054504, "epoch": 10.17, "learning_rate": 1.789918597370069e-05, "loss": 0.4698, "step": 12028, "task_loss": 0.3284510374069214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4110615849494934, "epoch": 10.17, "learning_rate": 1.7896055103318724e-05, "loss": 0.5419, "step": 12029, "task_loss": 0.055999573320150375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7471761107444763, "epoch": 10.17, "learning_rate": 1.789292423293676e-05, "loss": 0.528, "step": 12030, "task_loss": 1.1426992416381836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5584474802017212, "epoch": 10.17, "learning_rate": 1.788979336255479e-05, "loss": 0.5503, "step": 12031, "task_loss": 0.12276840209960938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3508461117744446, "epoch": 10.17, "learning_rate": 1.7886662492172826e-05, "loss": 0.4472, "step": 12032, "task_loss": 0.22514846920967102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9211220741271973, "epoch": 10.17, "learning_rate": 1.7883531621790858e-05, "loss": 0.7544, "step": 12033, "task_loss": 0.9317436218261719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6198623180389404, "epoch": 10.17, "learning_rate": 1.7880400751408893e-05, "loss": 0.491, "step": 12034, "task_loss": 0.6504527926445007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3975866138935089, "epoch": 10.17, "learning_rate": 1.7877269881026925e-05, "loss": 0.5965, "step": 12035, "task_loss": 0.37547704577445984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5710036754608154, "epoch": 10.17, "learning_rate": 1.787413901064496e-05, "loss": 0.5003, "step": 12036, "task_loss": 0.5649453401565552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2373313456773758, "epoch": 10.17, "learning_rate": 1.7871008140262992e-05, "loss": 0.3649, "step": 12037, "task_loss": 0.3479361832141876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4211253523826599, "epoch": 10.18, "learning_rate": 1.7867877269881027e-05, "loss": 0.4569, "step": 12038, "task_loss": 0.17443418502807617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7137956619262695, "epoch": 10.18, "learning_rate": 1.7864746399499062e-05, "loss": 0.5191, "step": 12039, "task_loss": 0.23349237442016602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6691463589668274, "epoch": 10.18, "learning_rate": 1.7861615529117094e-05, "loss": 0.7942, "step": 12040, "task_loss": 0.7262662053108215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7568945288658142, "epoch": 10.18, "learning_rate": 1.785848465873513e-05, "loss": 0.582, "step": 12041, "task_loss": 0.936819314956665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7748256921768188, "epoch": 10.18, "learning_rate": 1.7855353788353164e-05, "loss": 0.6297, "step": 12042, "task_loss": 1.015466570854187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8585696816444397, "epoch": 10.18, "learning_rate": 1.78522229179712e-05, "loss": 0.7016, "step": 12043, "task_loss": 0.2474176436662674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32613107562065125, "epoch": 10.18, "learning_rate": 1.784909204758923e-05, "loss": 0.3997, "step": 12044, "task_loss": 0.9145128130912781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5153183937072754, "epoch": 10.18, "learning_rate": 1.7845961177207266e-05, "loss": 0.4675, "step": 12045, "task_loss": 1.2591519355773926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2362176477909088, "epoch": 10.18, "learning_rate": 1.7842830306825298e-05, "loss": 0.3252, "step": 12046, "task_loss": 0.45949381589889526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7145122289657593, "epoch": 10.18, "learning_rate": 1.7839699436443333e-05, "loss": 0.6637, "step": 12047, "task_loss": 1.5078215599060059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5949344038963318, "epoch": 10.18, "learning_rate": 1.7836568566061365e-05, "loss": 0.5413, "step": 12048, "task_loss": 0.664148211479187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20267042517662048, "epoch": 10.19, "learning_rate": 1.78334376956794e-05, "loss": 0.3858, "step": 12049, "task_loss": 0.2480286955833435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5515130758285522, "epoch": 10.19, "learning_rate": 1.7830306825297432e-05, "loss": 0.6232, "step": 12050, "task_loss": 1.1592943668365479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4712703227996826, "epoch": 10.19, "learning_rate": 1.7827175954915467e-05, "loss": 0.6007, "step": 12051, "task_loss": 0.6107129454612732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3108679950237274, "epoch": 10.19, "learning_rate": 1.78240450845335e-05, "loss": 0.4902, "step": 12052, "task_loss": 1.1061323881149292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2998300790786743, "epoch": 10.19, "learning_rate": 1.7820914214151534e-05, "loss": 0.3785, "step": 12053, "task_loss": 0.38599419593811035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3487786054611206, "epoch": 10.19, "learning_rate": 1.781778334376957e-05, "loss": 0.3814, "step": 12054, "task_loss": 0.596051037311554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5298779010772705, "epoch": 10.19, "learning_rate": 1.78146524733876e-05, "loss": 0.6901, "step": 12055, "task_loss": 1.3664515018463135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3733927011489868, "epoch": 10.19, "learning_rate": 1.7811521603005636e-05, "loss": 0.401, "step": 12056, "task_loss": 0.9975472092628479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5162145495414734, "epoch": 10.19, "learning_rate": 1.780839073262367e-05, "loss": 0.4339, "step": 12057, "task_loss": 0.7391830086708069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4401974380016327, "epoch": 10.19, "learning_rate": 1.7805259862241703e-05, "loss": 0.5246, "step": 12058, "task_loss": 0.4944423735141754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7237913012504578, "epoch": 10.19, "learning_rate": 1.780212899185974e-05, "loss": 0.6115, "step": 12059, "task_loss": 1.9113237857818604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6287128925323486, "epoch": 10.19, "learning_rate": 1.7798998121477774e-05, "loss": 0.4814, "step": 12060, "task_loss": 0.7144003510475159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39218080043792725, "epoch": 10.2, "learning_rate": 1.7795867251095805e-05, "loss": 0.5447, "step": 12061, "task_loss": 1.4202492237091064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4332091808319092, "epoch": 10.2, "learning_rate": 1.779273638071384e-05, "loss": 0.4419, "step": 12062, "task_loss": 0.5966039299964905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5354828238487244, "epoch": 10.2, "learning_rate": 1.7789605510331872e-05, "loss": 0.5058, "step": 12063, "task_loss": 0.7709137797355652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6104151010513306, "epoch": 10.2, "learning_rate": 1.7786474639949908e-05, "loss": 0.677, "step": 12064, "task_loss": 1.089311122894287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4579690992832184, "epoch": 10.2, "learning_rate": 1.778334376956794e-05, "loss": 0.538, "step": 12065, "task_loss": 0.4620235860347748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5979282259941101, "epoch": 10.2, "learning_rate": 1.7780212899185974e-05, "loss": 0.5479, "step": 12066, "task_loss": 0.374456524848938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4693400263786316, "epoch": 10.2, "learning_rate": 1.777708202880401e-05, "loss": 0.4606, "step": 12067, "task_loss": 0.5529929995536804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25787633657455444, "epoch": 10.2, "learning_rate": 1.777395115842204e-05, "loss": 0.499, "step": 12068, "task_loss": 0.6285697221755981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6094725131988525, "epoch": 10.2, "learning_rate": 1.7770820288040077e-05, "loss": 0.4073, "step": 12069, "task_loss": 0.4138059914112091 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5640205144882202, "epoch": 10.2, "learning_rate": 1.776768941765811e-05, "loss": 0.4971, "step": 12070, "task_loss": 0.7809429168701172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7550443410873413, "epoch": 10.2, "learning_rate": 1.7764558547276144e-05, "loss": 0.6207, "step": 12071, "task_loss": 0.8398879170417786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3323984146118164, "epoch": 10.2, "learning_rate": 1.7761427676894175e-05, "loss": 0.5886, "step": 12072, "task_loss": 0.5666842460632324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3159791827201843, "epoch": 10.21, "learning_rate": 1.775829680651221e-05, "loss": 0.4098, "step": 12073, "task_loss": 0.2646048963069916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7762233018875122, "epoch": 10.21, "learning_rate": 1.7755165936130246e-05, "loss": 0.6624, "step": 12074, "task_loss": 1.2266242504119873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4130973219871521, "epoch": 10.21, "learning_rate": 1.775203506574828e-05, "loss": 0.4634, "step": 12075, "task_loss": 0.7562817931175232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5121117234230042, "epoch": 10.21, "learning_rate": 1.7748904195366313e-05, "loss": 0.5175, "step": 12076, "task_loss": 1.4286850690841675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4065849184989929, "epoch": 10.21, "learning_rate": 1.7745773324984348e-05, "loss": 0.6416, "step": 12077, "task_loss": 0.6912841796875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.50447678565979, "epoch": 10.21, "learning_rate": 1.774264245460238e-05, "loss": 0.4671, "step": 12078, "task_loss": 0.25848886370658875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4183650612831116, "epoch": 10.21, "learning_rate": 1.7739511584220415e-05, "loss": 0.451, "step": 12079, "task_loss": 0.527722179889679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6097018718719482, "epoch": 10.21, "learning_rate": 1.773638071383845e-05, "loss": 0.3811, "step": 12080, "task_loss": 0.3850257694721222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7559892535209656, "epoch": 10.21, "learning_rate": 1.773324984345648e-05, "loss": 0.5697, "step": 12081, "task_loss": 1.0737966299057007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4425274133682251, "epoch": 10.21, "learning_rate": 1.7730118973074517e-05, "loss": 0.3537, "step": 12082, "task_loss": 0.08130941540002823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7087022662162781, "epoch": 10.21, "learning_rate": 1.772698810269255e-05, "loss": 0.5567, "step": 12083, "task_loss": 0.6555995941162109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6517351865768433, "epoch": 10.21, "learning_rate": 1.7723857232310584e-05, "loss": 0.5335, "step": 12084, "task_loss": 0.7729477286338806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36969369649887085, "epoch": 10.22, "learning_rate": 1.7720726361928616e-05, "loss": 0.4908, "step": 12085, "task_loss": 0.6894257664680481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.517872154712677, "epoch": 10.22, "learning_rate": 1.771759549154665e-05, "loss": 0.5657, "step": 12086, "task_loss": 1.731475830078125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3724587559700012, "epoch": 10.22, "learning_rate": 1.7714464621164683e-05, "loss": 0.4739, "step": 12087, "task_loss": 0.1370657980442047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28910109400749207, "epoch": 10.22, "learning_rate": 1.7711333750782718e-05, "loss": 0.4626, "step": 12088, "task_loss": 0.9637137055397034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5147919058799744, "epoch": 10.22, "learning_rate": 1.770820288040075e-05, "loss": 0.5588, "step": 12089, "task_loss": 0.17413052916526794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6898822784423828, "epoch": 10.22, "learning_rate": 1.7705072010018785e-05, "loss": 0.6392, "step": 12090, "task_loss": 0.3913952708244324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0461584329605103, "epoch": 10.22, "learning_rate": 1.770194113963682e-05, "loss": 0.6091, "step": 12091, "task_loss": 1.3729164600372314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4589737057685852, "epoch": 10.22, "learning_rate": 1.7698810269254855e-05, "loss": 0.4898, "step": 12092, "task_loss": 0.1940767616033554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3833649754524231, "epoch": 10.22, "learning_rate": 1.769567939887289e-05, "loss": 0.4364, "step": 12093, "task_loss": 0.7980688810348511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29123055934906006, "epoch": 10.22, "learning_rate": 1.7692548528490922e-05, "loss": 0.5289, "step": 12094, "task_loss": 0.412523090839386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33498305082321167, "epoch": 10.22, "learning_rate": 1.7689417658108957e-05, "loss": 0.49, "step": 12095, "task_loss": 0.7191115021705627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5982953310012817, "epoch": 10.22, "learning_rate": 1.768628678772699e-05, "loss": 0.5643, "step": 12096, "task_loss": 1.2451001405715942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2948097586631775, "epoch": 10.23, "learning_rate": 1.7683155917345024e-05, "loss": 0.4162, "step": 12097, "task_loss": 0.4813440442085266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4312650263309479, "epoch": 10.23, "learning_rate": 1.7680025046963056e-05, "loss": 0.443, "step": 12098, "task_loss": 0.4428671598434448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1226814985275269, "epoch": 10.23, "learning_rate": 1.767689417658109e-05, "loss": 0.7598, "step": 12099, "task_loss": 1.325530767440796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41781941056251526, "epoch": 10.23, "learning_rate": 1.7673763306199123e-05, "loss": 0.4898, "step": 12100, "task_loss": 0.9023401737213135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2583785355091095, "epoch": 10.23, "learning_rate": 1.7670632435817158e-05, "loss": 0.4591, "step": 12101, "task_loss": 0.4332703948020935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28926944732666016, "epoch": 10.23, "learning_rate": 1.766750156543519e-05, "loss": 0.4441, "step": 12102, "task_loss": 0.13211257755756378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47958904504776, "epoch": 10.23, "learning_rate": 1.7664370695053225e-05, "loss": 0.5073, "step": 12103, "task_loss": 0.2032935619354248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.692255437374115, "epoch": 10.23, "learning_rate": 1.766123982467126e-05, "loss": 0.516, "step": 12104, "task_loss": 0.7967835068702698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8719264268875122, "epoch": 10.23, "learning_rate": 1.7658108954289292e-05, "loss": 0.6197, "step": 12105, "task_loss": 1.8209797143936157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4094051718711853, "epoch": 10.23, "learning_rate": 1.7654978083907327e-05, "loss": 0.6337, "step": 12106, "task_loss": 0.8741607666015625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6099573373794556, "epoch": 10.23, "learning_rate": 1.765184721352536e-05, "loss": 0.6222, "step": 12107, "task_loss": 0.7903013229370117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5944379568099976, "epoch": 10.23, "learning_rate": 1.7648716343143394e-05, "loss": 0.4809, "step": 12108, "task_loss": 0.7794653177261353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5122544169425964, "epoch": 10.24, "learning_rate": 1.764558547276143e-05, "loss": 0.4814, "step": 12109, "task_loss": 0.3959050178527832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4770876169204712, "epoch": 10.24, "learning_rate": 1.7642454602379464e-05, "loss": 0.3798, "step": 12110, "task_loss": 0.2645481526851654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1892295479774475, "epoch": 10.24, "learning_rate": 1.7639323731997496e-05, "loss": 0.3481, "step": 12111, "task_loss": 0.38886502385139465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7293222546577454, "epoch": 10.24, "learning_rate": 1.763619286161553e-05, "loss": 0.5811, "step": 12112, "task_loss": 0.9574298858642578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7497427463531494, "epoch": 10.24, "learning_rate": 1.7633061991233563e-05, "loss": 0.5637, "step": 12113, "task_loss": 1.4975786209106445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.67572021484375, "epoch": 10.24, "learning_rate": 1.7629931120851598e-05, "loss": 0.6841, "step": 12114, "task_loss": 1.651687741279602 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7337039709091187, "epoch": 10.24, "learning_rate": 1.762680025046963e-05, "loss": 0.7132, "step": 12115, "task_loss": 0.8532959222793579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.552988588809967, "epoch": 10.24, "learning_rate": 1.7623669380087665e-05, "loss": 0.4405, "step": 12116, "task_loss": 0.49354392290115356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39959484338760376, "epoch": 10.24, "learning_rate": 1.76205385097057e-05, "loss": 0.6258, "step": 12117, "task_loss": 0.8514547944068909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1248748302459717, "epoch": 10.24, "learning_rate": 1.7617407639323732e-05, "loss": 0.8035, "step": 12118, "task_loss": 1.1672132015228271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.585518479347229, "epoch": 10.24, "learning_rate": 1.7614276768941767e-05, "loss": 0.4123, "step": 12119, "task_loss": 0.5796045660972595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3688305616378784, "epoch": 10.24, "learning_rate": 1.76111458985598e-05, "loss": 0.4004, "step": 12120, "task_loss": 1.3265544176101685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38508957624435425, "epoch": 10.25, "learning_rate": 1.7608015028177834e-05, "loss": 0.4362, "step": 12121, "task_loss": 0.0689396932721138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31436505913734436, "epoch": 10.25, "learning_rate": 1.7604884157795866e-05, "loss": 0.4901, "step": 12122, "task_loss": 0.3027558922767639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2550210952758789, "epoch": 10.25, "learning_rate": 1.76017532874139e-05, "loss": 0.4442, "step": 12123, "task_loss": 0.6222748160362244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7247139811515808, "epoch": 10.25, "learning_rate": 1.7598622417031936e-05, "loss": 0.611, "step": 12124, "task_loss": 0.990831732749939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5541338324546814, "epoch": 10.25, "learning_rate": 1.7595491546649968e-05, "loss": 0.5055, "step": 12125, "task_loss": 0.5877610445022583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5502634048461914, "epoch": 10.25, "learning_rate": 1.7592360676268003e-05, "loss": 0.4975, "step": 12126, "task_loss": 0.5198662281036377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5692994594573975, "epoch": 10.25, "learning_rate": 1.758922980588604e-05, "loss": 0.5405, "step": 12127, "task_loss": 0.8563494086265564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39899617433547974, "epoch": 10.25, "learning_rate": 1.7586098935504074e-05, "loss": 0.4737, "step": 12128, "task_loss": 0.40708664059638977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8247205018997192, "epoch": 10.25, "learning_rate": 1.7582968065122105e-05, "loss": 0.4997, "step": 12129, "task_loss": 0.6439359188079834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5629087686538696, "epoch": 10.25, "learning_rate": 1.757983719474014e-05, "loss": 0.703, "step": 12130, "task_loss": 0.8603923916816711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5646600723266602, "epoch": 10.25, "learning_rate": 1.7576706324358172e-05, "loss": 0.7702, "step": 12131, "task_loss": 1.4253817796707153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9408529996871948, "epoch": 10.26, "learning_rate": 1.7573575453976208e-05, "loss": 0.5873, "step": 12132, "task_loss": 2.546365737915039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.298992395401001, "epoch": 10.26, "learning_rate": 1.757044458359424e-05, "loss": 0.5947, "step": 12133, "task_loss": 0.6273448467254639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3034610152244568, "epoch": 10.26, "learning_rate": 1.7567313713212275e-05, "loss": 0.3991, "step": 12134, "task_loss": 0.4971887171268463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8183097243309021, "epoch": 10.26, "learning_rate": 1.7564182842830306e-05, "loss": 0.6013, "step": 12135, "task_loss": 0.5642912983894348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6853615045547485, "epoch": 10.26, "learning_rate": 1.756105197244834e-05, "loss": 0.4068, "step": 12136, "task_loss": 0.6056121587753296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2095937281847, "epoch": 10.26, "learning_rate": 1.7557921102066373e-05, "loss": 0.4076, "step": 12137, "task_loss": 0.3784959614276886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33837854862213135, "epoch": 10.26, "learning_rate": 1.755479023168441e-05, "loss": 0.4875, "step": 12138, "task_loss": 0.47654134035110474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5795750021934509, "epoch": 10.26, "learning_rate": 1.755165936130244e-05, "loss": 0.5628, "step": 12139, "task_loss": 0.5714025497436523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2206198126077652, "epoch": 10.26, "learning_rate": 1.7548528490920475e-05, "loss": 0.3323, "step": 12140, "task_loss": 0.4442630112171173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8855122327804565, "epoch": 10.26, "learning_rate": 1.754539762053851e-05, "loss": 0.4963, "step": 12141, "task_loss": 0.27703234553337097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5920486450195312, "epoch": 10.26, "learning_rate": 1.7542266750156546e-05, "loss": 0.5779, "step": 12142, "task_loss": 0.2762804925441742 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39011240005493164, "epoch": 10.26, "learning_rate": 1.7539135879774577e-05, "loss": 0.4518, "step": 12143, "task_loss": 1.1141444444656372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4639725685119629, "epoch": 10.27, "learning_rate": 1.7536005009392613e-05, "loss": 0.5686, "step": 12144, "task_loss": 0.30392763018608093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3288063406944275, "epoch": 10.27, "learning_rate": 1.7532874139010648e-05, "loss": 0.4498, "step": 12145, "task_loss": 0.5114906430244446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33811482787132263, "epoch": 10.27, "learning_rate": 1.752974326862868e-05, "loss": 0.4504, "step": 12146, "task_loss": 1.084761142730713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5091581344604492, "epoch": 10.27, "learning_rate": 1.7526612398246715e-05, "loss": 0.5187, "step": 12147, "task_loss": 0.739599347114563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6010454893112183, "epoch": 10.27, "learning_rate": 1.7523481527864747e-05, "loss": 0.6833, "step": 12148, "task_loss": 1.2797173261642456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6224385499954224, "epoch": 10.27, "learning_rate": 1.7520350657482782e-05, "loss": 0.4896, "step": 12149, "task_loss": 0.7689597010612488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4768401086330414, "epoch": 10.27, "learning_rate": 1.7517219787100814e-05, "loss": 0.5809, "step": 12150, "task_loss": 0.8329519629478455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3430788516998291, "epoch": 10.27, "learning_rate": 1.751408891671885e-05, "loss": 0.4345, "step": 12151, "task_loss": 0.4872564971446991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4670379161834717, "epoch": 10.27, "learning_rate": 1.7510958046336884e-05, "loss": 0.4552, "step": 12152, "task_loss": 0.9554538726806641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5339517593383789, "epoch": 10.27, "learning_rate": 1.7507827175954916e-05, "loss": 0.492, "step": 12153, "task_loss": 0.20278915762901306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8736552000045776, "epoch": 10.27, "learning_rate": 1.750469630557295e-05, "loss": 0.6452, "step": 12154, "task_loss": 0.8815100789070129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5381532311439514, "epoch": 10.27, "learning_rate": 1.7501565435190983e-05, "loss": 0.5333, "step": 12155, "task_loss": 0.6295170187950134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.449758380651474, "epoch": 10.28, "learning_rate": 1.7498434564809018e-05, "loss": 0.5593, "step": 12156, "task_loss": 0.36644792556762695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5347714424133301, "epoch": 10.28, "learning_rate": 1.749530369442705e-05, "loss": 0.8123, "step": 12157, "task_loss": 0.3820698857307434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5335156917572021, "epoch": 10.28, "learning_rate": 1.7492172824045085e-05, "loss": 0.4865, "step": 12158, "task_loss": 1.1111325025558472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3091448247432709, "epoch": 10.28, "learning_rate": 1.748904195366312e-05, "loss": 0.3366, "step": 12159, "task_loss": 0.2085484117269516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4802117347717285, "epoch": 10.28, "learning_rate": 1.7485911083281155e-05, "loss": 0.618, "step": 12160, "task_loss": 1.4415159225463867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5550795793533325, "epoch": 10.28, "learning_rate": 1.7482780212899187e-05, "loss": 0.4485, "step": 12161, "task_loss": 0.9607769846916199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25969797372817993, "epoch": 10.28, "learning_rate": 1.7479649342517222e-05, "loss": 0.4299, "step": 12162, "task_loss": 0.2780642807483673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47950419783592224, "epoch": 10.28, "learning_rate": 1.7476518472135254e-05, "loss": 0.3911, "step": 12163, "task_loss": 0.6902692317962646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41208064556121826, "epoch": 10.28, "learning_rate": 1.747338760175329e-05, "loss": 0.4248, "step": 12164, "task_loss": 0.15303565561771393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3424745202064514, "epoch": 10.28, "learning_rate": 1.7470256731371324e-05, "loss": 0.3354, "step": 12165, "task_loss": 0.5274087190628052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.541602373123169, "epoch": 10.28, "learning_rate": 1.7467125860989356e-05, "loss": 0.5845, "step": 12166, "task_loss": 0.6942696571350098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3709469139575958, "epoch": 10.28, "learning_rate": 1.746399499060739e-05, "loss": 0.4667, "step": 12167, "task_loss": 1.3158189058303833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.1968079805374146, "epoch": 10.29, "learning_rate": 1.7460864120225423e-05, "loss": 0.663, "step": 12168, "task_loss": 1.0810977220535278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4244834780693054, "epoch": 10.29, "learning_rate": 1.7457733249843458e-05, "loss": 0.5346, "step": 12169, "task_loss": 0.5792749524116516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.548768162727356, "epoch": 10.29, "learning_rate": 1.745460237946149e-05, "loss": 0.4861, "step": 12170, "task_loss": 0.4075787663459778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8195865750312805, "epoch": 10.29, "learning_rate": 1.7451471509079525e-05, "loss": 0.7251, "step": 12171, "task_loss": 1.2254607677459717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4595829248428345, "epoch": 10.29, "learning_rate": 1.7448340638697557e-05, "loss": 0.5706, "step": 12172, "task_loss": 0.8418439030647278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37524357438087463, "epoch": 10.29, "learning_rate": 1.7445209768315592e-05, "loss": 0.6144, "step": 12173, "task_loss": 0.7548801302909851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.351808100938797, "epoch": 10.29, "learning_rate": 1.7442078897933624e-05, "loss": 0.5286, "step": 12174, "task_loss": 0.20178617537021637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35312336683273315, "epoch": 10.29, "learning_rate": 1.743894802755166e-05, "loss": 0.5638, "step": 12175, "task_loss": 0.5014525651931763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4214268922805786, "epoch": 10.29, "learning_rate": 1.7435817157169694e-05, "loss": 0.4556, "step": 12176, "task_loss": 0.7634850740432739 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5734293460845947, "epoch": 10.29, "learning_rate": 1.743268628678773e-05, "loss": 0.4598, "step": 12177, "task_loss": 0.4482349157333374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23997092247009277, "epoch": 10.29, "learning_rate": 1.7429555416405764e-05, "loss": 0.6078, "step": 12178, "task_loss": 0.021681414917111397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8235051035881042, "epoch": 10.29, "learning_rate": 1.7426424546023796e-05, "loss": 0.6131, "step": 12179, "task_loss": 0.6130803823471069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8186478614807129, "epoch": 10.3, "learning_rate": 1.742329367564183e-05, "loss": 0.6552, "step": 12180, "task_loss": 1.2082864046096802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6827694773674011, "epoch": 10.3, "learning_rate": 1.7420162805259863e-05, "loss": 0.5427, "step": 12181, "task_loss": 0.43797826766967773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3387342095375061, "epoch": 10.3, "learning_rate": 1.7417031934877898e-05, "loss": 0.3994, "step": 12182, "task_loss": 0.6677094101905823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22866083681583405, "epoch": 10.3, "learning_rate": 1.741390106449593e-05, "loss": 0.4692, "step": 12183, "task_loss": 0.39938992261886597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5735954642295837, "epoch": 10.3, "learning_rate": 1.7410770194113965e-05, "loss": 0.5651, "step": 12184, "task_loss": 0.2714807689189911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6572301983833313, "epoch": 10.3, "learning_rate": 1.7407639323731997e-05, "loss": 0.432, "step": 12185, "task_loss": 0.4191094934940338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4912635087966919, "epoch": 10.3, "learning_rate": 1.7404508453350032e-05, "loss": 0.5024, "step": 12186, "task_loss": 0.4234296679496765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5507910251617432, "epoch": 10.3, "learning_rate": 1.7401377582968064e-05, "loss": 0.6193, "step": 12187, "task_loss": 0.8246733546257019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6601153016090393, "epoch": 10.3, "learning_rate": 1.73982467125861e-05, "loss": 0.5215, "step": 12188, "task_loss": 1.04474937915802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4999523162841797, "epoch": 10.3, "learning_rate": 1.7395115842204134e-05, "loss": 0.4893, "step": 12189, "task_loss": 0.5610742568969727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7085359692573547, "epoch": 10.3, "learning_rate": 1.7391984971822166e-05, "loss": 0.4611, "step": 12190, "task_loss": 0.48162421584129333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4173715114593506, "epoch": 10.3, "learning_rate": 1.73888541014402e-05, "loss": 0.5023, "step": 12191, "task_loss": 0.9034230709075928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4331740736961365, "epoch": 10.31, "learning_rate": 1.7385723231058233e-05, "loss": 0.4689, "step": 12192, "task_loss": 1.1922953128814697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5675303936004639, "epoch": 10.31, "learning_rate": 1.7382592360676268e-05, "loss": 0.559, "step": 12193, "task_loss": 0.888352632522583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7231813073158264, "epoch": 10.31, "learning_rate": 1.7379461490294303e-05, "loss": 0.5844, "step": 12194, "task_loss": 0.715495228767395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2692994177341461, "epoch": 10.31, "learning_rate": 1.737633061991234e-05, "loss": 0.4257, "step": 12195, "task_loss": 0.08780968934297562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5697711706161499, "epoch": 10.31, "learning_rate": 1.737319974953037e-05, "loss": 0.6708, "step": 12196, "task_loss": 0.710485577583313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0372469425201416, "epoch": 10.31, "learning_rate": 1.7370068879148405e-05, "loss": 0.5986, "step": 12197, "task_loss": 0.2013901025056839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36944013833999634, "epoch": 10.31, "learning_rate": 1.7366938008766437e-05, "loss": 0.5026, "step": 12198, "task_loss": 0.7722318768501282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.531379222869873, "epoch": 10.31, "learning_rate": 1.7363807138384472e-05, "loss": 0.5159, "step": 12199, "task_loss": 0.18065273761749268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47422710061073303, "epoch": 10.31, "learning_rate": 1.7360676268002504e-05, "loss": 0.5259, "step": 12200, "task_loss": 0.8400028347969055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8568718433380127, "epoch": 10.31, "learning_rate": 1.735754539762054e-05, "loss": 0.6497, "step": 12201, "task_loss": 0.6886548399925232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6522707343101501, "epoch": 10.31, "learning_rate": 1.7354414527238575e-05, "loss": 0.4885, "step": 12202, "task_loss": 0.9855214357376099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3413834869861603, "epoch": 10.32, "learning_rate": 1.7351283656856606e-05, "loss": 0.4146, "step": 12203, "task_loss": 0.11795129626989365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37840113043785095, "epoch": 10.32, "learning_rate": 1.734815278647464e-05, "loss": 0.4439, "step": 12204, "task_loss": 0.17236557602882385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6404232978820801, "epoch": 10.32, "learning_rate": 1.7345021916092673e-05, "loss": 0.7254, "step": 12205, "task_loss": 0.40103477239608765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4390019178390503, "epoch": 10.32, "learning_rate": 1.734189104571071e-05, "loss": 0.4521, "step": 12206, "task_loss": 0.4988808035850525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5556554794311523, "epoch": 10.32, "learning_rate": 1.733876017532874e-05, "loss": 0.6874, "step": 12207, "task_loss": 0.6498032212257385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5765914916992188, "epoch": 10.32, "learning_rate": 1.7335629304946775e-05, "loss": 0.4352, "step": 12208, "task_loss": 1.6233189105987549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5411723852157593, "epoch": 10.32, "learning_rate": 1.733249843456481e-05, "loss": 0.571, "step": 12209, "task_loss": 1.6129049062728882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6225683689117432, "epoch": 10.32, "learning_rate": 1.7329367564182842e-05, "loss": 0.4976, "step": 12210, "task_loss": 0.6423135995864868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8439531326293945, "epoch": 10.32, "learning_rate": 1.7326236693800878e-05, "loss": 0.6191, "step": 12211, "task_loss": 0.3985973298549652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.765265941619873, "epoch": 10.32, "learning_rate": 1.7323105823418913e-05, "loss": 0.5385, "step": 12212, "task_loss": 0.7927666902542114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3803078234195709, "epoch": 10.32, "learning_rate": 1.7319974953036944e-05, "loss": 0.5381, "step": 12213, "task_loss": 1.3237972259521484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5011600255966187, "epoch": 10.32, "learning_rate": 1.731684408265498e-05, "loss": 0.5679, "step": 12214, "task_loss": 0.6040918231010437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3896874189376831, "epoch": 10.33, "learning_rate": 1.7313713212273015e-05, "loss": 0.4288, "step": 12215, "task_loss": 0.7434263229370117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2747374176979065, "epoch": 10.33, "learning_rate": 1.7310582341891047e-05, "loss": 0.6155, "step": 12216, "task_loss": 0.3529528081417084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2948361337184906, "epoch": 10.33, "learning_rate": 1.7307451471509082e-05, "loss": 0.3907, "step": 12217, "task_loss": 0.5203471779823303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6264392137527466, "epoch": 10.33, "learning_rate": 1.7304320601127114e-05, "loss": 0.4644, "step": 12218, "task_loss": 0.9097943305969238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6916968822479248, "epoch": 10.33, "learning_rate": 1.730118973074515e-05, "loss": 0.9255, "step": 12219, "task_loss": 1.1891075372695923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6629174947738647, "epoch": 10.33, "learning_rate": 1.729805886036318e-05, "loss": 0.5083, "step": 12220, "task_loss": 1.116591453552246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48602724075317383, "epoch": 10.33, "learning_rate": 1.7294927989981216e-05, "loss": 0.4826, "step": 12221, "task_loss": 0.09738582372665405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5736634731292725, "epoch": 10.33, "learning_rate": 1.7291797119599247e-05, "loss": 0.4693, "step": 12222, "task_loss": 0.6700268983840942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7391406893730164, "epoch": 10.33, "learning_rate": 1.7288666249217283e-05, "loss": 0.5782, "step": 12223, "task_loss": 0.7286933660507202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4050447940826416, "epoch": 10.33, "learning_rate": 1.7285535378835314e-05, "loss": 0.454, "step": 12224, "task_loss": 0.2854902446269989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34053921699523926, "epoch": 10.33, "learning_rate": 1.728240450845335e-05, "loss": 0.3791, "step": 12225, "task_loss": 0.34353652596473694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22024428844451904, "epoch": 10.33, "learning_rate": 1.7279273638071385e-05, "loss": 0.5084, "step": 12226, "task_loss": 0.3408294916152954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39716506004333496, "epoch": 10.34, "learning_rate": 1.727614276768942e-05, "loss": 0.48, "step": 12227, "task_loss": 0.40844520926475525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5761492848396301, "epoch": 10.34, "learning_rate": 1.7273011897307455e-05, "loss": 0.4828, "step": 12228, "task_loss": 1.0160927772521973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7320243120193481, "epoch": 10.34, "learning_rate": 1.7269881026925487e-05, "loss": 0.5331, "step": 12229, "task_loss": 0.660978376865387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2697523534297943, "epoch": 10.34, "learning_rate": 1.7266750156543522e-05, "loss": 0.4401, "step": 12230, "task_loss": 0.4118894934654236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34834277629852295, "epoch": 10.34, "learning_rate": 1.7263619286161554e-05, "loss": 0.5629, "step": 12231, "task_loss": 0.08823658525943756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27456793189048767, "epoch": 10.34, "learning_rate": 1.726048841577959e-05, "loss": 0.4078, "step": 12232, "task_loss": 0.28218990564346313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5704047679901123, "epoch": 10.34, "learning_rate": 1.725735754539762e-05, "loss": 0.4961, "step": 12233, "task_loss": 0.368924617767334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4881141185760498, "epoch": 10.34, "learning_rate": 1.7254226675015656e-05, "loss": 0.5204, "step": 12234, "task_loss": 0.8035383224487305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3036963939666748, "epoch": 10.34, "learning_rate": 1.7251095804633688e-05, "loss": 0.3655, "step": 12235, "task_loss": 0.5877688527107239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5915317535400391, "epoch": 10.34, "learning_rate": 1.7247964934251723e-05, "loss": 0.5099, "step": 12236, "task_loss": 0.9211133122444153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3020688593387604, "epoch": 10.34, "learning_rate": 1.7244834063869755e-05, "loss": 0.5487, "step": 12237, "task_loss": 0.7379892468452454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8382809162139893, "epoch": 10.34, "learning_rate": 1.724170319348779e-05, "loss": 0.6193, "step": 12238, "task_loss": 0.7261719107627869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7891133427619934, "epoch": 10.35, "learning_rate": 1.7238572323105825e-05, "loss": 0.5514, "step": 12239, "task_loss": 1.1004043817520142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4192657768726349, "epoch": 10.35, "learning_rate": 1.7235441452723857e-05, "loss": 0.447, "step": 12240, "task_loss": 0.7670013308525085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32054758071899414, "epoch": 10.35, "learning_rate": 1.7232310582341892e-05, "loss": 0.4434, "step": 12241, "task_loss": 0.5419958829879761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3918742537498474, "epoch": 10.35, "learning_rate": 1.7229179711959924e-05, "loss": 0.5362, "step": 12242, "task_loss": 0.8697637319564819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5483147501945496, "epoch": 10.35, "learning_rate": 1.722604884157796e-05, "loss": 0.5763, "step": 12243, "task_loss": 0.7845941185951233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42728525400161743, "epoch": 10.35, "learning_rate": 1.7222917971195994e-05, "loss": 0.3382, "step": 12244, "task_loss": 0.28603053092956543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2533744275569916, "epoch": 10.35, "learning_rate": 1.721978710081403e-05, "loss": 0.5509, "step": 12245, "task_loss": 0.13504517078399658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35888269543647766, "epoch": 10.35, "learning_rate": 1.721665623043206e-05, "loss": 0.5038, "step": 12246, "task_loss": 0.38439247012138367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7100124955177307, "epoch": 10.35, "learning_rate": 1.7213525360050096e-05, "loss": 0.6588, "step": 12247, "task_loss": 0.9679797887802124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49897757172584534, "epoch": 10.35, "learning_rate": 1.7210394489668128e-05, "loss": 0.5138, "step": 12248, "task_loss": 0.5328618884086609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6105197072029114, "epoch": 10.35, "learning_rate": 1.7207263619286163e-05, "loss": 0.4437, "step": 12249, "task_loss": 0.4941640794277191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32585203647613525, "epoch": 10.35, "learning_rate": 1.7204132748904198e-05, "loss": 0.4398, "step": 12250, "task_loss": 0.5088539719581604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5296124815940857, "epoch": 10.36, "learning_rate": 1.720100187852223e-05, "loss": 0.3938, "step": 12251, "task_loss": 0.7750639915466309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7517831325531006, "epoch": 10.36, "learning_rate": 1.7197871008140265e-05, "loss": 0.6655, "step": 12252, "task_loss": 1.4544979333877563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.533845841884613, "epoch": 10.36, "learning_rate": 1.7194740137758297e-05, "loss": 0.5206, "step": 12253, "task_loss": 0.8148673176765442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.727154552936554, "epoch": 10.36, "learning_rate": 1.7191609267376332e-05, "loss": 0.7482, "step": 12254, "task_loss": 0.7734328508377075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33595824241638184, "epoch": 10.36, "learning_rate": 1.7188478396994364e-05, "loss": 0.5418, "step": 12255, "task_loss": 1.254357933998108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4553719162940979, "epoch": 10.36, "learning_rate": 1.71853475266124e-05, "loss": 0.5266, "step": 12256, "task_loss": 0.16209007799625397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3668401837348938, "epoch": 10.36, "learning_rate": 1.718221665623043e-05, "loss": 0.464, "step": 12257, "task_loss": 0.4056311249732971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4568755030632019, "epoch": 10.36, "learning_rate": 1.7179085785848466e-05, "loss": 0.5027, "step": 12258, "task_loss": 0.8823167085647583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5354210138320923, "epoch": 10.36, "learning_rate": 1.7175954915466498e-05, "loss": 0.4576, "step": 12259, "task_loss": 1.293291687965393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5193758010864258, "epoch": 10.36, "learning_rate": 1.7172824045084533e-05, "loss": 0.4799, "step": 12260, "task_loss": 1.2169257402420044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5151838064193726, "epoch": 10.36, "learning_rate": 1.7169693174702568e-05, "loss": 0.5673, "step": 12261, "task_loss": 0.32707077264785767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4340287446975708, "epoch": 10.36, "learning_rate": 1.7166562304320603e-05, "loss": 0.4805, "step": 12262, "task_loss": 0.713024377822876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5026726722717285, "epoch": 10.37, "learning_rate": 1.716343143393864e-05, "loss": 0.4264, "step": 12263, "task_loss": 0.4953550696372986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6420920491218567, "epoch": 10.37, "learning_rate": 1.716030056355667e-05, "loss": 0.5906, "step": 12264, "task_loss": 0.5675892233848572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49645352363586426, "epoch": 10.37, "learning_rate": 1.7157169693174706e-05, "loss": 0.4946, "step": 12265, "task_loss": 0.3855322301387787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41374439001083374, "epoch": 10.37, "learning_rate": 1.7154038822792737e-05, "loss": 0.5889, "step": 12266, "task_loss": 1.4620635509490967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37082403898239136, "epoch": 10.37, "learning_rate": 1.7150907952410772e-05, "loss": 0.5309, "step": 12267, "task_loss": 0.10343547910451889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7614217400550842, "epoch": 10.37, "learning_rate": 1.7147777082028804e-05, "loss": 0.4293, "step": 12268, "task_loss": 1.0433911085128784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3330461382865906, "epoch": 10.37, "learning_rate": 1.714464621164684e-05, "loss": 0.3955, "step": 12269, "task_loss": 0.3990883529186249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4478636384010315, "epoch": 10.37, "learning_rate": 1.714151534126487e-05, "loss": 0.5609, "step": 12270, "task_loss": 0.7905673384666443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6165738105773926, "epoch": 10.37, "learning_rate": 1.7138384470882906e-05, "loss": 0.5414, "step": 12271, "task_loss": 0.19374816119670868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45530056953430176, "epoch": 10.37, "learning_rate": 1.7135253600500938e-05, "loss": 0.5715, "step": 12272, "task_loss": 0.28644248843193054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8567829132080078, "epoch": 10.37, "learning_rate": 1.7132122730118973e-05, "loss": 0.689, "step": 12273, "task_loss": 1.2195895910263062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5616340637207031, "epoch": 10.38, "learning_rate": 1.7128991859737005e-05, "loss": 0.491, "step": 12274, "task_loss": 0.4310693144798279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5032805800437927, "epoch": 10.38, "learning_rate": 1.712586098935504e-05, "loss": 0.3773, "step": 12275, "task_loss": 0.9398649334907532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5700476169586182, "epoch": 10.38, "learning_rate": 1.7122730118973075e-05, "loss": 0.5107, "step": 12276, "task_loss": 0.6020041108131409 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7694606781005859, "epoch": 10.38, "learning_rate": 1.7119599248591107e-05, "loss": 0.5623, "step": 12277, "task_loss": 1.308854579925537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.240865558385849, "epoch": 10.38, "learning_rate": 1.7116468378209142e-05, "loss": 0.5059, "step": 12278, "task_loss": 0.5923565626144409 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3567999601364136, "epoch": 10.38, "learning_rate": 1.7113337507827178e-05, "loss": 0.3847, "step": 12279, "task_loss": 0.21628281474113464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4050002992153168, "epoch": 10.38, "learning_rate": 1.7110206637445213e-05, "loss": 0.4302, "step": 12280, "task_loss": 1.063292384147644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6414216756820679, "epoch": 10.38, "learning_rate": 1.7107075767063245e-05, "loss": 0.502, "step": 12281, "task_loss": 0.2143091857433319 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37984931468963623, "epoch": 10.38, "learning_rate": 1.710394489668128e-05, "loss": 0.5156, "step": 12282, "task_loss": 0.493586927652359 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6817167401313782, "epoch": 10.38, "learning_rate": 1.710081402629931e-05, "loss": 0.595, "step": 12283, "task_loss": 1.2521718740463257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6356970071792603, "epoch": 10.38, "learning_rate": 1.7097683155917347e-05, "loss": 0.5415, "step": 12284, "task_loss": 0.6144841313362122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5662777423858643, "epoch": 10.38, "learning_rate": 1.709455228553538e-05, "loss": 0.4402, "step": 12285, "task_loss": 0.40835049748420715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5199501514434814, "epoch": 10.39, "learning_rate": 1.7091421415153414e-05, "loss": 0.5184, "step": 12286, "task_loss": 0.2665833532810211 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49974730610847473, "epoch": 10.39, "learning_rate": 1.708829054477145e-05, "loss": 0.5493, "step": 12287, "task_loss": 0.5451720952987671 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47618329524993896, "epoch": 10.39, "learning_rate": 1.708515967438948e-05, "loss": 0.4025, "step": 12288, "task_loss": 0.6389619708061218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29549553990364075, "epoch": 10.39, "learning_rate": 1.7082028804007516e-05, "loss": 0.3739, "step": 12289, "task_loss": 0.30647122859954834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48130083084106445, "epoch": 10.39, "learning_rate": 1.7078897933625547e-05, "loss": 0.5182, "step": 12290, "task_loss": 0.5374861359596252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7828546762466431, "epoch": 10.39, "learning_rate": 1.7075767063243583e-05, "loss": 0.5784, "step": 12291, "task_loss": 1.2901469469070435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40620285272598267, "epoch": 10.39, "learning_rate": 1.7072636192861614e-05, "loss": 0.6692, "step": 12292, "task_loss": 1.235414981842041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.782830536365509, "epoch": 10.39, "learning_rate": 1.706950532247965e-05, "loss": 0.7079, "step": 12293, "task_loss": 1.218846082687378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.577411413192749, "epoch": 10.39, "learning_rate": 1.7066374452097685e-05, "loss": 0.4503, "step": 12294, "task_loss": 0.3457411527633667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38234925270080566, "epoch": 10.39, "learning_rate": 1.7063243581715717e-05, "loss": 0.6871, "step": 12295, "task_loss": 0.9793305397033691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4326210618019104, "epoch": 10.39, "learning_rate": 1.7060112711333752e-05, "loss": 0.4779, "step": 12296, "task_loss": 0.22104841470718384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.596454381942749, "epoch": 10.39, "learning_rate": 1.7056981840951787e-05, "loss": 0.4515, "step": 12297, "task_loss": 0.3910447359085083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4186401963233948, "epoch": 10.4, "learning_rate": 1.705385097056982e-05, "loss": 0.4098, "step": 12298, "task_loss": 0.3806252181529999 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48662275075912476, "epoch": 10.4, "learning_rate": 1.7050720100187854e-05, "loss": 0.4946, "step": 12299, "task_loss": 0.30145952105522156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3112031817436218, "epoch": 10.4, "learning_rate": 1.704758922980589e-05, "loss": 0.5935, "step": 12300, "task_loss": 0.67006915807724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41914063692092896, "epoch": 10.4, "learning_rate": 1.704445835942392e-05, "loss": 0.5307, "step": 12301, "task_loss": 0.7782171964645386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6413158178329468, "epoch": 10.4, "learning_rate": 1.7041327489041956e-05, "loss": 0.4785, "step": 12302, "task_loss": 0.6725981831550598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42512789368629456, "epoch": 10.4, "learning_rate": 1.7038196618659988e-05, "loss": 0.5131, "step": 12303, "task_loss": 0.5381811261177063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8157612681388855, "epoch": 10.4, "learning_rate": 1.7035065748278023e-05, "loss": 0.5121, "step": 12304, "task_loss": 0.3281375765800476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28827178478240967, "epoch": 10.4, "learning_rate": 1.7031934877896055e-05, "loss": 0.5247, "step": 12305, "task_loss": 0.4648421108722687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41553795337677, "epoch": 10.4, "learning_rate": 1.702880400751409e-05, "loss": 0.5953, "step": 12306, "task_loss": 0.270324170589447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45487162470817566, "epoch": 10.4, "learning_rate": 1.702567313713212e-05, "loss": 0.5071, "step": 12307, "task_loss": 0.6547653675079346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.357261061668396, "epoch": 10.4, "learning_rate": 1.7022542266750157e-05, "loss": 0.408, "step": 12308, "task_loss": 0.4919606149196625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25473788380622864, "epoch": 10.4, "learning_rate": 1.701941139636819e-05, "loss": 0.3839, "step": 12309, "task_loss": 0.12997429072856903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6615239381790161, "epoch": 10.41, "learning_rate": 1.7016280525986224e-05, "loss": 0.6802, "step": 12310, "task_loss": 1.3572030067443848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6650342345237732, "epoch": 10.41, "learning_rate": 1.701314965560426e-05, "loss": 0.4741, "step": 12311, "task_loss": 0.4832237660884857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7184391021728516, "epoch": 10.41, "learning_rate": 1.7010018785222294e-05, "loss": 0.5568, "step": 12312, "task_loss": 1.2294588088989258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3721161186695099, "epoch": 10.41, "learning_rate": 1.700688791484033e-05, "loss": 0.4657, "step": 12313, "task_loss": 0.765754759311676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3845919966697693, "epoch": 10.41, "learning_rate": 1.700375704445836e-05, "loss": 0.4132, "step": 12314, "task_loss": 0.39494407176971436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.554307758808136, "epoch": 10.41, "learning_rate": 1.7000626174076396e-05, "loss": 0.5825, "step": 12315, "task_loss": 0.7759048938751221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3120424151420593, "epoch": 10.41, "learning_rate": 1.6997495303694428e-05, "loss": 0.5132, "step": 12316, "task_loss": 0.3829892575740814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3833746910095215, "epoch": 10.41, "learning_rate": 1.6994364433312463e-05, "loss": 0.6081, "step": 12317, "task_loss": 0.23813539743423462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4294738471508026, "epoch": 10.41, "learning_rate": 1.6991233562930495e-05, "loss": 0.4891, "step": 12318, "task_loss": 1.2044810056686401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5462110042572021, "epoch": 10.41, "learning_rate": 1.698810269254853e-05, "loss": 0.5835, "step": 12319, "task_loss": 0.7627325654029846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7123076319694519, "epoch": 10.41, "learning_rate": 1.6984971822166562e-05, "loss": 0.6619, "step": 12320, "task_loss": 0.8440114259719849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4917060136795044, "epoch": 10.41, "learning_rate": 1.6981840951784597e-05, "loss": 0.4493, "step": 12321, "task_loss": 0.784010648727417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.52115398645401, "epoch": 10.42, "learning_rate": 1.697871008140263e-05, "loss": 0.5197, "step": 12322, "task_loss": 0.9970873594284058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5352215766906738, "epoch": 10.42, "learning_rate": 1.6975579211020664e-05, "loss": 0.5101, "step": 12323, "task_loss": 0.25024116039276123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4464530348777771, "epoch": 10.42, "learning_rate": 1.69724483406387e-05, "loss": 0.5822, "step": 12324, "task_loss": 0.8409613966941833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4634682238101959, "epoch": 10.42, "learning_rate": 1.696931747025673e-05, "loss": 0.6498, "step": 12325, "task_loss": 0.31351298093795776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42263978719711304, "epoch": 10.42, "learning_rate": 1.6966186599874766e-05, "loss": 0.5043, "step": 12326, "task_loss": 0.42905542254447937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6689680218696594, "epoch": 10.42, "learning_rate": 1.6963055729492798e-05, "loss": 0.6305, "step": 12327, "task_loss": 1.2404260635375977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44589319825172424, "epoch": 10.42, "learning_rate": 1.6959924859110833e-05, "loss": 0.4203, "step": 12328, "task_loss": 0.9283291697502136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4498080909252167, "epoch": 10.42, "learning_rate": 1.6956793988728868e-05, "loss": 0.5226, "step": 12329, "task_loss": 0.0670446902513504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16393128037452698, "epoch": 10.42, "learning_rate": 1.6953663118346903e-05, "loss": 0.2027, "step": 12330, "task_loss": 0.023878000676631927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7711175084114075, "epoch": 10.42, "learning_rate": 1.6950532247964935e-05, "loss": 0.5569, "step": 12331, "task_loss": 0.9218887090682983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5217347145080566, "epoch": 10.42, "learning_rate": 1.694740137758297e-05, "loss": 0.4495, "step": 12332, "task_loss": 0.6408128142356873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34963053464889526, "epoch": 10.42, "learning_rate": 1.6944270507201002e-05, "loss": 0.6575, "step": 12333, "task_loss": 0.7822003364562988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4819610118865967, "epoch": 10.43, "learning_rate": 1.6941139636819037e-05, "loss": 0.6109, "step": 12334, "task_loss": 0.909914493560791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3376021683216095, "epoch": 10.43, "learning_rate": 1.693800876643707e-05, "loss": 0.5557, "step": 12335, "task_loss": 0.15435084700584412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7504376173019409, "epoch": 10.43, "learning_rate": 1.6934877896055104e-05, "loss": 0.5462, "step": 12336, "task_loss": 1.635261058807373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31656765937805176, "epoch": 10.43, "learning_rate": 1.693174702567314e-05, "loss": 0.601, "step": 12337, "task_loss": 0.4491124749183655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5915046334266663, "epoch": 10.43, "learning_rate": 1.692861615529117e-05, "loss": 0.579, "step": 12338, "task_loss": 0.8815596699714661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.361400306224823, "epoch": 10.43, "learning_rate": 1.6925485284909206e-05, "loss": 0.4174, "step": 12339, "task_loss": 1.2411081790924072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6631876230239868, "epoch": 10.43, "learning_rate": 1.6922354414527238e-05, "loss": 0.6385, "step": 12340, "task_loss": 0.6884949207305908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3292289674282074, "epoch": 10.43, "learning_rate": 1.6919223544145273e-05, "loss": 0.5125, "step": 12341, "task_loss": 0.42232534289360046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4712413251399994, "epoch": 10.43, "learning_rate": 1.6916092673763305e-05, "loss": 0.5897, "step": 12342, "task_loss": 0.6350299715995789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.616584300994873, "epoch": 10.43, "learning_rate": 1.691296180338134e-05, "loss": 0.5588, "step": 12343, "task_loss": 0.8559237718582153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7787129878997803, "epoch": 10.43, "learning_rate": 1.6909830932999372e-05, "loss": 0.7012, "step": 12344, "task_loss": 0.8923888206481934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28292185068130493, "epoch": 10.44, "learning_rate": 1.6906700062617407e-05, "loss": 0.5195, "step": 12345, "task_loss": 0.8947023153305054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37383949756622314, "epoch": 10.44, "learning_rate": 1.6903569192235442e-05, "loss": 0.4295, "step": 12346, "task_loss": 0.05950541794300079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5276881456375122, "epoch": 10.44, "learning_rate": 1.6900438321853478e-05, "loss": 0.4885, "step": 12347, "task_loss": 0.3524894118309021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6605134010314941, "epoch": 10.44, "learning_rate": 1.689730745147151e-05, "loss": 0.4972, "step": 12348, "task_loss": 1.2421077489852905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7562405467033386, "epoch": 10.44, "learning_rate": 1.6894176581089545e-05, "loss": 0.5262, "step": 12349, "task_loss": 0.8199668526649475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5466618537902832, "epoch": 10.44, "learning_rate": 1.689104571070758e-05, "loss": 0.4892, "step": 12350, "task_loss": 0.6097567081451416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6084431409835815, "epoch": 10.44, "learning_rate": 1.688791484032561e-05, "loss": 0.4177, "step": 12351, "task_loss": 0.4218100905418396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3900826871395111, "epoch": 10.44, "learning_rate": 1.6884783969943647e-05, "loss": 0.6217, "step": 12352, "task_loss": 0.9100607633590698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.270298033952713, "epoch": 10.44, "learning_rate": 1.688165309956168e-05, "loss": 0.6024, "step": 12353, "task_loss": 0.6020806431770325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6140271425247192, "epoch": 10.44, "learning_rate": 1.6878522229179714e-05, "loss": 0.501, "step": 12354, "task_loss": 0.24923928081989288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32782694697380066, "epoch": 10.44, "learning_rate": 1.6875391358797745e-05, "loss": 0.4913, "step": 12355, "task_loss": 0.3300352692604065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24197223782539368, "epoch": 10.44, "learning_rate": 1.687226048841578e-05, "loss": 0.5475, "step": 12356, "task_loss": 0.3087106943130493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8109524846076965, "epoch": 10.45, "learning_rate": 1.6869129618033812e-05, "loss": 0.5033, "step": 12357, "task_loss": 0.5193219184875488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6340799331665039, "epoch": 10.45, "learning_rate": 1.6865998747651848e-05, "loss": 0.5408, "step": 12358, "task_loss": 1.2620888948440552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36062368750572205, "epoch": 10.45, "learning_rate": 1.686286787726988e-05, "loss": 0.4704, "step": 12359, "task_loss": 0.529207170009613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6057044863700867, "epoch": 10.45, "learning_rate": 1.6859737006887914e-05, "loss": 0.4991, "step": 12360, "task_loss": 0.37151288986206055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42819732427597046, "epoch": 10.45, "learning_rate": 1.685660613650595e-05, "loss": 0.4228, "step": 12361, "task_loss": 0.562501847743988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2430780827999115, "epoch": 10.45, "learning_rate": 1.685347526612398e-05, "loss": 0.411, "step": 12362, "task_loss": 0.2594762146472931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45527490973472595, "epoch": 10.45, "learning_rate": 1.6850344395742017e-05, "loss": 0.3936, "step": 12363, "task_loss": 0.2557188868522644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4635499119758606, "epoch": 10.45, "learning_rate": 1.6847213525360052e-05, "loss": 0.3498, "step": 12364, "task_loss": 0.32127276062965393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37802568078041077, "epoch": 10.45, "learning_rate": 1.6844082654978087e-05, "loss": 0.4387, "step": 12365, "task_loss": 0.5516072511672974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3356698155403137, "epoch": 10.45, "learning_rate": 1.684095178459612e-05, "loss": 0.3917, "step": 12366, "task_loss": 0.4234967529773712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4311438202857971, "epoch": 10.45, "learning_rate": 1.6837820914214154e-05, "loss": 0.5367, "step": 12367, "task_loss": 0.763773500919342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6944614052772522, "epoch": 10.45, "learning_rate": 1.6834690043832186e-05, "loss": 0.607, "step": 12368, "task_loss": 1.841718077659607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5214617252349854, "epoch": 10.46, "learning_rate": 1.683155917345022e-05, "loss": 0.7859, "step": 12369, "task_loss": 1.3463878631591797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4892733693122864, "epoch": 10.46, "learning_rate": 1.6828428303068253e-05, "loss": 0.6287, "step": 12370, "task_loss": 0.6725708842277527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22061726450920105, "epoch": 10.46, "learning_rate": 1.6825297432686288e-05, "loss": 0.4793, "step": 12371, "task_loss": 0.09887600690126419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4367928206920624, "epoch": 10.46, "learning_rate": 1.682216656230432e-05, "loss": 0.4979, "step": 12372, "task_loss": 0.6031847596168518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5871248245239258, "epoch": 10.46, "learning_rate": 1.6819035691922355e-05, "loss": 0.3901, "step": 12373, "task_loss": 0.7840461730957031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23544813692569733, "epoch": 10.46, "learning_rate": 1.681590482154039e-05, "loss": 0.3339, "step": 12374, "task_loss": 0.22653529047966003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7365343570709229, "epoch": 10.46, "learning_rate": 1.681277395115842e-05, "loss": 0.5955, "step": 12375, "task_loss": 1.3005489110946655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8637133836746216, "epoch": 10.46, "learning_rate": 1.6809643080776457e-05, "loss": 0.7384, "step": 12376, "task_loss": 0.8265105485916138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5887594223022461, "epoch": 10.46, "learning_rate": 1.680651221039449e-05, "loss": 0.4183, "step": 12377, "task_loss": 0.40939533710479736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4588717818260193, "epoch": 10.46, "learning_rate": 1.6803381340012524e-05, "loss": 0.5709, "step": 12378, "task_loss": 0.4582872986793518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6252527832984924, "epoch": 10.46, "learning_rate": 1.680025046963056e-05, "loss": 0.5188, "step": 12379, "task_loss": 0.3067522943019867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7391865253448486, "epoch": 10.46, "learning_rate": 1.6797119599248594e-05, "loss": 0.5698, "step": 12380, "task_loss": 1.2034697532653809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7498160004615784, "epoch": 10.47, "learning_rate": 1.6793988728866626e-05, "loss": 0.5837, "step": 12381, "task_loss": 1.8775243759155273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26314085721969604, "epoch": 10.47, "learning_rate": 1.679085785848466e-05, "loss": 0.4614, "step": 12382, "task_loss": 0.44063588976860046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37479516863822937, "epoch": 10.47, "learning_rate": 1.6787726988102693e-05, "loss": 0.485, "step": 12383, "task_loss": 0.6873224377632141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47590649127960205, "epoch": 10.47, "learning_rate": 1.6784596117720728e-05, "loss": 0.6664, "step": 12384, "task_loss": 0.605383574962616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19061024487018585, "epoch": 10.47, "learning_rate": 1.6781465247338763e-05, "loss": 0.4273, "step": 12385, "task_loss": 0.03321922942996025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5453975200653076, "epoch": 10.47, "learning_rate": 1.6778334376956795e-05, "loss": 0.5149, "step": 12386, "task_loss": 0.9377149939537048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6122019290924072, "epoch": 10.47, "learning_rate": 1.677520350657483e-05, "loss": 0.5552, "step": 12387, "task_loss": 0.31632745265960693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3554687798023224, "epoch": 10.47, "learning_rate": 1.6772072636192862e-05, "loss": 0.4521, "step": 12388, "task_loss": 0.7773295044898987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43887859582901, "epoch": 10.47, "learning_rate": 1.6768941765810897e-05, "loss": 0.6824, "step": 12389, "task_loss": 0.4230102300643921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6166804432868958, "epoch": 10.47, "learning_rate": 1.676581089542893e-05, "loss": 0.6247, "step": 12390, "task_loss": 1.0344005823135376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8169925212860107, "epoch": 10.47, "learning_rate": 1.6762680025046964e-05, "loss": 0.5779, "step": 12391, "task_loss": 0.686436116695404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5013287663459778, "epoch": 10.47, "learning_rate": 1.6759549154664996e-05, "loss": 0.3963, "step": 12392, "task_loss": 0.8616618514060974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8235589265823364, "epoch": 10.48, "learning_rate": 1.675641828428303e-05, "loss": 0.5805, "step": 12393, "task_loss": 0.5786736607551575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.580960750579834, "epoch": 10.48, "learning_rate": 1.6753287413901063e-05, "loss": 0.4774, "step": 12394, "task_loss": 0.243538960814476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4001111388206482, "epoch": 10.48, "learning_rate": 1.6750156543519098e-05, "loss": 0.5776, "step": 12395, "task_loss": 0.20296069979667664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6293742060661316, "epoch": 10.48, "learning_rate": 1.6747025673137133e-05, "loss": 0.5697, "step": 12396, "task_loss": 1.0582407712936401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.632168710231781, "epoch": 10.48, "learning_rate": 1.6743894802755168e-05, "loss": 0.7271, "step": 12397, "task_loss": 0.27197980880737305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44642215967178345, "epoch": 10.48, "learning_rate": 1.6740763932373203e-05, "loss": 0.3619, "step": 12398, "task_loss": 0.11779285967350006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5842616558074951, "epoch": 10.48, "learning_rate": 1.6737633061991235e-05, "loss": 0.5898, "step": 12399, "task_loss": 1.1564198732376099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5603123307228088, "epoch": 10.48, "learning_rate": 1.673450219160927e-05, "loss": 0.5221, "step": 12400, "task_loss": 1.4576374292373657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1678169071674347, "epoch": 10.48, "learning_rate": 1.6731371321227302e-05, "loss": 0.3794, "step": 12401, "task_loss": 0.04925220087170601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6261307597160339, "epoch": 10.48, "learning_rate": 1.6728240450845337e-05, "loss": 0.5375, "step": 12402, "task_loss": 0.6225045323371887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3058100938796997, "epoch": 10.48, "learning_rate": 1.672510958046337e-05, "loss": 0.5089, "step": 12403, "task_loss": 0.3323676288127899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37875181436538696, "epoch": 10.48, "learning_rate": 1.6721978710081404e-05, "loss": 0.4846, "step": 12404, "task_loss": 0.5216951966285706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48818647861480713, "epoch": 10.49, "learning_rate": 1.6718847839699436e-05, "loss": 0.4141, "step": 12405, "task_loss": 0.41023072600364685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3979291319847107, "epoch": 10.49, "learning_rate": 1.671571696931747e-05, "loss": 0.4168, "step": 12406, "task_loss": 0.33532148599624634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47524505853652954, "epoch": 10.49, "learning_rate": 1.6712586098935503e-05, "loss": 0.3266, "step": 12407, "task_loss": 0.7678389549255371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196292400360107, "epoch": 10.49, "learning_rate": 1.6709455228553538e-05, "loss": 0.5673, "step": 12408, "task_loss": 0.5960555076599121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45745134353637695, "epoch": 10.49, "learning_rate": 1.670632435817157e-05, "loss": 0.656, "step": 12409, "task_loss": 0.693329930305481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5826972126960754, "epoch": 10.49, "learning_rate": 1.6703193487789605e-05, "loss": 0.5243, "step": 12410, "task_loss": 0.7235835194587708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23974579572677612, "epoch": 10.49, "learning_rate": 1.670006261740764e-05, "loss": 0.4786, "step": 12411, "task_loss": 0.6817932724952698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7538391947746277, "epoch": 10.49, "learning_rate": 1.6696931747025672e-05, "loss": 0.4604, "step": 12412, "task_loss": 0.22005829215049744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43159157037734985, "epoch": 10.49, "learning_rate": 1.6693800876643707e-05, "loss": 0.6524, "step": 12413, "task_loss": 0.4137827455997467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3823860287666321, "epoch": 10.49, "learning_rate": 1.6690670006261742e-05, "loss": 0.6504, "step": 12414, "task_loss": 0.5951184034347534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4945065975189209, "epoch": 10.49, "learning_rate": 1.6687539135879778e-05, "loss": 0.4626, "step": 12415, "task_loss": 0.2858210504055023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6572139859199524, "epoch": 10.5, "learning_rate": 1.668440826549781e-05, "loss": 0.6109, "step": 12416, "task_loss": 1.2861640453338623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5393455028533936, "epoch": 10.5, "learning_rate": 1.6681277395115845e-05, "loss": 0.5082, "step": 12417, "task_loss": 1.5057226419448853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3368762135505676, "epoch": 10.5, "learning_rate": 1.6678146524733876e-05, "loss": 0.4521, "step": 12418, "task_loss": 1.3656002283096313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6208356618881226, "epoch": 10.5, "learning_rate": 1.667501565435191e-05, "loss": 0.5359, "step": 12419, "task_loss": 0.654286801815033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3460594415664673, "epoch": 10.5, "learning_rate": 1.6671884783969943e-05, "loss": 0.4715, "step": 12420, "task_loss": 0.8962922692298889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4145790934562683, "epoch": 10.5, "learning_rate": 1.666875391358798e-05, "loss": 0.4598, "step": 12421, "task_loss": 0.6839534044265747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4224816560745239, "epoch": 10.5, "learning_rate": 1.6665623043206014e-05, "loss": 0.5865, "step": 12422, "task_loss": 0.7567062377929688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6888131499290466, "epoch": 10.5, "learning_rate": 1.6662492172824045e-05, "loss": 0.5631, "step": 12423, "task_loss": 0.36458513140678406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36849337816238403, "epoch": 10.5, "learning_rate": 1.665936130244208e-05, "loss": 0.5025, "step": 12424, "task_loss": 0.6194280385971069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43467286229133606, "epoch": 10.5, "learning_rate": 1.6656230432060112e-05, "loss": 0.4613, "step": 12425, "task_loss": 0.6329926252365112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5862665772438049, "epoch": 10.5, "learning_rate": 1.6653099561678148e-05, "loss": 0.6269, "step": 12426, "task_loss": 1.0347108840942383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5354701280593872, "epoch": 10.5, "learning_rate": 1.664996869129618e-05, "loss": 0.6135, "step": 12427, "task_loss": 1.4006736278533936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30947279930114746, "epoch": 10.51, "learning_rate": 1.6646837820914215e-05, "loss": 0.5278, "step": 12428, "task_loss": 0.5917579531669617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7481690049171448, "epoch": 10.51, "learning_rate": 1.6643706950532246e-05, "loss": 0.474, "step": 12429, "task_loss": 1.0044103860855103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4719312787055969, "epoch": 10.51, "learning_rate": 1.664057608015028e-05, "loss": 0.4514, "step": 12430, "task_loss": 0.4199010729789734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33865857124328613, "epoch": 10.51, "learning_rate": 1.6637445209768317e-05, "loss": 0.4176, "step": 12431, "task_loss": 1.1088862419128418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4196971654891968, "epoch": 10.51, "learning_rate": 1.6634314339386352e-05, "loss": 0.5002, "step": 12432, "task_loss": 0.27119994163513184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4561672508716583, "epoch": 10.51, "learning_rate": 1.6631183469004384e-05, "loss": 0.4884, "step": 12433, "task_loss": 0.359623521566391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5810227990150452, "epoch": 10.51, "learning_rate": 1.662805259862242e-05, "loss": 0.5758, "step": 12434, "task_loss": 0.8715783357620239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7542203664779663, "epoch": 10.51, "learning_rate": 1.6624921728240454e-05, "loss": 0.733, "step": 12435, "task_loss": 0.9402740597724915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4289530813694, "epoch": 10.51, "learning_rate": 1.6621790857858486e-05, "loss": 0.5381, "step": 12436, "task_loss": 0.8890682458877563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4514431059360504, "epoch": 10.51, "learning_rate": 1.661865998747652e-05, "loss": 0.4811, "step": 12437, "task_loss": 0.6732566356658936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6131711602210999, "epoch": 10.51, "learning_rate": 1.6615529117094553e-05, "loss": 0.4439, "step": 12438, "task_loss": 0.3249630331993103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46263694763183594, "epoch": 10.51, "learning_rate": 1.6612398246712588e-05, "loss": 0.3753, "step": 12439, "task_loss": 0.07484301179647446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7042872905731201, "epoch": 10.52, "learning_rate": 1.660926737633062e-05, "loss": 0.5859, "step": 12440, "task_loss": 0.9625197052955627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8005270957946777, "epoch": 10.52, "learning_rate": 1.6606136505948655e-05, "loss": 0.6522, "step": 12441, "task_loss": 0.7229359745979309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33988049626350403, "epoch": 10.52, "learning_rate": 1.6603005635566687e-05, "loss": 0.4672, "step": 12442, "task_loss": 0.6471412181854248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33861011266708374, "epoch": 10.52, "learning_rate": 1.6599874765184722e-05, "loss": 0.4691, "step": 12443, "task_loss": 0.19340603053569794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4018016755580902, "epoch": 10.52, "learning_rate": 1.6596743894802753e-05, "loss": 0.4284, "step": 12444, "task_loss": 0.6225508451461792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6568403840065002, "epoch": 10.52, "learning_rate": 1.659361302442079e-05, "loss": 0.4085, "step": 12445, "task_loss": 0.5843892097473145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4032161235809326, "epoch": 10.52, "learning_rate": 1.6590482154038824e-05, "loss": 0.4666, "step": 12446, "task_loss": 1.3815257549285889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4969712495803833, "epoch": 10.52, "learning_rate": 1.658735128365686e-05, "loss": 0.6034, "step": 12447, "task_loss": 0.7115610241889954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.54939866065979, "epoch": 10.52, "learning_rate": 1.658422041327489e-05, "loss": 0.4566, "step": 12448, "task_loss": 0.850694477558136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4757629632949829, "epoch": 10.52, "learning_rate": 1.6581089542892926e-05, "loss": 0.4393, "step": 12449, "task_loss": 0.23263248801231384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.466394305229187, "epoch": 10.52, "learning_rate": 1.657795867251096e-05, "loss": 0.5911, "step": 12450, "task_loss": 0.24758923053741455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5019265413284302, "epoch": 10.52, "learning_rate": 1.6574827802128993e-05, "loss": 0.5765, "step": 12451, "task_loss": 0.4615134298801422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3389222323894501, "epoch": 10.53, "learning_rate": 1.6571696931747028e-05, "loss": 0.525, "step": 12452, "task_loss": 0.5312319993972778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23250240087509155, "epoch": 10.53, "learning_rate": 1.656856606136506e-05, "loss": 0.404, "step": 12453, "task_loss": 0.5139718055725098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24876528978347778, "epoch": 10.53, "learning_rate": 1.6565435190983095e-05, "loss": 0.4686, "step": 12454, "task_loss": 0.19367915391921997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3424587845802307, "epoch": 10.53, "learning_rate": 1.6562304320601127e-05, "loss": 0.5226, "step": 12455, "task_loss": 0.3043158948421478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9089385271072388, "epoch": 10.53, "learning_rate": 1.6559173450219162e-05, "loss": 0.7984, "step": 12456, "task_loss": 1.2227623462677002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3975342810153961, "epoch": 10.53, "learning_rate": 1.6556042579837194e-05, "loss": 0.5098, "step": 12457, "task_loss": 0.7180137038230896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34020814299583435, "epoch": 10.53, "learning_rate": 1.655291170945523e-05, "loss": 0.431, "step": 12458, "task_loss": 0.18310758471488953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44596585631370544, "epoch": 10.53, "learning_rate": 1.6549780839073264e-05, "loss": 0.4831, "step": 12459, "task_loss": 0.39173251390457153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5384502410888672, "epoch": 10.53, "learning_rate": 1.6546649968691296e-05, "loss": 0.4612, "step": 12460, "task_loss": 0.325719952583313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5400130748748779, "epoch": 10.53, "learning_rate": 1.654351909830933e-05, "loss": 0.6355, "step": 12461, "task_loss": 0.5096601247787476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.583581805229187, "epoch": 10.53, "learning_rate": 1.6540388227927363e-05, "loss": 0.5884, "step": 12462, "task_loss": 0.5236307382583618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39938679337501526, "epoch": 10.53, "learning_rate": 1.6537257357545398e-05, "loss": 0.5654, "step": 12463, "task_loss": 0.1562756597995758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4626989960670471, "epoch": 10.54, "learning_rate": 1.6534126487163433e-05, "loss": 0.6831, "step": 12464, "task_loss": 0.7500542998313904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3211621344089508, "epoch": 10.54, "learning_rate": 1.653099561678147e-05, "loss": 0.394, "step": 12465, "task_loss": 0.6239553093910217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2399976253509521, "epoch": 10.54, "learning_rate": 1.65278647463995e-05, "loss": 0.7969, "step": 12466, "task_loss": 2.0350940227508545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32257014513015747, "epoch": 10.54, "learning_rate": 1.6524733876017535e-05, "loss": 0.4137, "step": 12467, "task_loss": 0.47391051054000854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6611876487731934, "epoch": 10.54, "learning_rate": 1.6521603005635567e-05, "loss": 0.6876, "step": 12468, "task_loss": 1.2143133878707886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29932701587677, "epoch": 10.54, "learning_rate": 1.6518472135253602e-05, "loss": 0.4752, "step": 12469, "task_loss": 0.8436862230300903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4649656414985657, "epoch": 10.54, "learning_rate": 1.6515341264871634e-05, "loss": 0.3899, "step": 12470, "task_loss": 0.25959697365760803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4533727765083313, "epoch": 10.54, "learning_rate": 1.651221039448967e-05, "loss": 0.5469, "step": 12471, "task_loss": 0.9056336879730225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5869718194007874, "epoch": 10.54, "learning_rate": 1.6509079524107704e-05, "loss": 0.5139, "step": 12472, "task_loss": 0.25065815448760986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5825937986373901, "epoch": 10.54, "learning_rate": 1.6505948653725736e-05, "loss": 0.4638, "step": 12473, "task_loss": 0.08579545468091965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.342678964138031, "epoch": 10.54, "learning_rate": 1.650281778334377e-05, "loss": 0.461, "step": 12474, "task_loss": 0.05556102842092514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6571763753890991, "epoch": 10.54, "learning_rate": 1.6499686912961803e-05, "loss": 0.5812, "step": 12475, "task_loss": 2.437140941619873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39404043555259705, "epoch": 10.55, "learning_rate": 1.6496556042579838e-05, "loss": 0.4299, "step": 12476, "task_loss": 0.3183479905128479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49310359358787537, "epoch": 10.55, "learning_rate": 1.649342517219787e-05, "loss": 0.4631, "step": 12477, "task_loss": 0.48143577575683594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.441850870847702, "epoch": 10.55, "learning_rate": 1.6490294301815905e-05, "loss": 0.4535, "step": 12478, "task_loss": 1.066569209098816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43468013405799866, "epoch": 10.55, "learning_rate": 1.6487163431433937e-05, "loss": 0.4672, "step": 12479, "task_loss": 1.024166226387024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6731959581375122, "epoch": 10.55, "learning_rate": 1.6484032561051972e-05, "loss": 0.4931, "step": 12480, "task_loss": 1.2129143476486206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49285516142845154, "epoch": 10.55, "learning_rate": 1.6480901690670007e-05, "loss": 0.6043, "step": 12481, "task_loss": 0.2416854202747345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3966233730316162, "epoch": 10.55, "learning_rate": 1.6477770820288042e-05, "loss": 0.3995, "step": 12482, "task_loss": 0.7882152795791626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6706939935684204, "epoch": 10.55, "learning_rate": 1.6474639949906078e-05, "loss": 0.5156, "step": 12483, "task_loss": 0.19471681118011475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6205955147743225, "epoch": 10.55, "learning_rate": 1.647150907952411e-05, "loss": 0.5429, "step": 12484, "task_loss": 0.6573511958122253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7065221667289734, "epoch": 10.55, "learning_rate": 1.6468378209142145e-05, "loss": 0.4686, "step": 12485, "task_loss": 0.9317858815193176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4714357256889343, "epoch": 10.55, "learning_rate": 1.6465247338760176e-05, "loss": 0.4603, "step": 12486, "task_loss": 0.831044614315033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7561823129653931, "epoch": 10.56, "learning_rate": 1.646211646837821e-05, "loss": 0.5351, "step": 12487, "task_loss": 1.2393789291381836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6883769035339355, "epoch": 10.56, "learning_rate": 1.6458985597996243e-05, "loss": 0.6023, "step": 12488, "task_loss": 1.1642683744430542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6906355619430542, "epoch": 10.56, "learning_rate": 1.645585472761428e-05, "loss": 0.4447, "step": 12489, "task_loss": 0.8214801549911499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4863533675670624, "epoch": 10.56, "learning_rate": 1.645272385723231e-05, "loss": 0.5618, "step": 12490, "task_loss": 0.5474005937576294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3708077371120453, "epoch": 10.56, "learning_rate": 1.6449592986850345e-05, "loss": 0.3881, "step": 12491, "task_loss": 0.04293157532811165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1870422512292862, "epoch": 10.56, "learning_rate": 1.6446462116468377e-05, "loss": 0.4339, "step": 12492, "task_loss": 1.0652281045913696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47413772344589233, "epoch": 10.56, "learning_rate": 1.6443331246086412e-05, "loss": 0.4284, "step": 12493, "task_loss": 0.6684250235557556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4920833706855774, "epoch": 10.56, "learning_rate": 1.6440200375704444e-05, "loss": 0.6688, "step": 12494, "task_loss": 1.265198826789856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7561609745025635, "epoch": 10.56, "learning_rate": 1.643706950532248e-05, "loss": 0.5824, "step": 12495, "task_loss": 0.815090537071228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.520534098148346, "epoch": 10.56, "learning_rate": 1.6433938634940515e-05, "loss": 0.4661, "step": 12496, "task_loss": 0.5134177207946777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4095068871974945, "epoch": 10.56, "learning_rate": 1.6430807764558546e-05, "loss": 0.6616, "step": 12497, "task_loss": 0.6808648705482483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4627792239189148, "epoch": 10.56, "learning_rate": 1.642767689417658e-05, "loss": 0.6115, "step": 12498, "task_loss": 0.1446375995874405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.473659873008728, "epoch": 10.57, "learning_rate": 1.6424546023794617e-05, "loss": 0.3766, "step": 12499, "task_loss": 0.3509414792060852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6995757818222046, "epoch": 10.57, "learning_rate": 1.6421415153412652e-05, "loss": 0.7103, "step": 12500, "task_loss": 0.5207656621932983 }, { "epoch": 10.57, "eval_accuracy": 0.9070495049504951, "eval_loss": 0.34786149859428406, "eval_runtime": 207.0384, "eval_samples_per_second": 121.958, "eval_steps_per_second": 0.956, "step": 12500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2861328423023224, "epoch": 10.57, "learning_rate": 1.6418284283030684e-05, "loss": 0.3611, "step": 12501, "task_loss": 0.34124261140823364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5139725804328918, "epoch": 10.57, "learning_rate": 1.641515341264872e-05, "loss": 0.5088, "step": 12502, "task_loss": 0.7876615524291992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4810686707496643, "epoch": 10.57, "learning_rate": 1.641202254226675e-05, "loss": 0.5569, "step": 12503, "task_loss": 1.2529563903808594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8265133500099182, "epoch": 10.57, "learning_rate": 1.6408891671884786e-05, "loss": 0.6789, "step": 12504, "task_loss": 1.3073722124099731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46941909193992615, "epoch": 10.57, "learning_rate": 1.6405760801502818e-05, "loss": 0.5017, "step": 12505, "task_loss": 0.6680939197540283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2248345613479614, "epoch": 10.57, "learning_rate": 1.6402629931120853e-05, "loss": 0.8395, "step": 12506, "task_loss": 1.0900497436523438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28609877824783325, "epoch": 10.57, "learning_rate": 1.6399499060738884e-05, "loss": 0.4563, "step": 12507, "task_loss": 0.3548286557197571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3833252489566803, "epoch": 10.57, "learning_rate": 1.639636819035692e-05, "loss": 0.4372, "step": 12508, "task_loss": 2.3745932579040527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23086440563201904, "epoch": 10.57, "learning_rate": 1.6393237319974955e-05, "loss": 0.3782, "step": 12509, "task_loss": 0.10303862392902374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5240297317504883, "epoch": 10.57, "learning_rate": 1.6390106449592987e-05, "loss": 0.5703, "step": 12510, "task_loss": 0.7920283079147339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3606717586517334, "epoch": 10.58, "learning_rate": 1.6386975579211022e-05, "loss": 0.6783, "step": 12511, "task_loss": 0.3953629434108734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3493741750717163, "epoch": 10.58, "learning_rate": 1.6383844708829054e-05, "loss": 0.4715, "step": 12512, "task_loss": 0.6332336664199829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37329548597335815, "epoch": 10.58, "learning_rate": 1.638071383844709e-05, "loss": 0.436, "step": 12513, "task_loss": 0.07807647436857224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.2408645153045654, "epoch": 10.58, "learning_rate": 1.637758296806512e-05, "loss": 0.6501, "step": 12514, "task_loss": 1.0746381282806396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4683522582054138, "epoch": 10.58, "learning_rate": 1.6374452097683156e-05, "loss": 0.479, "step": 12515, "task_loss": 0.7761314511299133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27030545473098755, "epoch": 10.58, "learning_rate": 1.637132122730119e-05, "loss": 0.3533, "step": 12516, "task_loss": 0.05047089606523514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3204413652420044, "epoch": 10.58, "learning_rate": 1.6368190356919226e-05, "loss": 0.4218, "step": 12517, "task_loss": 1.0924910306930542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37287628650665283, "epoch": 10.58, "learning_rate": 1.6365059486537258e-05, "loss": 0.4323, "step": 12518, "task_loss": 0.7428169250488281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33018919825553894, "epoch": 10.58, "learning_rate": 1.6361928616155293e-05, "loss": 0.4354, "step": 12519, "task_loss": 0.21316403150558472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6205664277076721, "epoch": 10.58, "learning_rate": 1.6358797745773328e-05, "loss": 0.5032, "step": 12520, "task_loss": 0.6995615363121033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5895015001296997, "epoch": 10.58, "learning_rate": 1.635566687539136e-05, "loss": 0.4216, "step": 12521, "task_loss": 0.6666021943092346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2809727191925049, "epoch": 10.58, "learning_rate": 1.6352536005009395e-05, "loss": 0.3883, "step": 12522, "task_loss": 0.5344029068946838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5009722709655762, "epoch": 10.59, "learning_rate": 1.6349405134627427e-05, "loss": 0.4795, "step": 12523, "task_loss": 0.6029967069625854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3129439353942871, "epoch": 10.59, "learning_rate": 1.6346274264245462e-05, "loss": 0.4953, "step": 12524, "task_loss": 0.39242228865623474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5202165842056274, "epoch": 10.59, "learning_rate": 1.6343143393863494e-05, "loss": 0.4895, "step": 12525, "task_loss": 0.7415745854377747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3917474150657654, "epoch": 10.59, "learning_rate": 1.634001252348153e-05, "loss": 0.5412, "step": 12526, "task_loss": 0.2776564657688141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6139740943908691, "epoch": 10.59, "learning_rate": 1.633688165309956e-05, "loss": 0.5476, "step": 12527, "task_loss": 0.5857943296432495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5889304280281067, "epoch": 10.59, "learning_rate": 1.6333750782717596e-05, "loss": 0.4843, "step": 12528, "task_loss": 1.204032301902771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49247246980667114, "epoch": 10.59, "learning_rate": 1.6330619912335628e-05, "loss": 0.5672, "step": 12529, "task_loss": 0.7264295816421509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5164080858230591, "epoch": 10.59, "learning_rate": 1.6327489041953663e-05, "loss": 0.3882, "step": 12530, "task_loss": 0.6496472358703613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6095845699310303, "epoch": 10.59, "learning_rate": 1.6324358171571698e-05, "loss": 0.6249, "step": 12531, "task_loss": 0.5765606164932251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5468679666519165, "epoch": 10.59, "learning_rate": 1.6321227301189733e-05, "loss": 0.5837, "step": 12532, "task_loss": 1.0722204446792603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5768659114837646, "epoch": 10.59, "learning_rate": 1.6318096430807765e-05, "loss": 0.5371, "step": 12533, "task_loss": 1.1232208013534546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2289595603942871, "epoch": 10.59, "learning_rate": 1.63149655604258e-05, "loss": 0.4454, "step": 12534, "task_loss": 0.050541043281555176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3683522045612335, "epoch": 10.6, "learning_rate": 1.6311834690043835e-05, "loss": 0.6059, "step": 12535, "task_loss": 0.525209367275238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5492563843727112, "epoch": 10.6, "learning_rate": 1.6308703819661867e-05, "loss": 0.4557, "step": 12536, "task_loss": 0.48103901743888855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6000148057937622, "epoch": 10.6, "learning_rate": 1.6305572949279902e-05, "loss": 0.5791, "step": 12537, "task_loss": 1.633821964263916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4703344702720642, "epoch": 10.6, "learning_rate": 1.6302442078897934e-05, "loss": 0.5038, "step": 12538, "task_loss": 0.4577834904193878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7112976312637329, "epoch": 10.6, "learning_rate": 1.629931120851597e-05, "loss": 0.5568, "step": 12539, "task_loss": 0.4450581669807434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5963632464408875, "epoch": 10.6, "learning_rate": 1.6296180338134e-05, "loss": 0.4808, "step": 12540, "task_loss": 1.0287545919418335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32339295744895935, "epoch": 10.6, "learning_rate": 1.6293049467752036e-05, "loss": 0.5293, "step": 12541, "task_loss": 0.22249051928520203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.347109317779541, "epoch": 10.6, "learning_rate": 1.6289918597370068e-05, "loss": 0.5077, "step": 12542, "task_loss": 1.0345464944839478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6882046461105347, "epoch": 10.6, "learning_rate": 1.6286787726988103e-05, "loss": 0.5515, "step": 12543, "task_loss": 1.4804680347442627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8051337003707886, "epoch": 10.6, "learning_rate": 1.6283656856606135e-05, "loss": 0.5332, "step": 12544, "task_loss": 1.3080577850341797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5574314594268799, "epoch": 10.6, "learning_rate": 1.628052598622417e-05, "loss": 0.5071, "step": 12545, "task_loss": 0.5508299469947815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4585030972957611, "epoch": 10.6, "learning_rate": 1.6277395115842205e-05, "loss": 0.6157, "step": 12546, "task_loss": 0.3279881775379181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5677977800369263, "epoch": 10.61, "learning_rate": 1.6274264245460237e-05, "loss": 0.499, "step": 12547, "task_loss": 0.31834253668785095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31841742992401123, "epoch": 10.61, "learning_rate": 1.6271133375078272e-05, "loss": 0.5336, "step": 12548, "task_loss": 0.37206417322158813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3097881078720093, "epoch": 10.61, "learning_rate": 1.6268002504696307e-05, "loss": 0.3776, "step": 12549, "task_loss": 0.5613054037094116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8546615839004517, "epoch": 10.61, "learning_rate": 1.6264871634314343e-05, "loss": 0.6551, "step": 12550, "task_loss": 1.2674899101257324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5590798258781433, "epoch": 10.61, "learning_rate": 1.6261740763932374e-05, "loss": 0.5613, "step": 12551, "task_loss": 0.7239428758621216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34411531686782837, "epoch": 10.61, "learning_rate": 1.625860989355041e-05, "loss": 0.3862, "step": 12552, "task_loss": 0.499724805355072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4876139163970947, "epoch": 10.61, "learning_rate": 1.625547902316844e-05, "loss": 0.4343, "step": 12553, "task_loss": 0.4872790277004242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7135711908340454, "epoch": 10.61, "learning_rate": 1.6252348152786476e-05, "loss": 0.4646, "step": 12554, "task_loss": 1.284379005432129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49208563566207886, "epoch": 10.61, "learning_rate": 1.6249217282404508e-05, "loss": 0.5495, "step": 12555, "task_loss": 0.32902318239212036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5395650267601013, "epoch": 10.61, "learning_rate": 1.6246086412022543e-05, "loss": 0.5796, "step": 12556, "task_loss": 0.8943939208984375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5176894664764404, "epoch": 10.61, "learning_rate": 1.624295554164058e-05, "loss": 0.4812, "step": 12557, "task_loss": 0.47635164856910706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3501155376434326, "epoch": 10.61, "learning_rate": 1.623982467125861e-05, "loss": 0.4119, "step": 12558, "task_loss": 0.2912565767765045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35782596468925476, "epoch": 10.62, "learning_rate": 1.6236693800876646e-05, "loss": 0.5457, "step": 12559, "task_loss": 0.6253226399421692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40256375074386597, "epoch": 10.62, "learning_rate": 1.6233562930494677e-05, "loss": 0.4386, "step": 12560, "task_loss": 0.572351336479187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3684041202068329, "epoch": 10.62, "learning_rate": 1.6230432060112712e-05, "loss": 0.5387, "step": 12561, "task_loss": 0.35365554690361023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34442389011383057, "epoch": 10.62, "learning_rate": 1.6227301189730744e-05, "loss": 0.5833, "step": 12562, "task_loss": 0.3562467396259308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3451375961303711, "epoch": 10.62, "learning_rate": 1.622417031934878e-05, "loss": 0.4236, "step": 12563, "task_loss": 0.4752061069011688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29523104429244995, "epoch": 10.62, "learning_rate": 1.622103944896681e-05, "loss": 0.4188, "step": 12564, "task_loss": 0.48665502667427063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6154934763908386, "epoch": 10.62, "learning_rate": 1.6217908578584846e-05, "loss": 0.4799, "step": 12565, "task_loss": 0.5292164087295532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3814394176006317, "epoch": 10.62, "learning_rate": 1.621477770820288e-05, "loss": 0.4675, "step": 12566, "task_loss": 0.30852413177490234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5842379331588745, "epoch": 10.62, "learning_rate": 1.6211646837820917e-05, "loss": 0.4801, "step": 12567, "task_loss": 0.12386069446802139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38717231154441833, "epoch": 10.62, "learning_rate": 1.620851596743895e-05, "loss": 0.3631, "step": 12568, "task_loss": 0.08734671026468277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6141087412834167, "epoch": 10.62, "learning_rate": 1.6205385097056984e-05, "loss": 0.4802, "step": 12569, "task_loss": 0.8781656622886658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5381922125816345, "epoch": 10.63, "learning_rate": 1.620225422667502e-05, "loss": 0.5335, "step": 12570, "task_loss": 0.24941767752170563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4967902898788452, "epoch": 10.63, "learning_rate": 1.619912335629305e-05, "loss": 0.4833, "step": 12571, "task_loss": 1.5853568315505981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6852172613143921, "epoch": 10.63, "learning_rate": 1.6195992485911086e-05, "loss": 0.6376, "step": 12572, "task_loss": 0.4055042862892151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34201812744140625, "epoch": 10.63, "learning_rate": 1.6192861615529118e-05, "loss": 0.3898, "step": 12573, "task_loss": 0.5251060724258423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5751321315765381, "epoch": 10.63, "learning_rate": 1.6189730745147153e-05, "loss": 0.5746, "step": 12574, "task_loss": 1.4724040031433105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4306270480155945, "epoch": 10.63, "learning_rate": 1.6186599874765184e-05, "loss": 0.6117, "step": 12575, "task_loss": 0.09865272790193558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4577740430831909, "epoch": 10.63, "learning_rate": 1.618346900438322e-05, "loss": 0.482, "step": 12576, "task_loss": 0.48857542872428894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4532328248023987, "epoch": 10.63, "learning_rate": 1.618033813400125e-05, "loss": 0.6236, "step": 12577, "task_loss": 1.4212743043899536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5893839597702026, "epoch": 10.63, "learning_rate": 1.6177207263619287e-05, "loss": 0.4946, "step": 12578, "task_loss": 0.0665440484881401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3268735706806183, "epoch": 10.63, "learning_rate": 1.617407639323732e-05, "loss": 0.4151, "step": 12579, "task_loss": 0.7247901558876038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5969240665435791, "epoch": 10.63, "learning_rate": 1.6170945522855354e-05, "loss": 0.4451, "step": 12580, "task_loss": 0.8419508337974548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27416908740997314, "epoch": 10.63, "learning_rate": 1.6167814652473385e-05, "loss": 0.4062, "step": 12581, "task_loss": 0.39180251955986023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41473904252052307, "epoch": 10.64, "learning_rate": 1.616468378209142e-05, "loss": 0.3042, "step": 12582, "task_loss": 0.32340526580810547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.461895227432251, "epoch": 10.64, "learning_rate": 1.6161552911709456e-05, "loss": 0.3852, "step": 12583, "task_loss": 0.8038643002510071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5679993629455566, "epoch": 10.64, "learning_rate": 1.615842204132749e-05, "loss": 0.7359, "step": 12584, "task_loss": 2.281252861022949 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5397638082504272, "epoch": 10.64, "learning_rate": 1.6155291170945526e-05, "loss": 0.518, "step": 12585, "task_loss": 0.7590453028678894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36722859740257263, "epoch": 10.64, "learning_rate": 1.6152160300563558e-05, "loss": 0.6128, "step": 12586, "task_loss": 0.3510221540927887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28357410430908203, "epoch": 10.64, "learning_rate": 1.6149029430181593e-05, "loss": 0.412, "step": 12587, "task_loss": 0.9911136627197266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7996985912322998, "epoch": 10.64, "learning_rate": 1.6145898559799625e-05, "loss": 0.5826, "step": 12588, "task_loss": 1.1715441942214966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6446834206581116, "epoch": 10.64, "learning_rate": 1.614276768941766e-05, "loss": 0.5655, "step": 12589, "task_loss": 0.7605506181716919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4817429482936859, "epoch": 10.64, "learning_rate": 1.6139636819035692e-05, "loss": 0.5632, "step": 12590, "task_loss": 1.0661756992340088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2835471034049988, "epoch": 10.64, "learning_rate": 1.6136505948653727e-05, "loss": 0.425, "step": 12591, "task_loss": 0.8421638607978821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5506187677383423, "epoch": 10.64, "learning_rate": 1.613337507827176e-05, "loss": 0.5657, "step": 12592, "task_loss": 0.4235389828681946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42375609278678894, "epoch": 10.64, "learning_rate": 1.6130244207889794e-05, "loss": 0.5406, "step": 12593, "task_loss": 1.0906567573547363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9454327821731567, "epoch": 10.65, "learning_rate": 1.612711333750783e-05, "loss": 0.5595, "step": 12594, "task_loss": 1.1610134840011597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5074527263641357, "epoch": 10.65, "learning_rate": 1.612398246712586e-05, "loss": 0.4339, "step": 12595, "task_loss": 0.8927974104881287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4488847255706787, "epoch": 10.65, "learning_rate": 1.6120851596743896e-05, "loss": 0.5908, "step": 12596, "task_loss": 1.194037914276123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31962496042251587, "epoch": 10.65, "learning_rate": 1.6117720726361928e-05, "loss": 0.5822, "step": 12597, "task_loss": 1.0872044563293457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2551916539669037, "epoch": 10.65, "learning_rate": 1.6114589855979963e-05, "loss": 0.4051, "step": 12598, "task_loss": 0.3835928738117218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7854132056236267, "epoch": 10.65, "learning_rate": 1.6111458985597998e-05, "loss": 0.5752, "step": 12599, "task_loss": 0.9586427807807922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6123063564300537, "epoch": 10.65, "learning_rate": 1.610832811521603e-05, "loss": 0.5705, "step": 12600, "task_loss": 0.4925985336303711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5782277584075928, "epoch": 10.65, "learning_rate": 1.6105197244834065e-05, "loss": 0.5251, "step": 12601, "task_loss": 0.6388105154037476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3188694417476654, "epoch": 10.65, "learning_rate": 1.61020663744521e-05, "loss": 0.438, "step": 12602, "task_loss": 0.4961451590061188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31207460165023804, "epoch": 10.65, "learning_rate": 1.6098935504070132e-05, "loss": 0.3329, "step": 12603, "task_loss": 0.2275175005197525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3878685534000397, "epoch": 10.65, "learning_rate": 1.6095804633688167e-05, "loss": 0.4562, "step": 12604, "task_loss": 0.1988787204027176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6456184387207031, "epoch": 10.65, "learning_rate": 1.60926737633062e-05, "loss": 0.56, "step": 12605, "task_loss": 0.8922989964485168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37201541662216187, "epoch": 10.66, "learning_rate": 1.6089542892924234e-05, "loss": 0.4109, "step": 12606, "task_loss": 0.5108391642570496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5496466159820557, "epoch": 10.66, "learning_rate": 1.608641202254227e-05, "loss": 0.4844, "step": 12607, "task_loss": 0.8435958027839661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7257152795791626, "epoch": 10.66, "learning_rate": 1.60832811521603e-05, "loss": 0.6268, "step": 12608, "task_loss": 1.7097502946853638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4024409055709839, "epoch": 10.66, "learning_rate": 1.6080150281778336e-05, "loss": 0.4858, "step": 12609, "task_loss": 0.5880594849586487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6264081597328186, "epoch": 10.66, "learning_rate": 1.6077019411396368e-05, "loss": 0.5096, "step": 12610, "task_loss": 0.873862624168396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38072827458381653, "epoch": 10.66, "learning_rate": 1.6073888541014403e-05, "loss": 0.3987, "step": 12611, "task_loss": 0.6321852803230286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7391407489776611, "epoch": 10.66, "learning_rate": 1.6070757670632435e-05, "loss": 0.5124, "step": 12612, "task_loss": 1.9026026725769043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5443798899650574, "epoch": 10.66, "learning_rate": 1.606762680025047e-05, "loss": 0.4388, "step": 12613, "task_loss": 0.6641151309013367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7975199818611145, "epoch": 10.66, "learning_rate": 1.6064495929868502e-05, "loss": 0.5572, "step": 12614, "task_loss": 1.2658989429473877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4005284309387207, "epoch": 10.66, "learning_rate": 1.6061365059486537e-05, "loss": 0.5544, "step": 12615, "task_loss": 0.20058459043502808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4770967364311218, "epoch": 10.66, "learning_rate": 1.6058234189104572e-05, "loss": 0.4882, "step": 12616, "task_loss": 0.3853207528591156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3428480327129364, "epoch": 10.66, "learning_rate": 1.6055103318722607e-05, "loss": 0.5203, "step": 12617, "task_loss": 0.6829301714897156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4370634853839874, "epoch": 10.67, "learning_rate": 1.605197244834064e-05, "loss": 0.3875, "step": 12618, "task_loss": 0.3026348352432251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39223241806030273, "epoch": 10.67, "learning_rate": 1.6048841577958674e-05, "loss": 0.6264, "step": 12619, "task_loss": 0.5128450989723206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5615884065628052, "epoch": 10.67, "learning_rate": 1.604571070757671e-05, "loss": 0.5116, "step": 12620, "task_loss": 1.1331369876861572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.235480397939682, "epoch": 10.67, "learning_rate": 1.604257983719474e-05, "loss": 0.6773, "step": 12621, "task_loss": 0.40588831901550293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39801689982414246, "epoch": 10.67, "learning_rate": 1.6039448966812776e-05, "loss": 0.4313, "step": 12622, "task_loss": 0.6402925848960876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4310901165008545, "epoch": 10.67, "learning_rate": 1.6036318096430808e-05, "loss": 0.5252, "step": 12623, "task_loss": 0.8352446556091309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4068133234977722, "epoch": 10.67, "learning_rate": 1.6033187226048843e-05, "loss": 0.5411, "step": 12624, "task_loss": 0.5497060418128967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33963584899902344, "epoch": 10.67, "learning_rate": 1.6030056355666875e-05, "loss": 0.4321, "step": 12625, "task_loss": 0.10877814888954163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6569900512695312, "epoch": 10.67, "learning_rate": 1.602692548528491e-05, "loss": 0.5171, "step": 12626, "task_loss": 2.0448968410491943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43728017807006836, "epoch": 10.67, "learning_rate": 1.6023794614902942e-05, "loss": 0.5024, "step": 12627, "task_loss": 1.0965694189071655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3616366982460022, "epoch": 10.67, "learning_rate": 1.6020663744520977e-05, "loss": 0.5341, "step": 12628, "task_loss": 0.3615300953388214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6937516927719116, "epoch": 10.67, "learning_rate": 1.601753287413901e-05, "loss": 0.477, "step": 12629, "task_loss": 0.43811777234077454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30358433723449707, "epoch": 10.68, "learning_rate": 1.6014402003757044e-05, "loss": 0.4283, "step": 12630, "task_loss": 0.5912361145019531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.481310099363327, "epoch": 10.68, "learning_rate": 1.601127113337508e-05, "loss": 0.5281, "step": 12631, "task_loss": 0.6369727253913879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35946133732795715, "epoch": 10.68, "learning_rate": 1.600814026299311e-05, "loss": 0.422, "step": 12632, "task_loss": 0.9086353182792664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4985516667366028, "epoch": 10.68, "learning_rate": 1.6005009392611146e-05, "loss": 0.5219, "step": 12633, "task_loss": 0.3115430176258087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45731836557388306, "epoch": 10.68, "learning_rate": 1.600187852222918e-05, "loss": 0.4998, "step": 12634, "task_loss": 0.42969146370887756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5902830958366394, "epoch": 10.68, "learning_rate": 1.5998747651847217e-05, "loss": 0.5129, "step": 12635, "task_loss": 0.8777862787246704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30888649821281433, "epoch": 10.68, "learning_rate": 1.599561678146525e-05, "loss": 0.5766, "step": 12636, "task_loss": 0.3693099021911621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6325458884239197, "epoch": 10.68, "learning_rate": 1.5992485911083284e-05, "loss": 0.586, "step": 12637, "task_loss": 0.7120633721351624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5283185243606567, "epoch": 10.68, "learning_rate": 1.5989355040701315e-05, "loss": 0.4827, "step": 12638, "task_loss": 0.6927450299263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4871745705604553, "epoch": 10.68, "learning_rate": 1.598622417031935e-05, "loss": 0.5261, "step": 12639, "task_loss": 0.6944361329078674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3359236419200897, "epoch": 10.68, "learning_rate": 1.5983093299937382e-05, "loss": 0.417, "step": 12640, "task_loss": 0.3529532849788666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3066575527191162, "epoch": 10.69, "learning_rate": 1.5979962429555418e-05, "loss": 0.3812, "step": 12641, "task_loss": 0.332918256521225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5837321281433105, "epoch": 10.69, "learning_rate": 1.597683155917345e-05, "loss": 0.5357, "step": 12642, "task_loss": 0.3250937759876251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43042802810668945, "epoch": 10.69, "learning_rate": 1.5973700688791485e-05, "loss": 0.503, "step": 12643, "task_loss": 0.34715455770492554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.398481547832489, "epoch": 10.69, "learning_rate": 1.597056981840952e-05, "loss": 0.4371, "step": 12644, "task_loss": 0.49270549416542053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5936612486839294, "epoch": 10.69, "learning_rate": 1.596743894802755e-05, "loss": 0.5119, "step": 12645, "task_loss": 0.3591997027397156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5401766896247864, "epoch": 10.69, "learning_rate": 1.5964308077645587e-05, "loss": 0.6036, "step": 12646, "task_loss": 0.7899668216705322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.423321396112442, "epoch": 10.69, "learning_rate": 1.596117720726362e-05, "loss": 0.5489, "step": 12647, "task_loss": 0.1586896777153015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5242501497268677, "epoch": 10.69, "learning_rate": 1.5958046336881654e-05, "loss": 0.4562, "step": 12648, "task_loss": 0.19388923048973083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5718816518783569, "epoch": 10.69, "learning_rate": 1.5954915466499685e-05, "loss": 0.5917, "step": 12649, "task_loss": 0.999021053314209 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2629321217536926, "epoch": 10.69, "learning_rate": 1.595178459611772e-05, "loss": 0.3966, "step": 12650, "task_loss": 0.16791029274463654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8443222045898438, "epoch": 10.69, "learning_rate": 1.5948653725735756e-05, "loss": 0.5247, "step": 12651, "task_loss": 0.8943113684654236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2902829349040985, "epoch": 10.69, "learning_rate": 1.594552285535379e-05, "loss": 0.4336, "step": 12652, "task_loss": 0.13428466022014618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3922235369682312, "epoch": 10.7, "learning_rate": 1.5942391984971823e-05, "loss": 0.4971, "step": 12653, "task_loss": 0.912013590335846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2770349383354187, "epoch": 10.7, "learning_rate": 1.5939261114589858e-05, "loss": 0.4769, "step": 12654, "task_loss": 0.36757203936576843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7118825912475586, "epoch": 10.7, "learning_rate": 1.5936130244207893e-05, "loss": 0.5858, "step": 12655, "task_loss": 0.29926663637161255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6759910583496094, "epoch": 10.7, "learning_rate": 1.5932999373825925e-05, "loss": 0.5617, "step": 12656, "task_loss": 1.2734037637710571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7953360080718994, "epoch": 10.7, "learning_rate": 1.592986850344396e-05, "loss": 0.7239, "step": 12657, "task_loss": 0.699284017086029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5727654099464417, "epoch": 10.7, "learning_rate": 1.5926737633061992e-05, "loss": 0.4653, "step": 12658, "task_loss": 0.17283675074577332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43393030762672424, "epoch": 10.7, "learning_rate": 1.5923606762680027e-05, "loss": 0.4456, "step": 12659, "task_loss": 0.45999735593795776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.551521360874176, "epoch": 10.7, "learning_rate": 1.592047589229806e-05, "loss": 0.5656, "step": 12660, "task_loss": 0.09617049992084503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7093237638473511, "epoch": 10.7, "learning_rate": 1.5917345021916094e-05, "loss": 0.7214, "step": 12661, "task_loss": 0.8017275333404541 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2655280828475952, "epoch": 10.7, "learning_rate": 1.5914214151534126e-05, "loss": 0.5342, "step": 12662, "task_loss": 0.034801580011844635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6086317300796509, "epoch": 10.7, "learning_rate": 1.591108328115216e-05, "loss": 0.5442, "step": 12663, "task_loss": 1.1232686042785645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4158373475074768, "epoch": 10.7, "learning_rate": 1.5907952410770193e-05, "loss": 0.4986, "step": 12664, "task_loss": 0.7418020963668823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6554831266403198, "epoch": 10.71, "learning_rate": 1.5904821540388228e-05, "loss": 0.5185, "step": 12665, "task_loss": 1.3102948665618896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3684961199760437, "epoch": 10.71, "learning_rate": 1.590169067000626e-05, "loss": 0.4628, "step": 12666, "task_loss": 0.6874048113822937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3481767773628235, "epoch": 10.71, "learning_rate": 1.5898559799624295e-05, "loss": 0.5302, "step": 12667, "task_loss": 0.8043579459190369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9640920162200928, "epoch": 10.71, "learning_rate": 1.589542892924233e-05, "loss": 0.6812, "step": 12668, "task_loss": 0.5991837978363037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8718572854995728, "epoch": 10.71, "learning_rate": 1.5892298058860365e-05, "loss": 0.5615, "step": 12669, "task_loss": 0.48388737440109253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5988047122955322, "epoch": 10.71, "learning_rate": 1.58891671884784e-05, "loss": 0.4958, "step": 12670, "task_loss": 0.8530166149139404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5657098293304443, "epoch": 10.71, "learning_rate": 1.5886036318096432e-05, "loss": 0.4523, "step": 12671, "task_loss": 0.9587454795837402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8781705498695374, "epoch": 10.71, "learning_rate": 1.5882905447714467e-05, "loss": 0.4925, "step": 12672, "task_loss": 1.0063350200653076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36614736914634705, "epoch": 10.71, "learning_rate": 1.58797745773325e-05, "loss": 0.421, "step": 12673, "task_loss": 0.3910221755504608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3624352216720581, "epoch": 10.71, "learning_rate": 1.5876643706950534e-05, "loss": 0.4506, "step": 12674, "task_loss": 1.1374804973602295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5085663199424744, "epoch": 10.71, "learning_rate": 1.5873512836568566e-05, "loss": 0.4852, "step": 12675, "task_loss": 0.9606655836105347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5956592559814453, "epoch": 10.71, "learning_rate": 1.58703819661866e-05, "loss": 0.5551, "step": 12676, "task_loss": 1.387351155281067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4529988467693329, "epoch": 10.72, "learning_rate": 1.5867251095804633e-05, "loss": 0.5139, "step": 12677, "task_loss": 0.3867698013782501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3248949944972992, "epoch": 10.72, "learning_rate": 1.5864120225422668e-05, "loss": 0.4794, "step": 12678, "task_loss": 0.3950401842594147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33228158950805664, "epoch": 10.72, "learning_rate": 1.58609893550407e-05, "loss": 0.3558, "step": 12679, "task_loss": 0.38830074667930603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6756575703620911, "epoch": 10.72, "learning_rate": 1.5857858484658735e-05, "loss": 0.6619, "step": 12680, "task_loss": 0.48109281063079834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3309782147407532, "epoch": 10.72, "learning_rate": 1.585472761427677e-05, "loss": 0.4312, "step": 12681, "task_loss": 0.00879033375531435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5328322649002075, "epoch": 10.72, "learning_rate": 1.5851596743894802e-05, "loss": 0.414, "step": 12682, "task_loss": 1.4887449741363525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24943293631076813, "epoch": 10.72, "learning_rate": 1.5848465873512837e-05, "loss": 0.4599, "step": 12683, "task_loss": 0.12068380415439606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5989769101142883, "epoch": 10.72, "learning_rate": 1.5845335003130872e-05, "loss": 0.5718, "step": 12684, "task_loss": 0.4484593868255615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5704302191734314, "epoch": 10.72, "learning_rate": 1.5842204132748904e-05, "loss": 0.4643, "step": 12685, "task_loss": 1.1043915748596191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5187911987304688, "epoch": 10.72, "learning_rate": 1.583907326236694e-05, "loss": 0.4756, "step": 12686, "task_loss": 0.5813544988632202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5258833765983582, "epoch": 10.72, "learning_rate": 1.5835942391984974e-05, "loss": 0.4438, "step": 12687, "task_loss": 0.9048635363578796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6250196099281311, "epoch": 10.72, "learning_rate": 1.5832811521603006e-05, "loss": 0.5874, "step": 12688, "task_loss": 1.6751710176467896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6167784929275513, "epoch": 10.73, "learning_rate": 1.582968065122104e-05, "loss": 0.7143, "step": 12689, "task_loss": 0.7738767266273499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7178137898445129, "epoch": 10.73, "learning_rate": 1.5826549780839073e-05, "loss": 0.5437, "step": 12690, "task_loss": 0.7693502306938171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4540122151374817, "epoch": 10.73, "learning_rate": 1.5823418910457108e-05, "loss": 0.5313, "step": 12691, "task_loss": 0.2393713742494583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4143012762069702, "epoch": 10.73, "learning_rate": 1.5820288040075143e-05, "loss": 0.5472, "step": 12692, "task_loss": 0.5604233741760254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.52079176902771, "epoch": 10.73, "learning_rate": 1.5817157169693175e-05, "loss": 0.4832, "step": 12693, "task_loss": 0.22334884107112885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5151841640472412, "epoch": 10.73, "learning_rate": 1.581402629931121e-05, "loss": 0.5528, "step": 12694, "task_loss": 1.6425691843032837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36868712306022644, "epoch": 10.73, "learning_rate": 1.5810895428929242e-05, "loss": 0.5395, "step": 12695, "task_loss": 0.3365870714187622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5898135900497437, "epoch": 10.73, "learning_rate": 1.5807764558547277e-05, "loss": 0.4471, "step": 12696, "task_loss": 0.5825769901275635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3836008608341217, "epoch": 10.73, "learning_rate": 1.580463368816531e-05, "loss": 0.4299, "step": 12697, "task_loss": 0.2819066047668457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4682512879371643, "epoch": 10.73, "learning_rate": 1.5801502817783344e-05, "loss": 0.5348, "step": 12698, "task_loss": 0.9686459302902222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7222830057144165, "epoch": 10.73, "learning_rate": 1.5798371947401376e-05, "loss": 0.4844, "step": 12699, "task_loss": 0.8110709190368652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4638531804084778, "epoch": 10.73, "learning_rate": 1.579524107701941e-05, "loss": 0.5464, "step": 12700, "task_loss": 0.40125831961631775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7392451763153076, "epoch": 10.74, "learning_rate": 1.5792110206637446e-05, "loss": 0.7688, "step": 12701, "task_loss": 1.0546914339065552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47463861107826233, "epoch": 10.74, "learning_rate": 1.578897933625548e-05, "loss": 0.4366, "step": 12702, "task_loss": 0.7405520677566528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.630129337310791, "epoch": 10.74, "learning_rate": 1.5785848465873513e-05, "loss": 0.5523, "step": 12703, "task_loss": 0.36984744668006897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6521474719047546, "epoch": 10.74, "learning_rate": 1.578271759549155e-05, "loss": 0.5402, "step": 12704, "task_loss": 0.4454151690006256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4658559560775757, "epoch": 10.74, "learning_rate": 1.5779586725109584e-05, "loss": 0.4632, "step": 12705, "task_loss": 1.0324056148529053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6627521514892578, "epoch": 10.74, "learning_rate": 1.5776455854727615e-05, "loss": 0.5141, "step": 12706, "task_loss": 0.24185839295387268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6582807898521423, "epoch": 10.74, "learning_rate": 1.577332498434565e-05, "loss": 0.7475, "step": 12707, "task_loss": 1.3700097799301147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42133307456970215, "epoch": 10.74, "learning_rate": 1.5770194113963682e-05, "loss": 0.3504, "step": 12708, "task_loss": 1.219004511833191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31237006187438965, "epoch": 10.74, "learning_rate": 1.5767063243581718e-05, "loss": 0.5009, "step": 12709, "task_loss": 0.20326709747314453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25210896134376526, "epoch": 10.74, "learning_rate": 1.576393237319975e-05, "loss": 0.4014, "step": 12710, "task_loss": 0.905098021030426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6249246597290039, "epoch": 10.74, "learning_rate": 1.5760801502817785e-05, "loss": 0.4597, "step": 12711, "task_loss": 0.11727923154830933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7793350219726562, "epoch": 10.75, "learning_rate": 1.5757670632435816e-05, "loss": 0.5708, "step": 12712, "task_loss": 1.0090863704681396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27613744139671326, "epoch": 10.75, "learning_rate": 1.575453976205385e-05, "loss": 0.4683, "step": 12713, "task_loss": 0.1000911220908165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4316602647304535, "epoch": 10.75, "learning_rate": 1.5751408891671883e-05, "loss": 0.4418, "step": 12714, "task_loss": 0.6343231201171875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4163796901702881, "epoch": 10.75, "learning_rate": 1.574827802128992e-05, "loss": 0.5413, "step": 12715, "task_loss": 0.24613262712955475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.358731746673584, "epoch": 10.75, "learning_rate": 1.574514715090795e-05, "loss": 0.4643, "step": 12716, "task_loss": 0.6359772682189941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5646160244941711, "epoch": 10.75, "learning_rate": 1.5742016280525985e-05, "loss": 0.5756, "step": 12717, "task_loss": 0.9304928779602051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4299827516078949, "epoch": 10.75, "learning_rate": 1.573888541014402e-05, "loss": 0.5355, "step": 12718, "task_loss": 0.6904782056808472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4733184576034546, "epoch": 10.75, "learning_rate": 1.5735754539762056e-05, "loss": 0.6345, "step": 12719, "task_loss": 0.5149893164634705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4782940745353699, "epoch": 10.75, "learning_rate": 1.573262366938009e-05, "loss": 0.4603, "step": 12720, "task_loss": 0.3528718948364258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3529214859008789, "epoch": 10.75, "learning_rate": 1.5729492798998123e-05, "loss": 0.4521, "step": 12721, "task_loss": 0.26036709547042847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2780185043811798, "epoch": 10.75, "learning_rate": 1.5726361928616158e-05, "loss": 0.4877, "step": 12722, "task_loss": 0.24430212378501892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5203900933265686, "epoch": 10.75, "learning_rate": 1.572323105823419e-05, "loss": 0.3966, "step": 12723, "task_loss": 0.5843032002449036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6800537109375, "epoch": 10.76, "learning_rate": 1.5720100187852225e-05, "loss": 0.5802, "step": 12724, "task_loss": 0.6362813115119934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5569829940795898, "epoch": 10.76, "learning_rate": 1.5716969317470257e-05, "loss": 0.4995, "step": 12725, "task_loss": 0.5593773722648621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38138726353645325, "epoch": 10.76, "learning_rate": 1.5713838447088292e-05, "loss": 0.6159, "step": 12726, "task_loss": 0.5159056186676025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3828917443752289, "epoch": 10.76, "learning_rate": 1.5710707576706324e-05, "loss": 0.4918, "step": 12727, "task_loss": 0.14301235973834991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43275684118270874, "epoch": 10.76, "learning_rate": 1.570757670632436e-05, "loss": 0.5853, "step": 12728, "task_loss": 1.252012014389038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5913571119308472, "epoch": 10.76, "learning_rate": 1.5704445835942394e-05, "loss": 0.4626, "step": 12729, "task_loss": 0.36528119444847107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45062339305877686, "epoch": 10.76, "learning_rate": 1.5701314965560426e-05, "loss": 0.5788, "step": 12730, "task_loss": 0.3145167827606201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45125052332878113, "epoch": 10.76, "learning_rate": 1.569818409517846e-05, "loss": 0.4545, "step": 12731, "task_loss": 0.7161767482757568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3793390393257141, "epoch": 10.76, "learning_rate": 1.5695053224796493e-05, "loss": 0.5016, "step": 12732, "task_loss": 1.1155364513397217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6610978841781616, "epoch": 10.76, "learning_rate": 1.5691922354414528e-05, "loss": 0.5654, "step": 12733, "task_loss": 1.3168280124664307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5613226294517517, "epoch": 10.76, "learning_rate": 1.568879148403256e-05, "loss": 0.5067, "step": 12734, "task_loss": 0.9575945734977722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8081387281417847, "epoch": 10.76, "learning_rate": 1.5685660613650595e-05, "loss": 0.5238, "step": 12735, "task_loss": 1.3258378505706787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5057939291000366, "epoch": 10.77, "learning_rate": 1.568252974326863e-05, "loss": 0.5578, "step": 12736, "task_loss": 0.5835580229759216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3336237370967865, "epoch": 10.77, "learning_rate": 1.5679398872886665e-05, "loss": 0.6403, "step": 12737, "task_loss": 1.1693696975708008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8029102683067322, "epoch": 10.77, "learning_rate": 1.5676268002504697e-05, "loss": 0.6574, "step": 12738, "task_loss": 0.600391149520874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6123663187026978, "epoch": 10.77, "learning_rate": 1.5673137132122732e-05, "loss": 0.3975, "step": 12739, "task_loss": 0.73946213722229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4133802056312561, "epoch": 10.77, "learning_rate": 1.5670006261740764e-05, "loss": 0.6091, "step": 12740, "task_loss": 0.7366372346878052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3945310413837433, "epoch": 10.77, "learning_rate": 1.56668753913588e-05, "loss": 0.5608, "step": 12741, "task_loss": 0.41016438603401184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8158407807350159, "epoch": 10.77, "learning_rate": 1.5663744520976834e-05, "loss": 0.4646, "step": 12742, "task_loss": 0.5057142376899719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4057363271713257, "epoch": 10.77, "learning_rate": 1.5660613650594866e-05, "loss": 0.5533, "step": 12743, "task_loss": 0.6531804203987122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4104726016521454, "epoch": 10.77, "learning_rate": 1.56574827802129e-05, "loss": 0.4725, "step": 12744, "task_loss": 0.46895191073417664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5845575928688049, "epoch": 10.77, "learning_rate": 1.5654351909830933e-05, "loss": 0.6377, "step": 12745, "task_loss": 0.8614857196807861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2530202269554138, "epoch": 10.77, "learning_rate": 1.5651221039448968e-05, "loss": 0.4961, "step": 12746, "task_loss": 0.49152660369873047 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45092225074768066, "epoch": 10.77, "learning_rate": 1.5648090169067e-05, "loss": 0.5049, "step": 12747, "task_loss": 1.0841782093048096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3878358006477356, "epoch": 10.78, "learning_rate": 1.5644959298685035e-05, "loss": 0.3987, "step": 12748, "task_loss": 0.12942691147327423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32123643159866333, "epoch": 10.78, "learning_rate": 1.5641828428303067e-05, "loss": 0.426, "step": 12749, "task_loss": 1.1930243968963623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38964444398880005, "epoch": 10.78, "learning_rate": 1.5638697557921102e-05, "loss": 0.4889, "step": 12750, "task_loss": 0.5713999271392822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31757333874702454, "epoch": 10.78, "learning_rate": 1.5635566687539137e-05, "loss": 0.498, "step": 12751, "task_loss": 0.22731442749500275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4187157452106476, "epoch": 10.78, "learning_rate": 1.563243581715717e-05, "loss": 0.4448, "step": 12752, "task_loss": 0.8707412481307983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.747075080871582, "epoch": 10.78, "learning_rate": 1.5629304946775204e-05, "loss": 0.6281, "step": 12753, "task_loss": 0.6780080199241638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8715403079986572, "epoch": 10.78, "learning_rate": 1.562617407639324e-05, "loss": 0.6553, "step": 12754, "task_loss": 0.7075722813606262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6105575561523438, "epoch": 10.78, "learning_rate": 1.5623043206011274e-05, "loss": 0.5647, "step": 12755, "task_loss": 0.8980018496513367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4538103938102722, "epoch": 10.78, "learning_rate": 1.5619912335629306e-05, "loss": 0.6633, "step": 12756, "task_loss": 0.38282573223114014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40719425678253174, "epoch": 10.78, "learning_rate": 1.561678146524734e-05, "loss": 0.5659, "step": 12757, "task_loss": 0.3483627438545227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1962966024875641, "epoch": 10.78, "learning_rate": 1.5613650594865373e-05, "loss": 0.454, "step": 12758, "task_loss": 0.5096330642700195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4721667766571045, "epoch": 10.78, "learning_rate": 1.561051972448341e-05, "loss": 0.4971, "step": 12759, "task_loss": 0.9457393884658813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7800027132034302, "epoch": 10.79, "learning_rate": 1.560738885410144e-05, "loss": 0.5613, "step": 12760, "task_loss": 0.6415376663208008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4542236924171448, "epoch": 10.79, "learning_rate": 1.5604257983719475e-05, "loss": 0.4687, "step": 12761, "task_loss": 0.4741625487804413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42666172981262207, "epoch": 10.79, "learning_rate": 1.5601127113337507e-05, "loss": 0.567, "step": 12762, "task_loss": 0.14345037937164307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5495996475219727, "epoch": 10.79, "learning_rate": 1.5597996242955542e-05, "loss": 0.5412, "step": 12763, "task_loss": 0.2736620306968689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3186270594596863, "epoch": 10.79, "learning_rate": 1.5594865372573574e-05, "loss": 0.5121, "step": 12764, "task_loss": 0.33065155148506165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45484718680381775, "epoch": 10.79, "learning_rate": 1.559173450219161e-05, "loss": 0.5596, "step": 12765, "task_loss": 1.037168264389038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44111162424087524, "epoch": 10.79, "learning_rate": 1.5588603631809644e-05, "loss": 0.4789, "step": 12766, "task_loss": 0.5657808184623718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4364088177680969, "epoch": 10.79, "learning_rate": 1.5585472761427676e-05, "loss": 0.4565, "step": 12767, "task_loss": 0.5176511406898499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4882410168647766, "epoch": 10.79, "learning_rate": 1.558234189104571e-05, "loss": 0.5795, "step": 12768, "task_loss": 0.6805793046951294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5414342880249023, "epoch": 10.79, "learning_rate": 1.5579211020663746e-05, "loss": 0.4159, "step": 12769, "task_loss": 0.9137556552886963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31543153524398804, "epoch": 10.79, "learning_rate": 1.5576080150281778e-05, "loss": 0.5238, "step": 12770, "task_loss": 0.2871803343296051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3948426842689514, "epoch": 10.79, "learning_rate": 1.5572949279899813e-05, "loss": 0.4618, "step": 12771, "task_loss": 0.2529677152633667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3082101345062256, "epoch": 10.8, "learning_rate": 1.556981840951785e-05, "loss": 0.3085, "step": 12772, "task_loss": 0.2720949649810791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8863597512245178, "epoch": 10.8, "learning_rate": 1.556668753913588e-05, "loss": 0.5597, "step": 12773, "task_loss": 0.5564403533935547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4445578455924988, "epoch": 10.8, "learning_rate": 1.5563556668753916e-05, "loss": 0.3594, "step": 12774, "task_loss": 0.3626253604888916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3852667808532715, "epoch": 10.8, "learning_rate": 1.5560425798371947e-05, "loss": 0.5791, "step": 12775, "task_loss": 0.33580565452575684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7871660590171814, "epoch": 10.8, "learning_rate": 1.5557294927989982e-05, "loss": 0.5918, "step": 12776, "task_loss": 1.1155492067337036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41727420687675476, "epoch": 10.8, "learning_rate": 1.5554164057608014e-05, "loss": 0.3614, "step": 12777, "task_loss": 0.6997470259666443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5463415384292603, "epoch": 10.8, "learning_rate": 1.555103318722605e-05, "loss": 0.4608, "step": 12778, "task_loss": 0.732166051864624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41985005140304565, "epoch": 10.8, "learning_rate": 1.5547902316844085e-05, "loss": 0.5213, "step": 12779, "task_loss": 0.45321178436279297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5995181798934937, "epoch": 10.8, "learning_rate": 1.5544771446462116e-05, "loss": 0.4945, "step": 12780, "task_loss": 0.7768793702125549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6910864114761353, "epoch": 10.8, "learning_rate": 1.554164057608015e-05, "loss": 0.5254, "step": 12781, "task_loss": 1.221524715423584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37395018339157104, "epoch": 10.8, "learning_rate": 1.5538509705698183e-05, "loss": 0.5563, "step": 12782, "task_loss": 0.9881718158721924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3316475749015808, "epoch": 10.81, "learning_rate": 1.553537883531622e-05, "loss": 0.4544, "step": 12783, "task_loss": 0.8298320174217224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7883398532867432, "epoch": 10.81, "learning_rate": 1.553224796493425e-05, "loss": 0.5517, "step": 12784, "task_loss": 0.5512227416038513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6283999085426331, "epoch": 10.81, "learning_rate": 1.5529117094552285e-05, "loss": 0.6955, "step": 12785, "task_loss": 0.6406662464141846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3931836187839508, "epoch": 10.81, "learning_rate": 1.552598622417032e-05, "loss": 0.54, "step": 12786, "task_loss": 0.37237393856048584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42268893122673035, "epoch": 10.81, "learning_rate": 1.5522855353788356e-05, "loss": 0.4657, "step": 12787, "task_loss": 0.5501884818077087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5650085210800171, "epoch": 10.81, "learning_rate": 1.5519724483406388e-05, "loss": 0.5012, "step": 12788, "task_loss": 1.3424909114837646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7824451327323914, "epoch": 10.81, "learning_rate": 1.5516593613024423e-05, "loss": 0.6154, "step": 12789, "task_loss": 0.4506768584251404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34361129999160767, "epoch": 10.81, "learning_rate": 1.5513462742642458e-05, "loss": 0.4703, "step": 12790, "task_loss": 0.4675077497959137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8395873308181763, "epoch": 10.81, "learning_rate": 1.551033187226049e-05, "loss": 0.5433, "step": 12791, "task_loss": 1.440772294998169 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3066089451313019, "epoch": 10.81, "learning_rate": 1.5507201001878525e-05, "loss": 0.481, "step": 12792, "task_loss": 0.7763055562973022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35358119010925293, "epoch": 10.81, "learning_rate": 1.5504070131496557e-05, "loss": 0.4901, "step": 12793, "task_loss": 0.42739397287368774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4348416328430176, "epoch": 10.81, "learning_rate": 1.5500939261114592e-05, "loss": 0.5243, "step": 12794, "task_loss": 0.6240849494934082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9420523047447205, "epoch": 10.82, "learning_rate": 1.5497808390732624e-05, "loss": 0.5164, "step": 12795, "task_loss": 0.35489827394485474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5926236510276794, "epoch": 10.82, "learning_rate": 1.549467752035066e-05, "loss": 0.437, "step": 12796, "task_loss": 0.6839828491210938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48098886013031006, "epoch": 10.82, "learning_rate": 1.549154664996869e-05, "loss": 0.473, "step": 12797, "task_loss": 1.62057363986969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6535744667053223, "epoch": 10.82, "learning_rate": 1.5488415779586726e-05, "loss": 0.5993, "step": 12798, "task_loss": 1.1384533643722534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5739275217056274, "epoch": 10.82, "learning_rate": 1.5485284909204758e-05, "loss": 0.5845, "step": 12799, "task_loss": 0.7802078723907471 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2681189179420471, "epoch": 10.82, "learning_rate": 1.5482154038822793e-05, "loss": 0.3516, "step": 12800, "task_loss": 0.022894661873579025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4181182384490967, "epoch": 10.82, "learning_rate": 1.5479023168440824e-05, "loss": 0.5274, "step": 12801, "task_loss": 0.3369516134262085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.970653235912323, "epoch": 10.82, "learning_rate": 1.547589229805886e-05, "loss": 0.6098, "step": 12802, "task_loss": 0.8197371363639832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6503778100013733, "epoch": 10.82, "learning_rate": 1.5472761427676895e-05, "loss": 0.4484, "step": 12803, "task_loss": 0.8802557587623596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.708260178565979, "epoch": 10.82, "learning_rate": 1.546963055729493e-05, "loss": 0.5817, "step": 12804, "task_loss": 0.2719804346561432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4247015714645386, "epoch": 10.82, "learning_rate": 1.5466499686912965e-05, "loss": 0.5471, "step": 12805, "task_loss": 0.9752165675163269 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38041216135025024, "epoch": 10.82, "learning_rate": 1.5463368816530997e-05, "loss": 0.4401, "step": 12806, "task_loss": 0.9101687669754028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42993542551994324, "epoch": 10.83, "learning_rate": 1.5460237946149032e-05, "loss": 0.3529, "step": 12807, "task_loss": 0.34772366285324097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40223246812820435, "epoch": 10.83, "learning_rate": 1.5457107075767064e-05, "loss": 0.5413, "step": 12808, "task_loss": 0.606217086315155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4346325695514679, "epoch": 10.83, "learning_rate": 1.54539762053851e-05, "loss": 0.3959, "step": 12809, "task_loss": 1.3492298126220703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.702835738658905, "epoch": 10.83, "learning_rate": 1.545084533500313e-05, "loss": 0.5568, "step": 12810, "task_loss": 0.7425141930580139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5245512127876282, "epoch": 10.83, "learning_rate": 1.5447714464621166e-05, "loss": 0.4911, "step": 12811, "task_loss": 1.5057623386383057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43983620405197144, "epoch": 10.83, "learning_rate": 1.5444583594239198e-05, "loss": 0.3529, "step": 12812, "task_loss": 0.3977017104625702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32536551356315613, "epoch": 10.83, "learning_rate": 1.5441452723857233e-05, "loss": 0.551, "step": 12813, "task_loss": 0.3238297402858734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7930103540420532, "epoch": 10.83, "learning_rate": 1.5438321853475265e-05, "loss": 0.6259, "step": 12814, "task_loss": 2.6887083053588867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35639041662216187, "epoch": 10.83, "learning_rate": 1.54351909830933e-05, "loss": 0.4225, "step": 12815, "task_loss": 0.5158243179321289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5048477649688721, "epoch": 10.83, "learning_rate": 1.5432060112711335e-05, "loss": 0.4078, "step": 12816, "task_loss": 0.2187536656856537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5767139196395874, "epoch": 10.83, "learning_rate": 1.5428929242329367e-05, "loss": 0.5896, "step": 12817, "task_loss": 0.8545069694519043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38463446497917175, "epoch": 10.83, "learning_rate": 1.5425798371947402e-05, "loss": 0.5545, "step": 12818, "task_loss": 0.9728146195411682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6697628498077393, "epoch": 10.84, "learning_rate": 1.5422667501565434e-05, "loss": 0.5451, "step": 12819, "task_loss": 0.7893702387809753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6149227023124695, "epoch": 10.84, "learning_rate": 1.541953663118347e-05, "loss": 0.5135, "step": 12820, "task_loss": 1.1641143560409546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5076012015342712, "epoch": 10.84, "learning_rate": 1.5416405760801504e-05, "loss": 0.5236, "step": 12821, "task_loss": 0.49111270904541016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5371319055557251, "epoch": 10.84, "learning_rate": 1.541327489041954e-05, "loss": 0.3724, "step": 12822, "task_loss": 0.4683212637901306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5334243774414062, "epoch": 10.84, "learning_rate": 1.541014402003757e-05, "loss": 0.5037, "step": 12823, "task_loss": 0.6849851012229919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27763134241104126, "epoch": 10.84, "learning_rate": 1.5407013149655606e-05, "loss": 0.5478, "step": 12824, "task_loss": 0.7202421426773071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5218371152877808, "epoch": 10.84, "learning_rate": 1.5403882279273638e-05, "loss": 0.4347, "step": 12825, "task_loss": 0.43332523107528687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5562266707420349, "epoch": 10.84, "learning_rate": 1.5400751408891673e-05, "loss": 0.5176, "step": 12826, "task_loss": 0.5799397230148315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.598837673664093, "epoch": 10.84, "learning_rate": 1.539762053850971e-05, "loss": 0.4151, "step": 12827, "task_loss": 1.0753505229949951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6763463616371155, "epoch": 10.84, "learning_rate": 1.539448966812774e-05, "loss": 0.541, "step": 12828, "task_loss": 0.7038050889968872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4453711211681366, "epoch": 10.84, "learning_rate": 1.5391358797745775e-05, "loss": 0.5049, "step": 12829, "task_loss": 0.4902510941028595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4780005216598511, "epoch": 10.84, "learning_rate": 1.5388227927363807e-05, "loss": 0.5381, "step": 12830, "task_loss": 0.20514324307441711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45659899711608887, "epoch": 10.85, "learning_rate": 1.5385097056981842e-05, "loss": 0.5349, "step": 12831, "task_loss": 0.7564250826835632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3050745725631714, "epoch": 10.85, "learning_rate": 1.5381966186599874e-05, "loss": 0.373, "step": 12832, "task_loss": 0.3445606529712677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43756359815597534, "epoch": 10.85, "learning_rate": 1.537883531621791e-05, "loss": 0.4138, "step": 12833, "task_loss": 0.016523420810699463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1843986064195633, "epoch": 10.85, "learning_rate": 1.537570444583594e-05, "loss": 0.4216, "step": 12834, "task_loss": 0.30423253774642944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32547152042388916, "epoch": 10.85, "learning_rate": 1.5372573575453976e-05, "loss": 0.4053, "step": 12835, "task_loss": 0.2504901885986328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48398250341415405, "epoch": 10.85, "learning_rate": 1.536944270507201e-05, "loss": 0.4528, "step": 12836, "task_loss": 1.6998077630996704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4274176061153412, "epoch": 10.85, "learning_rate": 1.5366311834690043e-05, "loss": 0.5799, "step": 12837, "task_loss": 0.39763444662094116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7369102239608765, "epoch": 10.85, "learning_rate": 1.5363180964308078e-05, "loss": 0.5619, "step": 12838, "task_loss": 1.290900468826294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5954046249389648, "epoch": 10.85, "learning_rate": 1.5360050093926113e-05, "loss": 0.6322, "step": 12839, "task_loss": 1.0652023553848267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3958621621131897, "epoch": 10.85, "learning_rate": 1.535691922354415e-05, "loss": 0.4309, "step": 12840, "task_loss": 0.18614304065704346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4223995804786682, "epoch": 10.85, "learning_rate": 1.535378835316218e-05, "loss": 0.4356, "step": 12841, "task_loss": 0.28902295231819153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3339812755584717, "epoch": 10.85, "learning_rate": 1.5350657482780216e-05, "loss": 0.6127, "step": 12842, "task_loss": 0.13121390342712402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30812087655067444, "epoch": 10.86, "learning_rate": 1.5347526612398247e-05, "loss": 0.3969, "step": 12843, "task_loss": 1.0218031406402588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7180724143981934, "epoch": 10.86, "learning_rate": 1.5344395742016283e-05, "loss": 0.6542, "step": 12844, "task_loss": 0.716809868812561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4471002519130707, "epoch": 10.86, "learning_rate": 1.5341264871634314e-05, "loss": 0.4314, "step": 12845, "task_loss": 0.3396604061126709 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41742151975631714, "epoch": 10.86, "learning_rate": 1.533813400125235e-05, "loss": 0.3819, "step": 12846, "task_loss": 0.5508880615234375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32282787561416626, "epoch": 10.86, "learning_rate": 1.533500313087038e-05, "loss": 0.5359, "step": 12847, "task_loss": 0.26426801085472107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7903274297714233, "epoch": 10.86, "learning_rate": 1.5331872260488416e-05, "loss": 0.6819, "step": 12848, "task_loss": 0.34806784987449646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3534128963947296, "epoch": 10.86, "learning_rate": 1.5328741390106448e-05, "loss": 0.592, "step": 12849, "task_loss": 0.5711828470230103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38979971408843994, "epoch": 10.86, "learning_rate": 1.5325610519724483e-05, "loss": 0.4561, "step": 12850, "task_loss": 1.3977922201156616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5685175657272339, "epoch": 10.86, "learning_rate": 1.532247964934252e-05, "loss": 0.4819, "step": 12851, "task_loss": 0.5231055617332458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5189343094825745, "epoch": 10.86, "learning_rate": 1.531934877896055e-05, "loss": 0.5743, "step": 12852, "task_loss": 2.011725425720215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5378477573394775, "epoch": 10.86, "learning_rate": 1.5316217908578585e-05, "loss": 0.4875, "step": 12853, "task_loss": 0.48432478308677673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6526554822921753, "epoch": 10.87, "learning_rate": 1.531308703819662e-05, "loss": 0.5987, "step": 12854, "task_loss": 1.3323031663894653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0548696517944336, "epoch": 10.87, "learning_rate": 1.5309956167814652e-05, "loss": 0.7259, "step": 12855, "task_loss": 1.0238010883331299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5036675930023193, "epoch": 10.87, "learning_rate": 1.5306825297432688e-05, "loss": 0.4171, "step": 12856, "task_loss": 0.42827117443084717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6843302249908447, "epoch": 10.87, "learning_rate": 1.5303694427050723e-05, "loss": 0.4811, "step": 12857, "task_loss": 0.33324164152145386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.186903715133667, "epoch": 10.87, "learning_rate": 1.5300563556668755e-05, "loss": 0.6678, "step": 12858, "task_loss": 1.3386549949645996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8639566898345947, "epoch": 10.87, "learning_rate": 1.529743268628679e-05, "loss": 0.5299, "step": 12859, "task_loss": 1.3680843114852905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28941792249679565, "epoch": 10.87, "learning_rate": 1.529430181590482e-05, "loss": 0.4311, "step": 12860, "task_loss": 0.25852569937705994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4175015985965729, "epoch": 10.87, "learning_rate": 1.5291170945522857e-05, "loss": 0.4725, "step": 12861, "task_loss": 0.48214325308799744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49453264474868774, "epoch": 10.87, "learning_rate": 1.528804007514089e-05, "loss": 0.4579, "step": 12862, "task_loss": 0.37277644872665405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.534419059753418, "epoch": 10.87, "learning_rate": 1.5284909204758924e-05, "loss": 0.5548, "step": 12863, "task_loss": 0.19659553468227386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4552343487739563, "epoch": 10.87, "learning_rate": 1.528177833437696e-05, "loss": 0.4451, "step": 12864, "task_loss": 0.7157099843025208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25352349877357483, "epoch": 10.87, "learning_rate": 1.527864746399499e-05, "loss": 0.4413, "step": 12865, "task_loss": 0.11901690065860748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34985291957855225, "epoch": 10.88, "learning_rate": 1.5275516593613026e-05, "loss": 0.503, "step": 12866, "task_loss": 0.7866591811180115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30115342140197754, "epoch": 10.88, "learning_rate": 1.5272385723231058e-05, "loss": 0.45, "step": 12867, "task_loss": 0.46104201674461365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7551948428153992, "epoch": 10.88, "learning_rate": 1.5269254852849093e-05, "loss": 0.6107, "step": 12868, "task_loss": 1.5394608974456787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.679541289806366, "epoch": 10.88, "learning_rate": 1.5266123982467124e-05, "loss": 0.425, "step": 12869, "task_loss": 0.495027095079422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7825802564620972, "epoch": 10.88, "learning_rate": 1.526299311208516e-05, "loss": 0.5735, "step": 12870, "task_loss": 0.860990047454834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7883006930351257, "epoch": 10.88, "learning_rate": 1.5259862241703195e-05, "loss": 0.4737, "step": 12871, "task_loss": 0.712565541267395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4142652750015259, "epoch": 10.88, "learning_rate": 1.5256731371321228e-05, "loss": 0.4959, "step": 12872, "task_loss": 0.0909951701760292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9128175973892212, "epoch": 10.88, "learning_rate": 1.525360050093926e-05, "loss": 0.5694, "step": 12873, "task_loss": 1.0372146368026733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2931479215621948, "epoch": 10.88, "learning_rate": 1.5250469630557295e-05, "loss": 0.4766, "step": 12874, "task_loss": 0.4450807273387909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37314385175704956, "epoch": 10.88, "learning_rate": 1.5247338760175329e-05, "loss": 0.387, "step": 12875, "task_loss": 0.22064079344272614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4085252285003662, "epoch": 10.88, "learning_rate": 1.5244207889793364e-05, "loss": 0.4467, "step": 12876, "task_loss": 1.4386588335037231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37106966972351074, "epoch": 10.88, "learning_rate": 1.5241077019411399e-05, "loss": 0.5642, "step": 12877, "task_loss": 0.30367597937583923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9864603281021118, "epoch": 10.89, "learning_rate": 1.5237946149029431e-05, "loss": 0.781, "step": 12878, "task_loss": 0.535528838634491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42650967836380005, "epoch": 10.89, "learning_rate": 1.5234815278647466e-05, "loss": 0.3789, "step": 12879, "task_loss": 0.32472240924835205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21971622109413147, "epoch": 10.89, "learning_rate": 1.5231684408265498e-05, "loss": 0.4206, "step": 12880, "task_loss": 0.5167831778526306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6583681106567383, "epoch": 10.89, "learning_rate": 1.5228553537883533e-05, "loss": 0.5391, "step": 12881, "task_loss": 0.5404089689254761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.074265956878662, "epoch": 10.89, "learning_rate": 1.5225422667501565e-05, "loss": 0.7077, "step": 12882, "task_loss": 1.3475638628005981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7244732975959778, "epoch": 10.89, "learning_rate": 1.52222917971196e-05, "loss": 0.5444, "step": 12883, "task_loss": 0.9020153284072876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4854426980018616, "epoch": 10.89, "learning_rate": 1.5219160926737633e-05, "loss": 0.5083, "step": 12884, "task_loss": 0.7650241851806641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5870682001113892, "epoch": 10.89, "learning_rate": 1.5216030056355669e-05, "loss": 0.5414, "step": 12885, "task_loss": 0.8088772296905518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4771483838558197, "epoch": 10.89, "learning_rate": 1.52128991859737e-05, "loss": 0.4688, "step": 12886, "task_loss": 0.285031259059906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42295581102371216, "epoch": 10.89, "learning_rate": 1.5209768315591736e-05, "loss": 0.6241, "step": 12887, "task_loss": 0.5746353268623352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3870317339897156, "epoch": 10.89, "learning_rate": 1.520663744520977e-05, "loss": 0.466, "step": 12888, "task_loss": 0.895117461681366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3416445851325989, "epoch": 10.89, "learning_rate": 1.5203506574827802e-05, "loss": 0.4454, "step": 12889, "task_loss": 0.3970315158367157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6371868848800659, "epoch": 10.9, "learning_rate": 1.5200375704445838e-05, "loss": 0.5367, "step": 12890, "task_loss": 1.3123793601989746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5199769735336304, "epoch": 10.9, "learning_rate": 1.519724483406387e-05, "loss": 0.5248, "step": 12891, "task_loss": 0.6719616651535034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2918190062046051, "epoch": 10.9, "learning_rate": 1.5194113963681905e-05, "loss": 0.4181, "step": 12892, "task_loss": 0.2829280197620392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46294957399368286, "epoch": 10.9, "learning_rate": 1.5190983093299938e-05, "loss": 0.4763, "step": 12893, "task_loss": 0.27679872512817383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3732497990131378, "epoch": 10.9, "learning_rate": 1.5187852222917973e-05, "loss": 0.5636, "step": 12894, "task_loss": 1.1364901065826416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27819621562957764, "epoch": 10.9, "learning_rate": 1.5184721352536005e-05, "loss": 0.3775, "step": 12895, "task_loss": 0.22748099267482758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33807435631752014, "epoch": 10.9, "learning_rate": 1.518159048215404e-05, "loss": 0.4776, "step": 12896, "task_loss": 0.9609302878379822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3004332184791565, "epoch": 10.9, "learning_rate": 1.5178459611772072e-05, "loss": 0.4593, "step": 12897, "task_loss": 0.1714378148317337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4040224552154541, "epoch": 10.9, "learning_rate": 1.5175328741390107e-05, "loss": 0.5468, "step": 12898, "task_loss": 0.9719327688217163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40206384658813477, "epoch": 10.9, "learning_rate": 1.5172197871008139e-05, "loss": 0.4721, "step": 12899, "task_loss": 0.1340421885251999 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8759273886680603, "epoch": 10.9, "learning_rate": 1.5169067000626174e-05, "loss": 0.6209, "step": 12900, "task_loss": 0.2504759430885315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5858975648880005, "epoch": 10.9, "learning_rate": 1.516593613024421e-05, "loss": 0.5697, "step": 12901, "task_loss": 0.9438489675521851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3881514370441437, "epoch": 10.91, "learning_rate": 1.5162805259862243e-05, "loss": 0.4293, "step": 12902, "task_loss": 0.8958790898323059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6489452123641968, "epoch": 10.91, "learning_rate": 1.5159674389480278e-05, "loss": 0.4016, "step": 12903, "task_loss": 0.7063891291618347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30103984475135803, "epoch": 10.91, "learning_rate": 1.515654351909831e-05, "loss": 0.4226, "step": 12904, "task_loss": 0.569516658782959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34334826469421387, "epoch": 10.91, "learning_rate": 1.5153412648716345e-05, "loss": 0.4917, "step": 12905, "task_loss": 0.32131877541542053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7050896883010864, "epoch": 10.91, "learning_rate": 1.5150281778334377e-05, "loss": 0.5697, "step": 12906, "task_loss": 0.1115395650267601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7843804359436035, "epoch": 10.91, "learning_rate": 1.5147150907952412e-05, "loss": 0.5766, "step": 12907, "task_loss": 1.250160813331604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6754249930381775, "epoch": 10.91, "learning_rate": 1.5144020037570444e-05, "loss": 0.4307, "step": 12908, "task_loss": 0.5749412178993225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3517855107784271, "epoch": 10.91, "learning_rate": 1.5140889167188479e-05, "loss": 0.4176, "step": 12909, "task_loss": 0.5149737596511841 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48996782302856445, "epoch": 10.91, "learning_rate": 1.5137758296806512e-05, "loss": 0.4115, "step": 12910, "task_loss": 0.5563344359397888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.754405677318573, "epoch": 10.91, "learning_rate": 1.5134627426424547e-05, "loss": 0.5074, "step": 12911, "task_loss": 0.7390317916870117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5851496458053589, "epoch": 10.91, "learning_rate": 1.513149655604258e-05, "loss": 0.4674, "step": 12912, "task_loss": 0.19400453567504883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22689059376716614, "epoch": 10.91, "learning_rate": 1.5128365685660614e-05, "loss": 0.4271, "step": 12913, "task_loss": 0.3806815445423126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030706524848938, "epoch": 10.92, "learning_rate": 1.512523481527865e-05, "loss": 0.4713, "step": 12914, "task_loss": 0.4394088089466095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4294728636741638, "epoch": 10.92, "learning_rate": 1.5122103944896681e-05, "loss": 0.4378, "step": 12915, "task_loss": 0.5344105362892151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47755247354507446, "epoch": 10.92, "learning_rate": 1.5118973074514716e-05, "loss": 0.6769, "step": 12916, "task_loss": 0.6348171234130859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45726001262664795, "epoch": 10.92, "learning_rate": 1.511584220413275e-05, "loss": 0.7682, "step": 12917, "task_loss": 0.5556889176368713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6019982099533081, "epoch": 10.92, "learning_rate": 1.5112711333750783e-05, "loss": 0.6682, "step": 12918, "task_loss": 1.7862690687179565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4576857388019562, "epoch": 10.92, "learning_rate": 1.5109580463368817e-05, "loss": 0.403, "step": 12919, "task_loss": 0.6708860993385315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6598764657974243, "epoch": 10.92, "learning_rate": 1.5106449592986852e-05, "loss": 0.5615, "step": 12920, "task_loss": 0.7253443002700806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5390857458114624, "epoch": 10.92, "learning_rate": 1.5103318722604884e-05, "loss": 0.4703, "step": 12921, "task_loss": 0.9200561046600342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8681755065917969, "epoch": 10.92, "learning_rate": 1.5100187852222919e-05, "loss": 0.4993, "step": 12922, "task_loss": 0.9365078806877136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47182267904281616, "epoch": 10.92, "learning_rate": 1.509705698184095e-05, "loss": 0.3869, "step": 12923, "task_loss": 0.4421626925468445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5713654160499573, "epoch": 10.92, "learning_rate": 1.5093926111458986e-05, "loss": 0.5862, "step": 12924, "task_loss": 0.514303982257843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30246150493621826, "epoch": 10.93, "learning_rate": 1.5090795241077021e-05, "loss": 0.3685, "step": 12925, "task_loss": 0.3434082567691803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21043947339057922, "epoch": 10.93, "learning_rate": 1.5087664370695055e-05, "loss": 0.4187, "step": 12926, "task_loss": 0.19982104003429413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3212679922580719, "epoch": 10.93, "learning_rate": 1.5084533500313088e-05, "loss": 0.5154, "step": 12927, "task_loss": 0.532741904258728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4298354387283325, "epoch": 10.93, "learning_rate": 1.5081402629931122e-05, "loss": 0.4762, "step": 12928, "task_loss": 0.6758939027786255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44252005219459534, "epoch": 10.93, "learning_rate": 1.5078271759549157e-05, "loss": 0.4517, "step": 12929, "task_loss": 0.7170485258102417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22628261148929596, "epoch": 10.93, "learning_rate": 1.5075140889167189e-05, "loss": 0.514, "step": 12930, "task_loss": 0.1961561143398285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47747528553009033, "epoch": 10.93, "learning_rate": 1.5072010018785224e-05, "loss": 0.4759, "step": 12931, "task_loss": 0.9369490146636963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35609957575798035, "epoch": 10.93, "learning_rate": 1.5068879148403255e-05, "loss": 0.4508, "step": 12932, "task_loss": 0.9029961228370667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5187788009643555, "epoch": 10.93, "learning_rate": 1.506574827802129e-05, "loss": 0.4138, "step": 12933, "task_loss": 0.3654021620750427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22957441210746765, "epoch": 10.93, "learning_rate": 1.5062617407639324e-05, "loss": 0.4985, "step": 12934, "task_loss": 0.3925546407699585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3773728609085083, "epoch": 10.93, "learning_rate": 1.505948653725736e-05, "loss": 0.4279, "step": 12935, "task_loss": 0.39819833636283875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2864890694618225, "epoch": 10.93, "learning_rate": 1.5056355666875391e-05, "loss": 0.3804, "step": 12936, "task_loss": 0.34117165207862854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.603461503982544, "epoch": 10.94, "learning_rate": 1.5053224796493426e-05, "loss": 0.5472, "step": 12937, "task_loss": 1.3411498069763184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40643274784088135, "epoch": 10.94, "learning_rate": 1.5050093926111461e-05, "loss": 0.5316, "step": 12938, "task_loss": 0.5569625496864319 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2630404829978943, "epoch": 10.94, "learning_rate": 1.5046963055729493e-05, "loss": 0.5795, "step": 12939, "task_loss": 0.38377511501312256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.504969596862793, "epoch": 10.94, "learning_rate": 1.5043832185347528e-05, "loss": 0.4456, "step": 12940, "task_loss": 0.6330620646476746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2759245038032532, "epoch": 10.94, "learning_rate": 1.504070131496556e-05, "loss": 0.4073, "step": 12941, "task_loss": 0.3068579435348511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26225167512893677, "epoch": 10.94, "learning_rate": 1.5037570444583595e-05, "loss": 0.4219, "step": 12942, "task_loss": 0.6231129169464111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6140240430831909, "epoch": 10.94, "learning_rate": 1.5034439574201629e-05, "loss": 0.5365, "step": 12943, "task_loss": 0.6517871618270874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5275040864944458, "epoch": 10.94, "learning_rate": 1.5031308703819664e-05, "loss": 0.594, "step": 12944, "task_loss": 0.7108621597290039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34361445903778076, "epoch": 10.94, "learning_rate": 1.5028177833437696e-05, "loss": 0.3875, "step": 12945, "task_loss": 0.4189951419830322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5503503084182739, "epoch": 10.94, "learning_rate": 1.5025046963055731e-05, "loss": 0.4993, "step": 12946, "task_loss": 0.624238908290863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49070343375205994, "epoch": 10.94, "learning_rate": 1.5021916092673763e-05, "loss": 0.5215, "step": 12947, "task_loss": 0.18108785152435303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43333983421325684, "epoch": 10.94, "learning_rate": 1.5018785222291798e-05, "loss": 0.5688, "step": 12948, "task_loss": 0.565432071685791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47431913018226624, "epoch": 10.95, "learning_rate": 1.501565435190983e-05, "loss": 0.5121, "step": 12949, "task_loss": 1.1016517877578735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35432493686676025, "epoch": 10.95, "learning_rate": 1.5012523481527865e-05, "loss": 0.469, "step": 12950, "task_loss": 0.6447552442550659 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6870264410972595, "epoch": 10.95, "learning_rate": 1.50093926111459e-05, "loss": 0.575, "step": 12951, "task_loss": 0.5808908939361572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5725613832473755, "epoch": 10.95, "learning_rate": 1.5006261740763933e-05, "loss": 0.5578, "step": 12952, "task_loss": 0.054525747895240784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5877391695976257, "epoch": 10.95, "learning_rate": 1.5003130870381969e-05, "loss": 0.5259, "step": 12953, "task_loss": 0.6204513907432556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37136924266815186, "epoch": 10.95, "learning_rate": 1.5e-05, "loss": 0.4641, "step": 12954, "task_loss": 0.6168534159660339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4418243169784546, "epoch": 10.95, "learning_rate": 1.4996869129618036e-05, "loss": 0.3966, "step": 12955, "task_loss": 0.39550620317459106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6478408575057983, "epoch": 10.95, "learning_rate": 1.4993738259236067e-05, "loss": 0.4594, "step": 12956, "task_loss": 0.5114001035690308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4959828853607178, "epoch": 10.95, "learning_rate": 1.4990607388854102e-05, "loss": 0.4684, "step": 12957, "task_loss": 0.22148513793945312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31866520643234253, "epoch": 10.95, "learning_rate": 1.4987476518472134e-05, "loss": 0.3934, "step": 12958, "task_loss": 0.5677955150604248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2551923990249634, "epoch": 10.95, "learning_rate": 1.498434564809017e-05, "loss": 0.3883, "step": 12959, "task_loss": 0.4117354452610016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4295051395893097, "epoch": 10.95, "learning_rate": 1.4981214777708203e-05, "loss": 0.5749, "step": 12960, "task_loss": 0.32496798038482666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5458429455757141, "epoch": 10.96, "learning_rate": 1.4978083907326238e-05, "loss": 0.5942, "step": 12961, "task_loss": 0.3899574279785156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6129829287528992, "epoch": 10.96, "learning_rate": 1.4974953036944273e-05, "loss": 0.4056, "step": 12962, "task_loss": 0.331462562084198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24586474895477295, "epoch": 10.96, "learning_rate": 1.4971822166562305e-05, "loss": 0.4831, "step": 12963, "task_loss": 0.2956916391849518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5967643857002258, "epoch": 10.96, "learning_rate": 1.496869129618034e-05, "loss": 0.6885, "step": 12964, "task_loss": 2.1182498931884766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4915872812271118, "epoch": 10.96, "learning_rate": 1.4965560425798372e-05, "loss": 0.6213, "step": 12965, "task_loss": 1.2763715982437134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7727789282798767, "epoch": 10.96, "learning_rate": 1.4962429555416407e-05, "loss": 0.6332, "step": 12966, "task_loss": 0.998192310333252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4409647583961487, "epoch": 10.96, "learning_rate": 1.4959298685034439e-05, "loss": 0.5704, "step": 12967, "task_loss": 1.3255927562713623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4961637854576111, "epoch": 10.96, "learning_rate": 1.4956167814652474e-05, "loss": 0.55, "step": 12968, "task_loss": 0.35189545154571533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4977482557296753, "epoch": 10.96, "learning_rate": 1.4953036944270508e-05, "loss": 0.3398, "step": 12969, "task_loss": 0.1881825476884842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4209232032299042, "epoch": 10.96, "learning_rate": 1.4949906073888543e-05, "loss": 0.4865, "step": 12970, "task_loss": 0.3455365002155304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6962898969650269, "epoch": 10.96, "learning_rate": 1.4946775203506575e-05, "loss": 0.6235, "step": 12971, "task_loss": 0.4658393859863281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3236977458000183, "epoch": 10.96, "learning_rate": 1.494364433312461e-05, "loss": 0.3508, "step": 12972, "task_loss": 0.8187462687492371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4864118695259094, "epoch": 10.97, "learning_rate": 1.4940513462742641e-05, "loss": 0.4204, "step": 12973, "task_loss": 0.7776699066162109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25884509086608887, "epoch": 10.97, "learning_rate": 1.4937382592360677e-05, "loss": 0.4884, "step": 12974, "task_loss": 0.5779843330383301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39435258507728577, "epoch": 10.97, "learning_rate": 1.4934251721978712e-05, "loss": 0.4251, "step": 12975, "task_loss": 0.13273724913597107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4285828471183777, "epoch": 10.97, "learning_rate": 1.4931120851596744e-05, "loss": 0.4938, "step": 12976, "task_loss": 0.17855019867420197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3454011082649231, "epoch": 10.97, "learning_rate": 1.4927989981214779e-05, "loss": 0.4926, "step": 12977, "task_loss": 0.8749127388000488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4585868716239929, "epoch": 10.97, "learning_rate": 1.4924859110832812e-05, "loss": 0.5856, "step": 12978, "task_loss": 0.31292784214019775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33053457736968994, "epoch": 10.97, "learning_rate": 1.4921728240450847e-05, "loss": 0.4374, "step": 12979, "task_loss": 0.4388846158981323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4259783923625946, "epoch": 10.97, "learning_rate": 1.491859737006888e-05, "loss": 0.4307, "step": 12980, "task_loss": 0.22678515315055847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43607020378112793, "epoch": 10.97, "learning_rate": 1.4915466499686914e-05, "loss": 0.4116, "step": 12981, "task_loss": 0.17978264391422272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44224196672439575, "epoch": 10.97, "learning_rate": 1.4912335629304946e-05, "loss": 0.6471, "step": 12982, "task_loss": 1.2296595573425293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4007852077484131, "epoch": 10.97, "learning_rate": 1.4909204758922981e-05, "loss": 0.4601, "step": 12983, "task_loss": 0.38692113757133484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47526317834854126, "epoch": 10.97, "learning_rate": 1.4906073888541013e-05, "loss": 0.4271, "step": 12984, "task_loss": 0.4098243713378906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15860290825366974, "epoch": 10.98, "learning_rate": 1.4902943018159048e-05, "loss": 0.5196, "step": 12985, "task_loss": 0.37783220410346985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3550984859466553, "epoch": 10.98, "learning_rate": 1.4899812147777083e-05, "loss": 0.5534, "step": 12986, "task_loss": 0.28729403018951416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30894744396209717, "epoch": 10.98, "learning_rate": 1.4896681277395117e-05, "loss": 0.6278, "step": 12987, "task_loss": 0.6254445910453796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5039688348770142, "epoch": 10.98, "learning_rate": 1.4893550407013152e-05, "loss": 0.4678, "step": 12988, "task_loss": 0.24265983700752258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44661813974380493, "epoch": 10.98, "learning_rate": 1.4890419536631184e-05, "loss": 0.4424, "step": 12989, "task_loss": 0.5152918100357056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4599907398223877, "epoch": 10.98, "learning_rate": 1.4887288666249219e-05, "loss": 0.4489, "step": 12990, "task_loss": 1.3579392433166504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5717059373855591, "epoch": 10.98, "learning_rate": 1.488415779586725e-05, "loss": 0.6188, "step": 12991, "task_loss": 0.39119723439216614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5931569337844849, "epoch": 10.98, "learning_rate": 1.4881026925485286e-05, "loss": 0.6539, "step": 12992, "task_loss": 1.3707789182662964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5461002588272095, "epoch": 10.98, "learning_rate": 1.487789605510332e-05, "loss": 0.505, "step": 12993, "task_loss": 1.2249609231948853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28133970499038696, "epoch": 10.98, "learning_rate": 1.4874765184721353e-05, "loss": 0.4805, "step": 12994, "task_loss": 0.0321483314037323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4012354612350464, "epoch": 10.98, "learning_rate": 1.4871634314339386e-05, "loss": 0.5395, "step": 12995, "task_loss": 0.3727371096611023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5444573163986206, "epoch": 10.99, "learning_rate": 1.4868503443957422e-05, "loss": 0.5705, "step": 12996, "task_loss": 1.093075156211853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5393609404563904, "epoch": 10.99, "learning_rate": 1.4865372573575453e-05, "loss": 0.4568, "step": 12997, "task_loss": 0.9585828185081482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5156993865966797, "epoch": 10.99, "learning_rate": 1.4862241703193489e-05, "loss": 0.5406, "step": 12998, "task_loss": 0.599909782409668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3674007058143616, "epoch": 10.99, "learning_rate": 1.4859110832811524e-05, "loss": 0.308, "step": 12999, "task_loss": 0.41116100549697876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8426975011825562, "epoch": 10.99, "learning_rate": 1.4855979962429555e-05, "loss": 0.6251, "step": 13000, "task_loss": 0.5863122344017029 }, { "epoch": 10.99, "eval_accuracy": 0.9072871287128713, "eval_loss": 0.34724152088165283, "eval_runtime": 208.7818, "eval_samples_per_second": 120.94, "eval_steps_per_second": 0.948, "step": 13000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23919108510017395, "epoch": 10.99, "learning_rate": 1.485284909204759e-05, "loss": 0.4913, "step": 13001, "task_loss": 0.25469347834587097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34446635842323303, "epoch": 10.99, "learning_rate": 1.4849718221665624e-05, "loss": 0.6204, "step": 13002, "task_loss": 0.6175515651702881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48932725191116333, "epoch": 10.99, "learning_rate": 1.4846587351283658e-05, "loss": 0.5252, "step": 13003, "task_loss": 0.4203982353210449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7349878549575806, "epoch": 10.99, "learning_rate": 1.4843456480901691e-05, "loss": 0.592, "step": 13004, "task_loss": 0.6738802194595337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5967287421226501, "epoch": 10.99, "learning_rate": 1.4840325610519726e-05, "loss": 0.492, "step": 13005, "task_loss": 1.011212706565857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3294861316680908, "epoch": 10.99, "learning_rate": 1.4837194740137758e-05, "loss": 0.4459, "step": 13006, "task_loss": 0.3251146674156189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45883846282958984, "epoch": 10.99, "learning_rate": 1.4834063869755793e-05, "loss": 0.4658, "step": 13007, "task_loss": 1.3146440982818604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7881167531013489, "epoch": 11.0, "learning_rate": 1.4830932999373825e-05, "loss": 0.528, "step": 13008, "task_loss": 0.6337814331054688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6913658380508423, "epoch": 11.0, "learning_rate": 1.482780212899186e-05, "loss": 0.588, "step": 13009, "task_loss": 1.8216040134429932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34649282693862915, "epoch": 11.0, "learning_rate": 1.4824671258609894e-05, "loss": 0.3633, "step": 13010, "task_loss": 0.5761299729347229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3623782992362976, "epoch": 11.0, "learning_rate": 1.4821540388227929e-05, "loss": 0.371, "step": 13011, "task_loss": 0.4070045053958893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43814903497695923, "epoch": 11.0, "learning_rate": 1.4818409517845962e-05, "loss": 0.5575, "step": 13012, "task_loss": 0.6980209946632385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.709821343421936, "epoch": 11.0, "learning_rate": 1.4815278647463996e-05, "loss": 0.4796, "step": 13013, "task_loss": 0.6388180255889893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27548983693122864, "epoch": 11.0, "learning_rate": 1.4812147777082031e-05, "loss": 0.8637, "step": 13014, "task_loss": 0.4260912537574768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37804198265075684, "epoch": 11.0, "learning_rate": 1.4809016906700063e-05, "loss": 0.5178, "step": 13015, "task_loss": 0.5690457224845886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3094586730003357, "epoch": 11.0, "learning_rate": 1.4805886036318098e-05, "loss": 0.419, "step": 13016, "task_loss": 0.6090420484542847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4734489321708679, "epoch": 11.0, "learning_rate": 1.480275516593613e-05, "loss": 0.4644, "step": 13017, "task_loss": 1.1837763786315918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8371721506118774, "epoch": 11.0, "learning_rate": 1.4799624295554165e-05, "loss": 0.7837, "step": 13018, "task_loss": 0.5424495339393616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.374021053314209, "epoch": 11.01, "learning_rate": 1.4796493425172198e-05, "loss": 0.3789, "step": 13019, "task_loss": 0.6061616539955139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40307849645614624, "epoch": 11.01, "learning_rate": 1.4793362554790233e-05, "loss": 0.499, "step": 13020, "task_loss": 0.43495601415634155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6927989721298218, "epoch": 11.01, "learning_rate": 1.4790231684408265e-05, "loss": 0.5074, "step": 13021, "task_loss": 0.8216361999511719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8790075182914734, "epoch": 11.01, "learning_rate": 1.47871008140263e-05, "loss": 0.7288, "step": 13022, "task_loss": 1.3761847019195557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5803523063659668, "epoch": 11.01, "learning_rate": 1.4783969943644336e-05, "loss": 0.6278, "step": 13023, "task_loss": 1.0695167779922485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2807234823703766, "epoch": 11.01, "learning_rate": 1.4780839073262367e-05, "loss": 0.4032, "step": 13024, "task_loss": 0.059708163142204285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6042443513870239, "epoch": 11.01, "learning_rate": 1.4777708202880403e-05, "loss": 0.5484, "step": 13025, "task_loss": 1.310827374458313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4612792134284973, "epoch": 11.01, "learning_rate": 1.4774577332498434e-05, "loss": 0.4642, "step": 13026, "task_loss": 0.2862667143344879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5681426525115967, "epoch": 11.01, "learning_rate": 1.477144646211647e-05, "loss": 0.5888, "step": 13027, "task_loss": 0.9930564165115356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3325341045856476, "epoch": 11.01, "learning_rate": 1.4768315591734503e-05, "loss": 0.4473, "step": 13028, "task_loss": 0.09000822901725769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20552915334701538, "epoch": 11.01, "learning_rate": 1.4765184721352538e-05, "loss": 0.4105, "step": 13029, "task_loss": 0.1405990719795227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5556410551071167, "epoch": 11.01, "learning_rate": 1.476205385097057e-05, "loss": 0.5158, "step": 13030, "task_loss": 0.2246694415807724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4171523153781891, "epoch": 11.02, "learning_rate": 1.4758922980588605e-05, "loss": 0.4919, "step": 13031, "task_loss": 0.453622043132782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9768198728561401, "epoch": 11.02, "learning_rate": 1.4755792110206637e-05, "loss": 0.5656, "step": 13032, "task_loss": 0.9923115372657776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5550410747528076, "epoch": 11.02, "learning_rate": 1.4752661239824672e-05, "loss": 0.4298, "step": 13033, "task_loss": 1.274035930633545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4834396243095398, "epoch": 11.02, "learning_rate": 1.4749530369442704e-05, "loss": 0.6078, "step": 13034, "task_loss": 0.46644145250320435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48726943135261536, "epoch": 11.02, "learning_rate": 1.4746399499060739e-05, "loss": 0.4571, "step": 13035, "task_loss": 0.9161332249641418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6496454477310181, "epoch": 11.02, "learning_rate": 1.4743268628678774e-05, "loss": 0.6239, "step": 13036, "task_loss": 0.7586998343467712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5942457914352417, "epoch": 11.02, "learning_rate": 1.4740137758296808e-05, "loss": 0.4568, "step": 13037, "task_loss": 0.8755385875701904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3230292797088623, "epoch": 11.02, "learning_rate": 1.4737006887914843e-05, "loss": 0.4027, "step": 13038, "task_loss": 0.8470308780670166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32889121770858765, "epoch": 11.02, "learning_rate": 1.4733876017532875e-05, "loss": 0.3667, "step": 13039, "task_loss": 0.24688993394374847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.553609311580658, "epoch": 11.02, "learning_rate": 1.473074514715091e-05, "loss": 0.5254, "step": 13040, "task_loss": 0.40914538502693176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3094998598098755, "epoch": 11.02, "learning_rate": 1.4727614276768942e-05, "loss": 0.5061, "step": 13041, "task_loss": 0.3710198700428009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37613046169281006, "epoch": 11.02, "learning_rate": 1.4724483406386977e-05, "loss": 0.4215, "step": 13042, "task_loss": 0.550649881362915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34503450989723206, "epoch": 11.03, "learning_rate": 1.4721352536005008e-05, "loss": 0.4736, "step": 13043, "task_loss": 0.3880934715270996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6477611660957336, "epoch": 11.03, "learning_rate": 1.4718221665623044e-05, "loss": 0.4702, "step": 13044, "task_loss": 0.4261380732059479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.380681574344635, "epoch": 11.03, "learning_rate": 1.4715090795241077e-05, "loss": 0.4832, "step": 13045, "task_loss": 0.4904899001121521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6270726919174194, "epoch": 11.03, "learning_rate": 1.4711959924859112e-05, "loss": 0.6243, "step": 13046, "task_loss": 0.5206149220466614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5371562242507935, "epoch": 11.03, "learning_rate": 1.4708829054477144e-05, "loss": 0.5223, "step": 13047, "task_loss": 0.07386890798807144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26792752742767334, "epoch": 11.03, "learning_rate": 1.470569818409518e-05, "loss": 0.455, "step": 13048, "task_loss": 0.11710811406373978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3144639730453491, "epoch": 11.03, "learning_rate": 1.4702567313713214e-05, "loss": 0.4735, "step": 13049, "task_loss": 0.5641815066337585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37841689586639404, "epoch": 11.03, "learning_rate": 1.4699436443331246e-05, "loss": 0.4538, "step": 13050, "task_loss": 0.8870931267738342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46718740463256836, "epoch": 11.03, "learning_rate": 1.4696305572949281e-05, "loss": 0.5679, "step": 13051, "task_loss": 0.32034680247306824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.675365149974823, "epoch": 11.03, "learning_rate": 1.4693174702567313e-05, "loss": 0.4549, "step": 13052, "task_loss": 0.1369076818227768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27478235960006714, "epoch": 11.03, "learning_rate": 1.4690043832185348e-05, "loss": 0.4835, "step": 13053, "task_loss": 0.36385348439216614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47826409339904785, "epoch": 11.03, "learning_rate": 1.4686912961803382e-05, "loss": 0.538, "step": 13054, "task_loss": 0.9481008648872375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6090825200080872, "epoch": 11.04, "learning_rate": 1.4683782091421417e-05, "loss": 0.4774, "step": 13055, "task_loss": 0.08098180592060089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4145662784576416, "epoch": 11.04, "learning_rate": 1.4680651221039449e-05, "loss": 0.4894, "step": 13056, "task_loss": 0.9175491333007812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43668776750564575, "epoch": 11.04, "learning_rate": 1.4677520350657484e-05, "loss": 0.4967, "step": 13057, "task_loss": 0.5389097332954407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.14516428112983704, "epoch": 11.04, "learning_rate": 1.4674389480275516e-05, "loss": 0.3823, "step": 13058, "task_loss": 0.5384541749954224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46949875354766846, "epoch": 11.04, "learning_rate": 1.4671258609893551e-05, "loss": 0.5532, "step": 13059, "task_loss": 0.429742693901062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2740248739719391, "epoch": 11.04, "learning_rate": 1.4668127739511586e-05, "loss": 0.5067, "step": 13060, "task_loss": 0.2545311748981476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4605979323387146, "epoch": 11.04, "learning_rate": 1.4664996869129618e-05, "loss": 0.4741, "step": 13061, "task_loss": 0.7304402589797974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36735352873802185, "epoch": 11.04, "learning_rate": 1.4661865998747653e-05, "loss": 0.5146, "step": 13062, "task_loss": 0.3009168803691864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6556042432785034, "epoch": 11.04, "learning_rate": 1.4658735128365686e-05, "loss": 0.4564, "step": 13063, "task_loss": 0.281253844499588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40933674573898315, "epoch": 11.04, "learning_rate": 1.4655604257983722e-05, "loss": 0.5361, "step": 13064, "task_loss": 0.5920759439468384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2715068459510803, "epoch": 11.04, "learning_rate": 1.4652473387601753e-05, "loss": 0.4603, "step": 13065, "task_loss": 0.4270116984844208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2581900358200073, "epoch": 11.04, "learning_rate": 1.4649342517219789e-05, "loss": 0.4335, "step": 13066, "task_loss": 0.33291009068489075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30019044876098633, "epoch": 11.05, "learning_rate": 1.464621164683782e-05, "loss": 0.4069, "step": 13067, "task_loss": 0.12497202306985855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33908239006996155, "epoch": 11.05, "learning_rate": 1.4643080776455856e-05, "loss": 0.3465, "step": 13068, "task_loss": 0.034111641347408295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.443665087223053, "epoch": 11.05, "learning_rate": 1.4639949906073889e-05, "loss": 0.6, "step": 13069, "task_loss": 0.6236209869384766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5121778249740601, "epoch": 11.05, "learning_rate": 1.4636819035691922e-05, "loss": 0.3727, "step": 13070, "task_loss": 0.12494290620088577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3509316146373749, "epoch": 11.05, "learning_rate": 1.4633688165309956e-05, "loss": 0.4025, "step": 13071, "task_loss": 0.25240346789360046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5096230506896973, "epoch": 11.05, "learning_rate": 1.4630557294927991e-05, "loss": 0.5374, "step": 13072, "task_loss": 0.4059157967567444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5372920036315918, "epoch": 11.05, "learning_rate": 1.4627426424546026e-05, "loss": 0.5555, "step": 13073, "task_loss": 1.1186754703521729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40931856632232666, "epoch": 11.05, "learning_rate": 1.4624295554164058e-05, "loss": 0.3997, "step": 13074, "task_loss": 0.16304762661457062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33590632677078247, "epoch": 11.05, "learning_rate": 1.4621164683782093e-05, "loss": 0.4391, "step": 13075, "task_loss": 0.24554000794887543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33866339921951294, "epoch": 11.05, "learning_rate": 1.4618033813400125e-05, "loss": 0.4395, "step": 13076, "task_loss": 0.1453637182712555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2938881814479828, "epoch": 11.05, "learning_rate": 1.461490294301816e-05, "loss": 0.5101, "step": 13077, "task_loss": 1.006699562072754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40744754672050476, "epoch": 11.05, "learning_rate": 1.4611772072636194e-05, "loss": 0.5144, "step": 13078, "task_loss": 0.6520411968231201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30532020330429077, "epoch": 11.06, "learning_rate": 1.4608641202254227e-05, "loss": 0.5207, "step": 13079, "task_loss": 0.17396670579910278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5393151044845581, "epoch": 11.06, "learning_rate": 1.460551033187226e-05, "loss": 0.5334, "step": 13080, "task_loss": 0.7658246755599976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3279956877231598, "epoch": 11.06, "learning_rate": 1.4602379461490296e-05, "loss": 0.498, "step": 13081, "task_loss": 0.43916431069374084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5923595428466797, "epoch": 11.06, "learning_rate": 1.4599248591108328e-05, "loss": 0.5535, "step": 13082, "task_loss": 0.776157557964325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9341280460357666, "epoch": 11.06, "learning_rate": 1.4596117720726363e-05, "loss": 0.5969, "step": 13083, "task_loss": 0.9745886921882629 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8020813465118408, "epoch": 11.06, "learning_rate": 1.4592986850344395e-05, "loss": 0.6046, "step": 13084, "task_loss": 1.5725806951522827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6425871253013611, "epoch": 11.06, "learning_rate": 1.458985597996243e-05, "loss": 0.5134, "step": 13085, "task_loss": 0.9547855854034424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4116685390472412, "epoch": 11.06, "learning_rate": 1.4586725109580465e-05, "loss": 0.444, "step": 13086, "task_loss": 0.37350016832351685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3192428648471832, "epoch": 11.06, "learning_rate": 1.4583594239198498e-05, "loss": 0.3526, "step": 13087, "task_loss": 0.23310734331607819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5088917016983032, "epoch": 11.06, "learning_rate": 1.4580463368816532e-05, "loss": 0.4851, "step": 13088, "task_loss": 0.5020776391029358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6137377619743347, "epoch": 11.06, "learning_rate": 1.4577332498434565e-05, "loss": 0.5166, "step": 13089, "task_loss": 0.8358263373374939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2568938136100769, "epoch": 11.07, "learning_rate": 1.45742016280526e-05, "loss": 0.5607, "step": 13090, "task_loss": 0.4387325942516327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46591275930404663, "epoch": 11.07, "learning_rate": 1.4571070757670632e-05, "loss": 0.4979, "step": 13091, "task_loss": 1.1312452554702759 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6518567800521851, "epoch": 11.07, "learning_rate": 1.4567939887288667e-05, "loss": 0.4982, "step": 13092, "task_loss": 0.515493631362915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5061675906181335, "epoch": 11.07, "learning_rate": 1.45648090169067e-05, "loss": 0.5739, "step": 13093, "task_loss": 1.3903473615646362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.459321528673172, "epoch": 11.07, "learning_rate": 1.4561678146524734e-05, "loss": 0.516, "step": 13094, "task_loss": 0.6768597960472107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38846221566200256, "epoch": 11.07, "learning_rate": 1.4558547276142768e-05, "loss": 0.6549, "step": 13095, "task_loss": 0.8264008164405823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6356916427612305, "epoch": 11.07, "learning_rate": 1.4555416405760803e-05, "loss": 0.5127, "step": 13096, "task_loss": 1.188524842262268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27215903997421265, "epoch": 11.07, "learning_rate": 1.4552285535378836e-05, "loss": 0.4572, "step": 13097, "task_loss": 1.6521813869476318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3036625385284424, "epoch": 11.07, "learning_rate": 1.454915466499687e-05, "loss": 0.4905, "step": 13098, "task_loss": 0.8045509457588196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2629750370979309, "epoch": 11.07, "learning_rate": 1.4546023794614905e-05, "loss": 0.4463, "step": 13099, "task_loss": 0.11819253116846085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3638831377029419, "epoch": 11.07, "learning_rate": 1.4542892924232937e-05, "loss": 0.5423, "step": 13100, "task_loss": 0.5149403810501099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5072722434997559, "epoch": 11.07, "learning_rate": 1.4539762053850972e-05, "loss": 0.4775, "step": 13101, "task_loss": 0.5872755646705627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5935639142990112, "epoch": 11.08, "learning_rate": 1.4536631183469004e-05, "loss": 0.4242, "step": 13102, "task_loss": 0.333985835313797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7243087291717529, "epoch": 11.08, "learning_rate": 1.4533500313087039e-05, "loss": 0.509, "step": 13103, "task_loss": 0.7240154147148132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8439787030220032, "epoch": 11.08, "learning_rate": 1.4530369442705072e-05, "loss": 0.6707, "step": 13104, "task_loss": 0.5600200295448303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41004839539527893, "epoch": 11.08, "learning_rate": 1.4527238572323108e-05, "loss": 0.4946, "step": 13105, "task_loss": 1.2563351392745972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26586633920669556, "epoch": 11.08, "learning_rate": 1.452410770194114e-05, "loss": 0.372, "step": 13106, "task_loss": 0.10178890079259872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44184553623199463, "epoch": 11.08, "learning_rate": 1.4520976831559175e-05, "loss": 0.3657, "step": 13107, "task_loss": 0.4144172966480255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43662118911743164, "epoch": 11.08, "learning_rate": 1.4517845961177206e-05, "loss": 0.4846, "step": 13108, "task_loss": 0.8669463396072388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3287960886955261, "epoch": 11.08, "learning_rate": 1.4514715090795242e-05, "loss": 0.4059, "step": 13109, "task_loss": 0.6806660890579224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3585226535797119, "epoch": 11.08, "learning_rate": 1.4511584220413277e-05, "loss": 0.5311, "step": 13110, "task_loss": 0.5416846871376038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4868028461933136, "epoch": 11.08, "learning_rate": 1.4508453350031309e-05, "loss": 0.5919, "step": 13111, "task_loss": 0.20589281618595123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6657900214195251, "epoch": 11.08, "learning_rate": 1.4505322479649344e-05, "loss": 0.4991, "step": 13112, "task_loss": 0.6526437997817993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6722291707992554, "epoch": 11.08, "learning_rate": 1.4502191609267377e-05, "loss": 0.4327, "step": 13113, "task_loss": 0.5224518775939941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.565173864364624, "epoch": 11.09, "learning_rate": 1.4499060738885412e-05, "loss": 0.5147, "step": 13114, "task_loss": 0.3616068363189697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4649798572063446, "epoch": 11.09, "learning_rate": 1.4495929868503444e-05, "loss": 0.5192, "step": 13115, "task_loss": 1.0644752979278564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5138081312179565, "epoch": 11.09, "learning_rate": 1.449279899812148e-05, "loss": 0.4835, "step": 13116, "task_loss": 0.9669243693351746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7655719518661499, "epoch": 11.09, "learning_rate": 1.4489668127739511e-05, "loss": 0.5272, "step": 13117, "task_loss": 0.5530925393104553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34662705659866333, "epoch": 11.09, "learning_rate": 1.4486537257357546e-05, "loss": 0.6359, "step": 13118, "task_loss": 0.7508710026741028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.520668625831604, "epoch": 11.09, "learning_rate": 1.4483406386975578e-05, "loss": 0.5159, "step": 13119, "task_loss": 1.1597384214401245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5002853870391846, "epoch": 11.09, "learning_rate": 1.4480275516593613e-05, "loss": 0.398, "step": 13120, "task_loss": 0.13805168867111206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38929468393325806, "epoch": 11.09, "learning_rate": 1.4477144646211648e-05, "loss": 0.5841, "step": 13121, "task_loss": 1.1159552335739136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5328454971313477, "epoch": 11.09, "learning_rate": 1.4474013775829682e-05, "loss": 0.5102, "step": 13122, "task_loss": 1.0846034288406372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.645319402217865, "epoch": 11.09, "learning_rate": 1.4470882905447717e-05, "loss": 0.3994, "step": 13123, "task_loss": 0.7471615076065063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.578123152256012, "epoch": 11.09, "learning_rate": 1.4467752035065749e-05, "loss": 0.4353, "step": 13124, "task_loss": 0.7879818081855774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40190935134887695, "epoch": 11.09, "learning_rate": 1.4464621164683784e-05, "loss": 0.429, "step": 13125, "task_loss": 0.28055495023727417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45387428998947144, "epoch": 11.1, "learning_rate": 1.4461490294301816e-05, "loss": 0.4576, "step": 13126, "task_loss": 0.5251865983009338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4863726794719696, "epoch": 11.1, "learning_rate": 1.4458359423919851e-05, "loss": 0.7318, "step": 13127, "task_loss": 1.4485580921173096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3584274351596832, "epoch": 11.1, "learning_rate": 1.4455228553537883e-05, "loss": 0.542, "step": 13128, "task_loss": 0.20114322006702423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2880040109157562, "epoch": 11.1, "learning_rate": 1.4452097683155918e-05, "loss": 0.4942, "step": 13129, "task_loss": 0.9713569283485413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3380356431007385, "epoch": 11.1, "learning_rate": 1.4448966812773951e-05, "loss": 0.5209, "step": 13130, "task_loss": 0.5317831635475159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4938916563987732, "epoch": 11.1, "learning_rate": 1.4445835942391986e-05, "loss": 0.3531, "step": 13131, "task_loss": 0.24766191840171814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4069693088531494, "epoch": 11.1, "learning_rate": 1.4442705072010018e-05, "loss": 0.4859, "step": 13132, "task_loss": 0.039769530296325684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38436320424079895, "epoch": 11.1, "learning_rate": 1.4439574201628053e-05, "loss": 0.4086, "step": 13133, "task_loss": 0.18400083482265472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29641613364219666, "epoch": 11.1, "learning_rate": 1.4436443331246089e-05, "loss": 0.4183, "step": 13134, "task_loss": 1.2936887741088867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4508240222930908, "epoch": 11.1, "learning_rate": 1.443331246086412e-05, "loss": 0.5807, "step": 13135, "task_loss": 0.6858047246932983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6286863088607788, "epoch": 11.1, "learning_rate": 1.4430181590482156e-05, "loss": 0.6119, "step": 13136, "task_loss": 0.914941668510437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5569921731948853, "epoch": 11.1, "learning_rate": 1.4427050720100187e-05, "loss": 0.4823, "step": 13137, "task_loss": 1.0559589862823486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6006746292114258, "epoch": 11.11, "learning_rate": 1.4423919849718223e-05, "loss": 0.5839, "step": 13138, "task_loss": 0.7467256188392639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2709413170814514, "epoch": 11.11, "learning_rate": 1.4420788979336256e-05, "loss": 0.465, "step": 13139, "task_loss": 0.5315667986869812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5258362889289856, "epoch": 11.11, "learning_rate": 1.4417658108954291e-05, "loss": 0.5506, "step": 13140, "task_loss": 0.5237521529197693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35446035861968994, "epoch": 11.11, "learning_rate": 1.4414527238572323e-05, "loss": 0.5931, "step": 13141, "task_loss": 1.33598792552948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3887791931629181, "epoch": 11.11, "learning_rate": 1.4411396368190358e-05, "loss": 0.5213, "step": 13142, "task_loss": 0.18797440826892853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27110421657562256, "epoch": 11.11, "learning_rate": 1.440826549780839e-05, "loss": 0.4081, "step": 13143, "task_loss": 0.5429233312606812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40161311626434326, "epoch": 11.11, "learning_rate": 1.4405134627426425e-05, "loss": 0.4622, "step": 13144, "task_loss": 0.27172499895095825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29671236872673035, "epoch": 11.11, "learning_rate": 1.4402003757044459e-05, "loss": 0.3923, "step": 13145, "task_loss": 0.4694913327693939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5721628665924072, "epoch": 11.11, "learning_rate": 1.4398872886662492e-05, "loss": 0.5097, "step": 13146, "task_loss": 0.9745218753814697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6207249164581299, "epoch": 11.11, "learning_rate": 1.4395742016280527e-05, "loss": 0.5304, "step": 13147, "task_loss": 1.3214672803878784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5831857919692993, "epoch": 11.11, "learning_rate": 1.439261114589856e-05, "loss": 0.5176, "step": 13148, "task_loss": 1.0753369331359863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4719722270965576, "epoch": 11.11, "learning_rate": 1.4389480275516596e-05, "loss": 0.3926, "step": 13149, "task_loss": 0.5998983383178711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2952868938446045, "epoch": 11.12, "learning_rate": 1.4386349405134628e-05, "loss": 0.3159, "step": 13150, "task_loss": 0.5453146696090698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.770187497138977, "epoch": 11.12, "learning_rate": 1.4383218534752663e-05, "loss": 0.5872, "step": 13151, "task_loss": 0.8995476365089417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4915262460708618, "epoch": 11.12, "learning_rate": 1.4380087664370695e-05, "loss": 0.4565, "step": 13152, "task_loss": 0.19908545911312103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7197775840759277, "epoch": 11.12, "learning_rate": 1.437695679398873e-05, "loss": 0.5296, "step": 13153, "task_loss": 1.9401793479919434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7649619579315186, "epoch": 11.12, "learning_rate": 1.4373825923606763e-05, "loss": 0.633, "step": 13154, "task_loss": 0.6499426364898682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3979906737804413, "epoch": 11.12, "learning_rate": 1.4370695053224797e-05, "loss": 0.54, "step": 13155, "task_loss": 0.5579807758331299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3885861039161682, "epoch": 11.12, "learning_rate": 1.436756418284283e-05, "loss": 0.4301, "step": 13156, "task_loss": 0.2763407230377197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36842235922813416, "epoch": 11.12, "learning_rate": 1.4364433312460865e-05, "loss": 0.5671, "step": 13157, "task_loss": 0.23565638065338135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6021231412887573, "epoch": 11.12, "learning_rate": 1.43613024420789e-05, "loss": 0.5558, "step": 13158, "task_loss": 0.733422040939331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5044393539428711, "epoch": 11.12, "learning_rate": 1.4358171571696932e-05, "loss": 0.5466, "step": 13159, "task_loss": 1.65714693069458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.444097638130188, "epoch": 11.12, "learning_rate": 1.4355040701314967e-05, "loss": 0.4414, "step": 13160, "task_loss": 0.08199509233236313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32984820008277893, "epoch": 11.13, "learning_rate": 1.4351909830933e-05, "loss": 0.409, "step": 13161, "task_loss": 0.6242150068283081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6031490564346313, "epoch": 11.13, "learning_rate": 1.4348778960551034e-05, "loss": 0.5983, "step": 13162, "task_loss": 0.8867161870002747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4407880902290344, "epoch": 11.13, "learning_rate": 1.4345648090169068e-05, "loss": 0.3641, "step": 13163, "task_loss": 0.8426035642623901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4901919364929199, "epoch": 11.13, "learning_rate": 1.4342517219787101e-05, "loss": 0.4977, "step": 13164, "task_loss": 0.33343127369880676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.577907919883728, "epoch": 11.13, "learning_rate": 1.4339386349405135e-05, "loss": 0.4889, "step": 13165, "task_loss": 1.0197012424468994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9596951007843018, "epoch": 11.13, "learning_rate": 1.433625547902317e-05, "loss": 0.5917, "step": 13166, "task_loss": 0.845793604850769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4152587354183197, "epoch": 11.13, "learning_rate": 1.4333124608641202e-05, "loss": 0.5845, "step": 13167, "task_loss": 0.9543878436088562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3973087668418884, "epoch": 11.13, "learning_rate": 1.4329993738259237e-05, "loss": 0.513, "step": 13168, "task_loss": 0.7785236239433289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3125378489494324, "epoch": 11.13, "learning_rate": 1.4326862867877269e-05, "loss": 0.4471, "step": 13169, "task_loss": 0.5805445313453674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5977421402931213, "epoch": 11.13, "learning_rate": 1.4323731997495304e-05, "loss": 0.5588, "step": 13170, "task_loss": 0.4986657202243805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5492769479751587, "epoch": 11.13, "learning_rate": 1.4320601127113339e-05, "loss": 0.4401, "step": 13171, "task_loss": 0.2458019107580185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2321518361568451, "epoch": 11.13, "learning_rate": 1.4317470256731373e-05, "loss": 0.5861, "step": 13172, "task_loss": 0.44427430629730225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7436115145683289, "epoch": 11.14, "learning_rate": 1.4314339386349408e-05, "loss": 0.5869, "step": 13173, "task_loss": 0.9824748635292053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.438058465719223, "epoch": 11.14, "learning_rate": 1.431120851596744e-05, "loss": 0.4992, "step": 13174, "task_loss": 0.3955536484718323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29221001267433167, "epoch": 11.14, "learning_rate": 1.4308077645585475e-05, "loss": 0.465, "step": 13175, "task_loss": 0.4074991047382355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7542937994003296, "epoch": 11.14, "learning_rate": 1.4304946775203506e-05, "loss": 0.5794, "step": 13176, "task_loss": 2.6634442806243896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8386727571487427, "epoch": 11.14, "learning_rate": 1.4301815904821542e-05, "loss": 0.4867, "step": 13177, "task_loss": 0.9797796010971069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45879340171813965, "epoch": 11.14, "learning_rate": 1.4298685034439573e-05, "loss": 0.4841, "step": 13178, "task_loss": 0.4060736894607544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2710523009300232, "epoch": 11.14, "learning_rate": 1.4295554164057609e-05, "loss": 0.4513, "step": 13179, "task_loss": 0.7370314002037048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6449586749076843, "epoch": 11.14, "learning_rate": 1.4292423293675642e-05, "loss": 0.4785, "step": 13180, "task_loss": 1.022552251815796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3287096917629242, "epoch": 11.14, "learning_rate": 1.4289292423293677e-05, "loss": 0.5893, "step": 13181, "task_loss": 0.40475884079933167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33463582396507263, "epoch": 11.14, "learning_rate": 1.4286161552911709e-05, "loss": 0.4925, "step": 13182, "task_loss": 0.403172105550766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35579749941825867, "epoch": 11.14, "learning_rate": 1.4283030682529744e-05, "loss": 0.5838, "step": 13183, "task_loss": 0.5845431685447693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23319938778877258, "epoch": 11.14, "learning_rate": 1.427989981214778e-05, "loss": 0.4146, "step": 13184, "task_loss": 0.6056206226348877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4693854749202728, "epoch": 11.15, "learning_rate": 1.4276768941765811e-05, "loss": 0.562, "step": 13185, "task_loss": 0.4751952886581421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5419719219207764, "epoch": 11.15, "learning_rate": 1.4273638071383846e-05, "loss": 0.3901, "step": 13186, "task_loss": 0.6327316761016846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2388828992843628, "epoch": 11.15, "learning_rate": 1.4270507201001878e-05, "loss": 0.4488, "step": 13187, "task_loss": 0.980025053024292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5559141635894775, "epoch": 11.15, "learning_rate": 1.4267376330619913e-05, "loss": 0.4101, "step": 13188, "task_loss": 0.3900673985481262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3895641565322876, "epoch": 11.15, "learning_rate": 1.4264245460237947e-05, "loss": 0.4007, "step": 13189, "task_loss": 0.1313076615333557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36458319425582886, "epoch": 11.15, "learning_rate": 1.4261114589855982e-05, "loss": 0.4744, "step": 13190, "task_loss": 0.45044463872909546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4178618788719177, "epoch": 11.15, "learning_rate": 1.4257983719474014e-05, "loss": 0.391, "step": 13191, "task_loss": 0.6893482208251953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8138992190361023, "epoch": 11.15, "learning_rate": 1.4254852849092049e-05, "loss": 0.6976, "step": 13192, "task_loss": 0.742251455783844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3954484760761261, "epoch": 11.15, "learning_rate": 1.425172197871008e-05, "loss": 0.4981, "step": 13193, "task_loss": 0.1761525422334671 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5883210897445679, "epoch": 11.15, "learning_rate": 1.4248591108328116e-05, "loss": 0.4663, "step": 13194, "task_loss": 0.8451892137527466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42798855900764465, "epoch": 11.15, "learning_rate": 1.4245460237946151e-05, "loss": 0.514, "step": 13195, "task_loss": 0.7902276515960693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6536253690719604, "epoch": 11.15, "learning_rate": 1.4242329367564183e-05, "loss": 0.4995, "step": 13196, "task_loss": 0.7941808700561523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5552327632904053, "epoch": 11.16, "learning_rate": 1.4239198497182218e-05, "loss": 0.6094, "step": 13197, "task_loss": 0.4238063395023346 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5888668298721313, "epoch": 11.16, "learning_rate": 1.4236067626800251e-05, "loss": 0.4935, "step": 13198, "task_loss": 0.6445841193199158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.419008731842041, "epoch": 11.16, "learning_rate": 1.4232936756418287e-05, "loss": 0.4347, "step": 13199, "task_loss": 1.1716899871826172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4404166340827942, "epoch": 11.16, "learning_rate": 1.4229805886036318e-05, "loss": 0.4249, "step": 13200, "task_loss": 0.5378589034080505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38625362515449524, "epoch": 11.16, "learning_rate": 1.4226675015654353e-05, "loss": 0.5414, "step": 13201, "task_loss": 0.5382758975028992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43437549471855164, "epoch": 11.16, "learning_rate": 1.4223544145272385e-05, "loss": 0.4246, "step": 13202, "task_loss": 1.0732619762420654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7864535450935364, "epoch": 11.16, "learning_rate": 1.422041327489042e-05, "loss": 0.5137, "step": 13203, "task_loss": 0.9043009877204895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5922012329101562, "epoch": 11.16, "learning_rate": 1.4217282404508452e-05, "loss": 0.4071, "step": 13204, "task_loss": 1.0140758752822876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5887234210968018, "epoch": 11.16, "learning_rate": 1.4214151534126487e-05, "loss": 0.5469, "step": 13205, "task_loss": 0.564475953578949 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6358050107955933, "epoch": 11.16, "learning_rate": 1.4211020663744521e-05, "loss": 0.4783, "step": 13206, "task_loss": 0.6870816349983215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3665807843208313, "epoch": 11.16, "learning_rate": 1.4207889793362556e-05, "loss": 0.367, "step": 13207, "task_loss": 0.6418401598930359 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5650041103363037, "epoch": 11.16, "learning_rate": 1.4204758922980591e-05, "loss": 0.5072, "step": 13208, "task_loss": 0.8451234102249146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4286324679851532, "epoch": 11.17, "learning_rate": 1.4201628052598623e-05, "loss": 0.5213, "step": 13209, "task_loss": 0.6251382827758789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33412104845046997, "epoch": 11.17, "learning_rate": 1.4198497182216658e-05, "loss": 0.4741, "step": 13210, "task_loss": 0.3120884299278259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5967466235160828, "epoch": 11.17, "learning_rate": 1.419536631183469e-05, "loss": 0.6685, "step": 13211, "task_loss": 0.8653411865234375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2374182939529419, "epoch": 11.17, "learning_rate": 1.4192235441452725e-05, "loss": 0.4215, "step": 13212, "task_loss": 0.10742975026369095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39385032653808594, "epoch": 11.17, "learning_rate": 1.4189104571070757e-05, "loss": 0.4081, "step": 13213, "task_loss": 0.832152783870697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5562753677368164, "epoch": 11.17, "learning_rate": 1.4185973700688792e-05, "loss": 0.474, "step": 13214, "task_loss": 0.37352967262268066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46753329038619995, "epoch": 11.17, "learning_rate": 1.4182842830306826e-05, "loss": 0.5001, "step": 13215, "task_loss": 0.6786234378814697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3917957544326782, "epoch": 11.17, "learning_rate": 1.417971195992486e-05, "loss": 0.4588, "step": 13216, "task_loss": 1.0042731761932373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4132229685783386, "epoch": 11.17, "learning_rate": 1.4176581089542892e-05, "loss": 0.4966, "step": 13217, "task_loss": 0.942795991897583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.436436265707016, "epoch": 11.17, "learning_rate": 1.4173450219160928e-05, "loss": 0.5121, "step": 13218, "task_loss": 0.40166106820106506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28221839666366577, "epoch": 11.17, "learning_rate": 1.4170319348778963e-05, "loss": 0.3783, "step": 13219, "task_loss": 0.5945196151733398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.365206241607666, "epoch": 11.17, "learning_rate": 1.4167188478396995e-05, "loss": 0.393, "step": 13220, "task_loss": 0.5832639932632446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5386824011802673, "epoch": 11.18, "learning_rate": 1.416405760801503e-05, "loss": 0.3974, "step": 13221, "task_loss": 0.5870251059532166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3097211718559265, "epoch": 11.18, "learning_rate": 1.4160926737633062e-05, "loss": 0.4295, "step": 13222, "task_loss": 0.34883496165275574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5134425163269043, "epoch": 11.18, "learning_rate": 1.4157795867251097e-05, "loss": 0.5394, "step": 13223, "task_loss": 1.0657379627227783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37326186895370483, "epoch": 11.18, "learning_rate": 1.415466499686913e-05, "loss": 0.5199, "step": 13224, "task_loss": 0.5661720633506775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6725691556930542, "epoch": 11.18, "learning_rate": 1.4151534126487165e-05, "loss": 0.6219, "step": 13225, "task_loss": 0.4823896288871765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2519083023071289, "epoch": 11.18, "learning_rate": 1.4148403256105197e-05, "loss": 0.5105, "step": 13226, "task_loss": 0.685124397277832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20544150471687317, "epoch": 11.18, "learning_rate": 1.4145272385723232e-05, "loss": 0.4893, "step": 13227, "task_loss": 0.02814428322017193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32645291090011597, "epoch": 11.18, "learning_rate": 1.4142141515341264e-05, "loss": 0.4846, "step": 13228, "task_loss": 0.4470427632331848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4351954162120819, "epoch": 11.18, "learning_rate": 1.41390106449593e-05, "loss": 0.3742, "step": 13229, "task_loss": 0.847345232963562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5972213745117188, "epoch": 11.18, "learning_rate": 1.4135879774577333e-05, "loss": 0.5903, "step": 13230, "task_loss": 0.9307670593261719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.950821042060852, "epoch": 11.18, "learning_rate": 1.4132748904195366e-05, "loss": 0.5741, "step": 13231, "task_loss": 1.1865476369857788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3103441894054413, "epoch": 11.19, "learning_rate": 1.4129618033813401e-05, "loss": 0.5034, "step": 13232, "task_loss": 0.43033361434936523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49941837787628174, "epoch": 11.19, "learning_rate": 1.4126487163431435e-05, "loss": 0.5316, "step": 13233, "task_loss": 0.4832957684993744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31886744499206543, "epoch": 11.19, "learning_rate": 1.412335629304947e-05, "loss": 0.5597, "step": 13234, "task_loss": 0.45855340361595154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5851520299911499, "epoch": 11.19, "learning_rate": 1.4120225422667502e-05, "loss": 0.5536, "step": 13235, "task_loss": 0.4923832416534424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5552916526794434, "epoch": 11.19, "learning_rate": 1.4117094552285537e-05, "loss": 0.3587, "step": 13236, "task_loss": 1.0047193765640259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30003952980041504, "epoch": 11.19, "learning_rate": 1.4113963681903569e-05, "loss": 0.5149, "step": 13237, "task_loss": 0.9781810641288757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3893689513206482, "epoch": 11.19, "learning_rate": 1.4110832811521604e-05, "loss": 0.4777, "step": 13238, "task_loss": 0.6712619066238403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48092395067214966, "epoch": 11.19, "learning_rate": 1.4107701941139637e-05, "loss": 0.5737, "step": 13239, "task_loss": 1.3951265811920166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47012215852737427, "epoch": 11.19, "learning_rate": 1.4104571070757671e-05, "loss": 0.4719, "step": 13240, "task_loss": 0.41817939281463623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20759472250938416, "epoch": 11.19, "learning_rate": 1.4101440200375704e-05, "loss": 0.4756, "step": 13241, "task_loss": 0.37488269805908203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3982604444026947, "epoch": 11.19, "learning_rate": 1.409830932999374e-05, "loss": 0.4644, "step": 13242, "task_loss": 0.8639428615570068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5310929417610168, "epoch": 11.19, "learning_rate": 1.4095178459611771e-05, "loss": 0.5041, "step": 13243, "task_loss": 0.7785470485687256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4226257801055908, "epoch": 11.2, "learning_rate": 1.4092047589229806e-05, "loss": 0.4931, "step": 13244, "task_loss": 1.1175748109817505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3900502026081085, "epoch": 11.2, "learning_rate": 1.4088916718847842e-05, "loss": 0.4733, "step": 13245, "task_loss": 0.30633729696273804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5064839124679565, "epoch": 11.2, "learning_rate": 1.4085785848465873e-05, "loss": 0.5667, "step": 13246, "task_loss": 0.3366199731826782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21139848232269287, "epoch": 11.2, "learning_rate": 1.4082654978083909e-05, "loss": 0.5294, "step": 13247, "task_loss": 0.04004925116896629 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28339287638664246, "epoch": 11.2, "learning_rate": 1.4079524107701942e-05, "loss": 0.454, "step": 13248, "task_loss": 0.6925181150436401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40065956115722656, "epoch": 11.2, "learning_rate": 1.4076393237319977e-05, "loss": 0.4641, "step": 13249, "task_loss": 0.7961905598640442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5779479742050171, "epoch": 11.2, "learning_rate": 1.4073262366938009e-05, "loss": 0.5821, "step": 13250, "task_loss": 0.47718751430511475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8104084730148315, "epoch": 11.2, "learning_rate": 1.4070131496556044e-05, "loss": 0.6283, "step": 13251, "task_loss": 0.3491058051586151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4351203441619873, "epoch": 11.2, "learning_rate": 1.4067000626174076e-05, "loss": 0.4585, "step": 13252, "task_loss": 0.2968767583370209 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3597986102104187, "epoch": 11.2, "learning_rate": 1.4063869755792111e-05, "loss": 0.4362, "step": 13253, "task_loss": 0.26886072754859924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4179709255695343, "epoch": 11.2, "learning_rate": 1.4060738885410143e-05, "loss": 0.5111, "step": 13254, "task_loss": 0.7210264205932617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23404669761657715, "epoch": 11.2, "learning_rate": 1.4057608015028178e-05, "loss": 0.4181, "step": 13255, "task_loss": 0.250771164894104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3950282037258148, "epoch": 11.21, "learning_rate": 1.4054477144646213e-05, "loss": 0.4377, "step": 13256, "task_loss": 0.4034884572029114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39902400970458984, "epoch": 11.21, "learning_rate": 1.4051346274264247e-05, "loss": 0.5858, "step": 13257, "task_loss": 0.9891909956932068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5925943851470947, "epoch": 11.21, "learning_rate": 1.4048215403882282e-05, "loss": 0.4648, "step": 13258, "task_loss": 0.38543227314949036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0347236394882202, "epoch": 11.21, "learning_rate": 1.4045084533500314e-05, "loss": 0.733, "step": 13259, "task_loss": 1.341744065284729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5951594114303589, "epoch": 11.21, "learning_rate": 1.4041953663118349e-05, "loss": 0.582, "step": 13260, "task_loss": 0.4697588384151459 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.708172082901001, "epoch": 11.21, "learning_rate": 1.403882279273638e-05, "loss": 0.6316, "step": 13261, "task_loss": 1.0934303998947144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5608752965927124, "epoch": 11.21, "learning_rate": 1.4035691922354416e-05, "loss": 0.5448, "step": 13262, "task_loss": 0.1071329414844513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5456868410110474, "epoch": 11.21, "learning_rate": 1.4032561051972448e-05, "loss": 0.484, "step": 13263, "task_loss": 1.0444893836975098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4548012614250183, "epoch": 11.21, "learning_rate": 1.4029430181590483e-05, "loss": 0.4773, "step": 13264, "task_loss": 0.7040666341781616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4911329746246338, "epoch": 11.21, "learning_rate": 1.4026299311208516e-05, "loss": 0.4541, "step": 13265, "task_loss": 0.28080007433891296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26982492208480835, "epoch": 11.21, "learning_rate": 1.4023168440826551e-05, "loss": 0.4276, "step": 13266, "task_loss": 0.08006531745195389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38239923119544983, "epoch": 11.21, "learning_rate": 1.4020037570444583e-05, "loss": 0.4807, "step": 13267, "task_loss": 0.1813192069530487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5898103713989258, "epoch": 11.22, "learning_rate": 1.4016906700062618e-05, "loss": 0.4208, "step": 13268, "task_loss": 0.9345130920410156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4213978052139282, "epoch": 11.22, "learning_rate": 1.4013775829680654e-05, "loss": 0.3761, "step": 13269, "task_loss": 0.4766980707645416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45255059003829956, "epoch": 11.22, "learning_rate": 1.4010644959298685e-05, "loss": 0.5529, "step": 13270, "task_loss": 0.5100817084312439 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31460732221603394, "epoch": 11.22, "learning_rate": 1.400751408891672e-05, "loss": 0.5645, "step": 13271, "task_loss": 0.0837274044752121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5770588517189026, "epoch": 11.22, "learning_rate": 1.4004383218534752e-05, "loss": 0.5022, "step": 13272, "task_loss": 0.5942583680152893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41492772102355957, "epoch": 11.22, "learning_rate": 1.4001252348152787e-05, "loss": 0.5456, "step": 13273, "task_loss": 0.8636767864227295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3011431097984314, "epoch": 11.22, "learning_rate": 1.3998121477770821e-05, "loss": 0.4147, "step": 13274, "task_loss": 0.2655058801174164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5572324395179749, "epoch": 11.22, "learning_rate": 1.3994990607388856e-05, "loss": 0.5637, "step": 13275, "task_loss": 1.798477053642273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5172795653343201, "epoch": 11.22, "learning_rate": 1.3991859737006888e-05, "loss": 0.4987, "step": 13276, "task_loss": 0.16365541517734528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.577811062335968, "epoch": 11.22, "learning_rate": 1.3988728866624923e-05, "loss": 0.4881, "step": 13277, "task_loss": 1.481948971748352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3606798052787781, "epoch": 11.22, "learning_rate": 1.3985597996242955e-05, "loss": 0.4123, "step": 13278, "task_loss": 0.6718652844429016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.573148787021637, "epoch": 11.22, "learning_rate": 1.398246712586099e-05, "loss": 0.5835, "step": 13279, "task_loss": 0.4622833728790283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3187636137008667, "epoch": 11.23, "learning_rate": 1.3979336255479022e-05, "loss": 0.4981, "step": 13280, "task_loss": 0.38378626108169556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5403709411621094, "epoch": 11.23, "learning_rate": 1.3976205385097057e-05, "loss": 0.5349, "step": 13281, "task_loss": 1.3844964504241943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4301987886428833, "epoch": 11.23, "learning_rate": 1.3973074514715092e-05, "loss": 0.4652, "step": 13282, "task_loss": 0.5958496928215027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6901684403419495, "epoch": 11.23, "learning_rate": 1.3969943644333126e-05, "loss": 0.5324, "step": 13283, "task_loss": 1.3842175006866455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49343442916870117, "epoch": 11.23, "learning_rate": 1.396681277395116e-05, "loss": 0.4873, "step": 13284, "task_loss": 0.8048269152641296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5978333353996277, "epoch": 11.23, "learning_rate": 1.3963681903569193e-05, "loss": 0.569, "step": 13285, "task_loss": 0.3228929340839386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29157164692878723, "epoch": 11.23, "learning_rate": 1.3960551033187228e-05, "loss": 0.3482, "step": 13286, "task_loss": 0.16487431526184082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42945200204849243, "epoch": 11.23, "learning_rate": 1.395742016280526e-05, "loss": 0.4387, "step": 13287, "task_loss": 0.6698407530784607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6149923205375671, "epoch": 11.23, "learning_rate": 1.3954289292423295e-05, "loss": 0.5342, "step": 13288, "task_loss": 0.2782036364078522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3913559913635254, "epoch": 11.23, "learning_rate": 1.3951158422041326e-05, "loss": 0.4487, "step": 13289, "task_loss": 0.6002231240272522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3359200954437256, "epoch": 11.23, "learning_rate": 1.3948027551659362e-05, "loss": 0.3556, "step": 13290, "task_loss": 0.6874136328697205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7455294728279114, "epoch": 11.23, "learning_rate": 1.3944896681277395e-05, "loss": 0.8372, "step": 13291, "task_loss": 1.2128944396972656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5527111291885376, "epoch": 11.24, "learning_rate": 1.394176581089543e-05, "loss": 0.6204, "step": 13292, "task_loss": 0.5505964756011963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.485586941242218, "epoch": 11.24, "learning_rate": 1.3938634940513465e-05, "loss": 0.6543, "step": 13293, "task_loss": 0.3310738801956177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5086438655853271, "epoch": 11.24, "learning_rate": 1.3935504070131497e-05, "loss": 0.5155, "step": 13294, "task_loss": 0.732918381690979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40117931365966797, "epoch": 11.24, "learning_rate": 1.3932373199749532e-05, "loss": 0.4738, "step": 13295, "task_loss": 0.9306606650352478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6628032922744751, "epoch": 11.24, "learning_rate": 1.3929242329367564e-05, "loss": 0.5077, "step": 13296, "task_loss": 1.3659861087799072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24670808017253876, "epoch": 11.24, "learning_rate": 1.39261114589856e-05, "loss": 0.3611, "step": 13297, "task_loss": 0.08270201832056046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32237881422042847, "epoch": 11.24, "learning_rate": 1.3922980588603631e-05, "loss": 0.3379, "step": 13298, "task_loss": 0.494404673576355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4081307649612427, "epoch": 11.24, "learning_rate": 1.3919849718221666e-05, "loss": 0.6222, "step": 13299, "task_loss": 0.4576394259929657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3797703683376312, "epoch": 11.24, "learning_rate": 1.39167188478397e-05, "loss": 0.5543, "step": 13300, "task_loss": 1.1302056312561035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31177374720573425, "epoch": 11.24, "learning_rate": 1.3913587977457735e-05, "loss": 0.382, "step": 13301, "task_loss": 0.3063073754310608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5067688822746277, "epoch": 11.24, "learning_rate": 1.3910457107075767e-05, "loss": 0.5184, "step": 13302, "task_loss": 0.8469496369361877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6522015929222107, "epoch": 11.24, "learning_rate": 1.3907326236693802e-05, "loss": 0.5202, "step": 13303, "task_loss": 1.3286445140838623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5628464221954346, "epoch": 11.25, "learning_rate": 1.3904195366311834e-05, "loss": 0.5205, "step": 13304, "task_loss": 0.6996123790740967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9741288423538208, "epoch": 11.25, "learning_rate": 1.3901064495929869e-05, "loss": 0.6283, "step": 13305, "task_loss": 0.758793830871582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45977067947387695, "epoch": 11.25, "learning_rate": 1.3897933625547904e-05, "loss": 0.4698, "step": 13306, "task_loss": 0.5161727070808411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3759758472442627, "epoch": 11.25, "learning_rate": 1.3894802755165936e-05, "loss": 0.4427, "step": 13307, "task_loss": 0.31706011295318604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38641953468322754, "epoch": 11.25, "learning_rate": 1.3891671884783971e-05, "loss": 0.5185, "step": 13308, "task_loss": 0.3599022328853607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.504600465297699, "epoch": 11.25, "learning_rate": 1.3888541014402004e-05, "loss": 0.4927, "step": 13309, "task_loss": 0.30941009521484375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36224132776260376, "epoch": 11.25, "learning_rate": 1.388541014402004e-05, "loss": 0.5525, "step": 13310, "task_loss": 0.05257333815097809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29038944840431213, "epoch": 11.25, "learning_rate": 1.3882279273638071e-05, "loss": 0.4391, "step": 13311, "task_loss": 0.06849151104688644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45091572403907776, "epoch": 11.25, "learning_rate": 1.3879148403256107e-05, "loss": 0.4464, "step": 13312, "task_loss": 0.43930551409721375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.652741551399231, "epoch": 11.25, "learning_rate": 1.3876017532874138e-05, "loss": 0.464, "step": 13313, "task_loss": 0.4403991997241974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2343963086605072, "epoch": 11.25, "learning_rate": 1.3872886662492173e-05, "loss": 0.3666, "step": 13314, "task_loss": 0.3988073766231537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49259334802627563, "epoch": 11.26, "learning_rate": 1.3869755792110207e-05, "loss": 0.3985, "step": 13315, "task_loss": 0.8962003588676453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7120488882064819, "epoch": 11.26, "learning_rate": 1.386662492172824e-05, "loss": 0.5916, "step": 13316, "task_loss": 1.275599718093872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5264216661453247, "epoch": 11.26, "learning_rate": 1.3863494051346274e-05, "loss": 0.5753, "step": 13317, "task_loss": 0.6812021136283875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5454494953155518, "epoch": 11.26, "learning_rate": 1.3860363180964309e-05, "loss": 0.4361, "step": 13318, "task_loss": 0.7324299812316895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6395057439804077, "epoch": 11.26, "learning_rate": 1.3857232310582344e-05, "loss": 0.523, "step": 13319, "task_loss": 1.1497973203659058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5345066785812378, "epoch": 11.26, "learning_rate": 1.3854101440200376e-05, "loss": 0.4844, "step": 13320, "task_loss": 2.3618900775909424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2876392602920532, "epoch": 11.26, "learning_rate": 1.3850970569818411e-05, "loss": 0.4538, "step": 13321, "task_loss": 0.7020166516304016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47661080956459045, "epoch": 11.26, "learning_rate": 1.3847839699436443e-05, "loss": 0.4991, "step": 13322, "task_loss": 0.46099716424942017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.474174827337265, "epoch": 11.26, "learning_rate": 1.3844708829054478e-05, "loss": 0.4758, "step": 13323, "task_loss": 0.5283194184303284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4762096107006073, "epoch": 11.26, "learning_rate": 1.3841577958672512e-05, "loss": 0.3867, "step": 13324, "task_loss": 0.6484405994415283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4052409529685974, "epoch": 11.26, "learning_rate": 1.3838447088290547e-05, "loss": 0.3284, "step": 13325, "task_loss": 0.5821306109428406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6057902574539185, "epoch": 11.26, "learning_rate": 1.3835316217908579e-05, "loss": 0.5493, "step": 13326, "task_loss": 1.297049641609192 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3911043107509613, "epoch": 11.27, "learning_rate": 1.3832185347526614e-05, "loss": 0.5168, "step": 13327, "task_loss": 0.3837048411369324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.521282434463501, "epoch": 11.27, "learning_rate": 1.3829054477144645e-05, "loss": 0.5438, "step": 13328, "task_loss": 0.347763329744339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4856293499469757, "epoch": 11.27, "learning_rate": 1.382592360676268e-05, "loss": 0.4673, "step": 13329, "task_loss": 0.842688262462616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5539947748184204, "epoch": 11.27, "learning_rate": 1.3822792736380716e-05, "loss": 0.6081, "step": 13330, "task_loss": 1.4849334955215454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48663508892059326, "epoch": 11.27, "learning_rate": 1.3819661865998748e-05, "loss": 0.4223, "step": 13331, "task_loss": 0.5939164757728577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5013359785079956, "epoch": 11.27, "learning_rate": 1.3816530995616783e-05, "loss": 0.5087, "step": 13332, "task_loss": 0.7028133869171143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7055884003639221, "epoch": 11.27, "learning_rate": 1.3813400125234816e-05, "loss": 0.6429, "step": 13333, "task_loss": 0.32000863552093506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5064519643783569, "epoch": 11.27, "learning_rate": 1.3810269254852851e-05, "loss": 0.5267, "step": 13334, "task_loss": 1.181858777999878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8874207139015198, "epoch": 11.27, "learning_rate": 1.3807138384470883e-05, "loss": 0.606, "step": 13335, "task_loss": 0.8572090268135071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.390458345413208, "epoch": 11.27, "learning_rate": 1.3804007514088918e-05, "loss": 0.4712, "step": 13336, "task_loss": 0.3224963843822479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3175746202468872, "epoch": 11.27, "learning_rate": 1.380087664370695e-05, "loss": 0.4805, "step": 13337, "task_loss": 1.0958672761917114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45938149094581604, "epoch": 11.27, "learning_rate": 1.3797745773324985e-05, "loss": 0.4665, "step": 13338, "task_loss": 0.24571096897125244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2977871894836426, "epoch": 11.28, "learning_rate": 1.3794614902943017e-05, "loss": 0.4943, "step": 13339, "task_loss": 0.3642612397670746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30152615904808044, "epoch": 11.28, "learning_rate": 1.3791484032561052e-05, "loss": 0.3167, "step": 13340, "task_loss": 0.4939119517803192 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4273010492324829, "epoch": 11.28, "learning_rate": 1.3788353162179086e-05, "loss": 0.3992, "step": 13341, "task_loss": 0.4114125370979309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33184847235679626, "epoch": 11.28, "learning_rate": 1.3785222291797121e-05, "loss": 0.4232, "step": 13342, "task_loss": 0.6853185296058655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40984219312667847, "epoch": 11.28, "learning_rate": 1.3782091421415156e-05, "loss": 0.4371, "step": 13343, "task_loss": 0.2970816195011139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5458592176437378, "epoch": 11.28, "learning_rate": 1.3778960551033188e-05, "loss": 0.471, "step": 13344, "task_loss": 0.6990772485733032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49724289774894714, "epoch": 11.28, "learning_rate": 1.3775829680651223e-05, "loss": 0.4138, "step": 13345, "task_loss": 0.7524357438087463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3764515221118927, "epoch": 11.28, "learning_rate": 1.3772698810269255e-05, "loss": 0.4156, "step": 13346, "task_loss": 0.16859135031700134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4897491931915283, "epoch": 11.28, "learning_rate": 1.376956793988729e-05, "loss": 0.4485, "step": 13347, "task_loss": 0.5890968441963196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2898651957511902, "epoch": 11.28, "learning_rate": 1.3766437069505322e-05, "loss": 0.4772, "step": 13348, "task_loss": 0.26680994033813477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.667726993560791, "epoch": 11.28, "learning_rate": 1.3763306199123357e-05, "loss": 0.5852, "step": 13349, "task_loss": 1.096417784690857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5777534246444702, "epoch": 11.28, "learning_rate": 1.376017532874139e-05, "loss": 0.5432, "step": 13350, "task_loss": 0.8183897733688354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4297376275062561, "epoch": 11.29, "learning_rate": 1.3757044458359426e-05, "loss": 0.3974, "step": 13351, "task_loss": 0.4582257866859436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.495836466550827, "epoch": 11.29, "learning_rate": 1.3753913587977457e-05, "loss": 0.4591, "step": 13352, "task_loss": 1.2667330503463745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5677151679992676, "epoch": 11.29, "learning_rate": 1.3750782717595493e-05, "loss": 0.5127, "step": 13353, "task_loss": 0.6865700483322144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5268344283103943, "epoch": 11.29, "learning_rate": 1.3747651847213528e-05, "loss": 0.5592, "step": 13354, "task_loss": 1.17003333568573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48578524589538574, "epoch": 11.29, "learning_rate": 1.374452097683156e-05, "loss": 0.4526, "step": 13355, "task_loss": 0.07372177392244339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5387701988220215, "epoch": 11.29, "learning_rate": 1.3741390106449595e-05, "loss": 0.4289, "step": 13356, "task_loss": 0.8387320041656494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5043855309486389, "epoch": 11.29, "learning_rate": 1.3738259236067626e-05, "loss": 0.323, "step": 13357, "task_loss": 0.23875297605991364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5808480978012085, "epoch": 11.29, "learning_rate": 1.3735128365685662e-05, "loss": 0.5059, "step": 13358, "task_loss": 0.5326516032218933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41285234689712524, "epoch": 11.29, "learning_rate": 1.3731997495303695e-05, "loss": 0.4043, "step": 13359, "task_loss": 0.5473989844322205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6750186681747437, "epoch": 11.29, "learning_rate": 1.372886662492173e-05, "loss": 0.6351, "step": 13360, "task_loss": 0.36334073543548584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3663714826107025, "epoch": 11.29, "learning_rate": 1.3725735754539762e-05, "loss": 0.3081, "step": 13361, "task_loss": 0.971801221370697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45355087518692017, "epoch": 11.29, "learning_rate": 1.3722604884157797e-05, "loss": 0.6565, "step": 13362, "task_loss": 1.2883436679840088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3375697135925293, "epoch": 11.3, "learning_rate": 1.3719474013775829e-05, "loss": 0.3749, "step": 13363, "task_loss": 0.058856990188360214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5056964159011841, "epoch": 11.3, "learning_rate": 1.3716343143393864e-05, "loss": 0.4517, "step": 13364, "task_loss": 1.0236470699310303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5667839646339417, "epoch": 11.3, "learning_rate": 1.3713212273011896e-05, "loss": 0.5362, "step": 13365, "task_loss": 0.7686445713043213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25451910495758057, "epoch": 11.3, "learning_rate": 1.3710081402629931e-05, "loss": 0.546, "step": 13366, "task_loss": 0.6259620189666748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33604294061660767, "epoch": 11.3, "learning_rate": 1.3706950532247966e-05, "loss": 0.5896, "step": 13367, "task_loss": 0.7130985260009766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42202168703079224, "epoch": 11.3, "learning_rate": 1.3703819661866e-05, "loss": 0.5629, "step": 13368, "task_loss": 0.7841446399688721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6884379386901855, "epoch": 11.3, "learning_rate": 1.3700688791484035e-05, "loss": 0.528, "step": 13369, "task_loss": 1.361886978149414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4109395146369934, "epoch": 11.3, "learning_rate": 1.3697557921102067e-05, "loss": 0.6591, "step": 13370, "task_loss": 0.6874579191207886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5568269491195679, "epoch": 11.3, "learning_rate": 1.3694427050720102e-05, "loss": 0.5019, "step": 13371, "task_loss": 0.6558979749679565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44843512773513794, "epoch": 11.3, "learning_rate": 1.3691296180338134e-05, "loss": 0.4707, "step": 13372, "task_loss": 0.28383398056030273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45432907342910767, "epoch": 11.3, "learning_rate": 1.3688165309956169e-05, "loss": 0.4567, "step": 13373, "task_loss": 1.2079734802246094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.386044442653656, "epoch": 11.3, "learning_rate": 1.36850344395742e-05, "loss": 0.4573, "step": 13374, "task_loss": 0.054436925798654556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3672822415828705, "epoch": 11.31, "learning_rate": 1.3681903569192236e-05, "loss": 0.5553, "step": 13375, "task_loss": 0.5753524303436279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2382594645023346, "epoch": 11.31, "learning_rate": 1.367877269881027e-05, "loss": 0.4945, "step": 13376, "task_loss": 0.4129157066345215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6497118473052979, "epoch": 11.31, "learning_rate": 1.3675641828428304e-05, "loss": 0.5116, "step": 13377, "task_loss": 0.9274224042892456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41200608015060425, "epoch": 11.31, "learning_rate": 1.3672510958046336e-05, "loss": 0.4536, "step": 13378, "task_loss": 1.176707148551941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4407302439212799, "epoch": 11.31, "learning_rate": 1.3669380087664371e-05, "loss": 0.6338, "step": 13379, "task_loss": 0.06862740218639374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3836440443992615, "epoch": 11.31, "learning_rate": 1.3666249217282407e-05, "loss": 0.4164, "step": 13380, "task_loss": 0.41880226135253906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34630173444747925, "epoch": 11.31, "learning_rate": 1.3663118346900438e-05, "loss": 0.4403, "step": 13381, "task_loss": 0.08816014975309372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3904734253883362, "epoch": 11.31, "learning_rate": 1.3659987476518473e-05, "loss": 0.4016, "step": 13382, "task_loss": 0.3079637885093689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5237665772438049, "epoch": 11.31, "learning_rate": 1.3656856606136505e-05, "loss": 0.5121, "step": 13383, "task_loss": 0.2636469304561615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4129066467285156, "epoch": 11.31, "learning_rate": 1.365372573575454e-05, "loss": 0.3599, "step": 13384, "task_loss": 0.4263373613357544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6301016211509705, "epoch": 11.31, "learning_rate": 1.3650594865372574e-05, "loss": 0.4466, "step": 13385, "task_loss": 0.9277878403663635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2782844603061676, "epoch": 11.32, "learning_rate": 1.3647463994990609e-05, "loss": 0.4931, "step": 13386, "task_loss": 0.09378264099359512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25975513458251953, "epoch": 11.32, "learning_rate": 1.3644333124608641e-05, "loss": 0.4915, "step": 13387, "task_loss": 0.3365277051925659 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.586678147315979, "epoch": 11.32, "learning_rate": 1.3641202254226676e-05, "loss": 0.4775, "step": 13388, "task_loss": 0.45373642444610596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4524852931499481, "epoch": 11.32, "learning_rate": 1.3638071383844708e-05, "loss": 0.4756, "step": 13389, "task_loss": 0.5014271140098572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46272388100624084, "epoch": 11.32, "learning_rate": 1.3634940513462743e-05, "loss": 0.4441, "step": 13390, "task_loss": 0.12681461870670319 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4037438631057739, "epoch": 11.32, "learning_rate": 1.3631809643080778e-05, "loss": 0.3733, "step": 13391, "task_loss": 0.3803998529911041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41954532265663147, "epoch": 11.32, "learning_rate": 1.362867877269881e-05, "loss": 0.5659, "step": 13392, "task_loss": 0.6400060057640076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5114694237709045, "epoch": 11.32, "learning_rate": 1.3625547902316845e-05, "loss": 0.5769, "step": 13393, "task_loss": 0.8380044102668762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49967673420906067, "epoch": 11.32, "learning_rate": 1.3622417031934879e-05, "loss": 0.5613, "step": 13394, "task_loss": 0.3630809187889099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5169315338134766, "epoch": 11.32, "learning_rate": 1.3619286161552914e-05, "loss": 0.5838, "step": 13395, "task_loss": 0.6414481997489929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44580867886543274, "epoch": 11.32, "learning_rate": 1.3616155291170946e-05, "loss": 0.566, "step": 13396, "task_loss": 0.6656661629676819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23438279330730438, "epoch": 11.32, "learning_rate": 1.361302442078898e-05, "loss": 0.4967, "step": 13397, "task_loss": 0.44450104236602783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4562116861343384, "epoch": 11.33, "learning_rate": 1.3609893550407012e-05, "loss": 0.3785, "step": 13398, "task_loss": 0.2262602299451828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4760447144508362, "epoch": 11.33, "learning_rate": 1.3606762680025048e-05, "loss": 0.5058, "step": 13399, "task_loss": 1.25832200050354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3436952829360962, "epoch": 11.33, "learning_rate": 1.3603631809643081e-05, "loss": 0.4664, "step": 13400, "task_loss": 0.5014289617538452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4205661714076996, "epoch": 11.33, "learning_rate": 1.3600500939261116e-05, "loss": 0.4209, "step": 13401, "task_loss": 1.2450330257415771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7762722969055176, "epoch": 11.33, "learning_rate": 1.3597370068879148e-05, "loss": 0.4344, "step": 13402, "task_loss": 0.8572439551353455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6550551652908325, "epoch": 11.33, "learning_rate": 1.3594239198497183e-05, "loss": 0.4542, "step": 13403, "task_loss": 0.6343236565589905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3930063843727112, "epoch": 11.33, "learning_rate": 1.3591108328115218e-05, "loss": 0.4188, "step": 13404, "task_loss": 0.3986065089702606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.433918297290802, "epoch": 11.33, "learning_rate": 1.358797745773325e-05, "loss": 0.3191, "step": 13405, "task_loss": 0.4140155613422394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6488302946090698, "epoch": 11.33, "learning_rate": 1.3584846587351285e-05, "loss": 0.562, "step": 13406, "task_loss": 0.19816847145557404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48100948333740234, "epoch": 11.33, "learning_rate": 1.3581715716969317e-05, "loss": 0.4733, "step": 13407, "task_loss": 0.38867056369781494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24857836961746216, "epoch": 11.33, "learning_rate": 1.3578584846587352e-05, "loss": 0.398, "step": 13408, "task_loss": 0.24509504437446594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3802242577075958, "epoch": 11.33, "learning_rate": 1.3575453976205386e-05, "loss": 0.3496, "step": 13409, "task_loss": 0.3202233612537384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47925683856010437, "epoch": 11.34, "learning_rate": 1.3572323105823421e-05, "loss": 0.5076, "step": 13410, "task_loss": 0.13921737670898438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5100857019424438, "epoch": 11.34, "learning_rate": 1.3569192235441453e-05, "loss": 0.5024, "step": 13411, "task_loss": 0.8136407136917114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4130246341228485, "epoch": 11.34, "learning_rate": 1.3566061365059488e-05, "loss": 0.3474, "step": 13412, "task_loss": 1.0072972774505615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3904590606689453, "epoch": 11.34, "learning_rate": 1.356293049467752e-05, "loss": 0.3893, "step": 13413, "task_loss": 0.5007895827293396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3351680636405945, "epoch": 11.34, "learning_rate": 1.3559799624295555e-05, "loss": 0.5122, "step": 13414, "task_loss": 0.6062806844711304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.875203013420105, "epoch": 11.34, "learning_rate": 1.3556668753913587e-05, "loss": 0.5877, "step": 13415, "task_loss": 0.9651710391044617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4931524097919464, "epoch": 11.34, "learning_rate": 1.3553537883531622e-05, "loss": 0.4047, "step": 13416, "task_loss": 1.030369758605957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5621304512023926, "epoch": 11.34, "learning_rate": 1.3550407013149657e-05, "loss": 0.4601, "step": 13417, "task_loss": 0.8106245994567871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4540729522705078, "epoch": 11.34, "learning_rate": 1.354727614276769e-05, "loss": 0.4557, "step": 13418, "task_loss": 0.7694531679153442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31840750575065613, "epoch": 11.34, "learning_rate": 1.3544145272385726e-05, "loss": 0.4614, "step": 13419, "task_loss": 0.2681237757205963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6150997877120972, "epoch": 11.34, "learning_rate": 1.3541014402003757e-05, "loss": 0.5291, "step": 13420, "task_loss": 0.49860239028930664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2548239827156067, "epoch": 11.34, "learning_rate": 1.3537883531621793e-05, "loss": 0.5047, "step": 13421, "task_loss": 0.8994212746620178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3094787895679474, "epoch": 11.35, "learning_rate": 1.3534752661239824e-05, "loss": 0.4736, "step": 13422, "task_loss": 0.7753906846046448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4576607942581177, "epoch": 11.35, "learning_rate": 1.353162179085786e-05, "loss": 0.509, "step": 13423, "task_loss": 0.8873574137687683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5469491481781006, "epoch": 11.35, "learning_rate": 1.3528490920475891e-05, "loss": 0.598, "step": 13424, "task_loss": 1.108163595199585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4785671830177307, "epoch": 11.35, "learning_rate": 1.3525360050093926e-05, "loss": 0.4861, "step": 13425, "task_loss": 0.4308520257472992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.154972791671753, "epoch": 11.35, "learning_rate": 1.352222917971196e-05, "loss": 0.6499, "step": 13426, "task_loss": 0.5532777309417725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34761595726013184, "epoch": 11.35, "learning_rate": 1.3519098309329995e-05, "loss": 0.468, "step": 13427, "task_loss": 0.8235142827033997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2776043117046356, "epoch": 11.35, "learning_rate": 1.351596743894803e-05, "loss": 0.3995, "step": 13428, "task_loss": 0.6162019371986389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43348342180252075, "epoch": 11.35, "learning_rate": 1.3512836568566062e-05, "loss": 0.5181, "step": 13429, "task_loss": 0.501488983631134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36238986253738403, "epoch": 11.35, "learning_rate": 1.3509705698184097e-05, "loss": 0.4256, "step": 13430, "task_loss": 0.7634019255638123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23057928681373596, "epoch": 11.35, "learning_rate": 1.3506574827802129e-05, "loss": 0.401, "step": 13431, "task_loss": 0.4590986371040344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2874537706375122, "epoch": 11.35, "learning_rate": 1.3503443957420164e-05, "loss": 0.3712, "step": 13432, "task_loss": 0.5492112636566162 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3312625586986542, "epoch": 11.35, "learning_rate": 1.3500313087038196e-05, "loss": 0.4188, "step": 13433, "task_loss": 0.11307858675718307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9354593753814697, "epoch": 11.36, "learning_rate": 1.3497182216656231e-05, "loss": 0.6233, "step": 13434, "task_loss": 1.5848944187164307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43516606092453003, "epoch": 11.36, "learning_rate": 1.3494051346274265e-05, "loss": 0.4265, "step": 13435, "task_loss": 0.3258773684501648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3711320161819458, "epoch": 11.36, "learning_rate": 1.34909204758923e-05, "loss": 0.3255, "step": 13436, "task_loss": 0.12808597087860107 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.776048481464386, "epoch": 11.36, "learning_rate": 1.3487789605510332e-05, "loss": 0.5103, "step": 13437, "task_loss": 0.7046303749084473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46114909648895264, "epoch": 11.36, "learning_rate": 1.3484658735128367e-05, "loss": 0.4028, "step": 13438, "task_loss": 0.45602840185165405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4125756621360779, "epoch": 11.36, "learning_rate": 1.3481527864746399e-05, "loss": 0.48, "step": 13439, "task_loss": 0.6102780699729919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47756028175354004, "epoch": 11.36, "learning_rate": 1.3478396994364434e-05, "loss": 0.5404, "step": 13440, "task_loss": 0.6832168102264404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4421721398830414, "epoch": 11.36, "learning_rate": 1.3475266123982469e-05, "loss": 0.4626, "step": 13441, "task_loss": 1.069075107574463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6344483494758606, "epoch": 11.36, "learning_rate": 1.34721352536005e-05, "loss": 0.5835, "step": 13442, "task_loss": 1.025761365890503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5571324825286865, "epoch": 11.36, "learning_rate": 1.3469004383218536e-05, "loss": 0.5733, "step": 13443, "task_loss": 0.9477716088294983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5051746368408203, "epoch": 11.36, "learning_rate": 1.346587351283657e-05, "loss": 0.5465, "step": 13444, "task_loss": 0.37902286648750305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29703861474990845, "epoch": 11.36, "learning_rate": 1.3462742642454604e-05, "loss": 0.5373, "step": 13445, "task_loss": 0.44304999709129333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5473580360412598, "epoch": 11.37, "learning_rate": 1.3459611772072636e-05, "loss": 0.5305, "step": 13446, "task_loss": 1.0028460025787354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4304829239845276, "epoch": 11.37, "learning_rate": 1.3456480901690671e-05, "loss": 0.4717, "step": 13447, "task_loss": 0.5336629152297974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6534134149551392, "epoch": 11.37, "learning_rate": 1.3453350031308703e-05, "loss": 0.6508, "step": 13448, "task_loss": 0.26125165820121765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6263030767440796, "epoch": 11.37, "learning_rate": 1.3450219160926738e-05, "loss": 0.5739, "step": 13449, "task_loss": 0.9009964466094971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6634045839309692, "epoch": 11.37, "learning_rate": 1.344708829054477e-05, "loss": 0.4708, "step": 13450, "task_loss": 0.7667508721351624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2528965175151825, "epoch": 11.37, "learning_rate": 1.3443957420162805e-05, "loss": 0.4041, "step": 13451, "task_loss": 0.44988200068473816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.403778612613678, "epoch": 11.37, "learning_rate": 1.344082654978084e-05, "loss": 0.4387, "step": 13452, "task_loss": 0.7008814215660095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6918982267379761, "epoch": 11.37, "learning_rate": 1.3437695679398874e-05, "loss": 0.5381, "step": 13453, "task_loss": 0.9185293912887573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30877506732940674, "epoch": 11.37, "learning_rate": 1.3434564809016909e-05, "loss": 0.5913, "step": 13454, "task_loss": 0.3754591941833496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3243224322795868, "epoch": 11.37, "learning_rate": 1.3431433938634941e-05, "loss": 0.5074, "step": 13455, "task_loss": 0.6020894050598145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3592853248119354, "epoch": 11.37, "learning_rate": 1.3428303068252976e-05, "loss": 0.3911, "step": 13456, "task_loss": 1.2577414512634277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39873141050338745, "epoch": 11.38, "learning_rate": 1.3425172197871008e-05, "loss": 0.4945, "step": 13457, "task_loss": 0.5454912185668945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32221081852912903, "epoch": 11.38, "learning_rate": 1.3422041327489043e-05, "loss": 0.4007, "step": 13458, "task_loss": 0.5369566679000854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35904696583747864, "epoch": 11.38, "learning_rate": 1.3418910457107075e-05, "loss": 0.3637, "step": 13459, "task_loss": 0.23569725453853607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30779266357421875, "epoch": 11.38, "learning_rate": 1.341577958672511e-05, "loss": 0.4442, "step": 13460, "task_loss": 0.20098252594470978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5619673132896423, "epoch": 11.38, "learning_rate": 1.3412648716343143e-05, "loss": 0.3831, "step": 13461, "task_loss": 0.455183744430542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3579810857772827, "epoch": 11.38, "learning_rate": 1.3409517845961179e-05, "loss": 0.5269, "step": 13462, "task_loss": 0.7468730807304382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6393323540687561, "epoch": 11.38, "learning_rate": 1.340638697557921e-05, "loss": 0.5225, "step": 13463, "task_loss": 0.43784138560295105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48542821407318115, "epoch": 11.38, "learning_rate": 1.3403256105197246e-05, "loss": 0.3878, "step": 13464, "task_loss": 0.3089962303638458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3112417459487915, "epoch": 11.38, "learning_rate": 1.340012523481528e-05, "loss": 0.5684, "step": 13465, "task_loss": 0.9153579473495483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4815816581249237, "epoch": 11.38, "learning_rate": 1.3396994364433313e-05, "loss": 0.5301, "step": 13466, "task_loss": 0.6106700301170349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30273866653442383, "epoch": 11.38, "learning_rate": 1.3393863494051348e-05, "loss": 0.4597, "step": 13467, "task_loss": 0.1974809169769287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.224589541554451, "epoch": 11.38, "learning_rate": 1.339073262366938e-05, "loss": 0.5102, "step": 13468, "task_loss": 0.06221473962068558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3997344970703125, "epoch": 11.39, "learning_rate": 1.3387601753287415e-05, "loss": 0.4984, "step": 13469, "task_loss": 0.8279926776885986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.422261118888855, "epoch": 11.39, "learning_rate": 1.3384470882905448e-05, "loss": 0.6218, "step": 13470, "task_loss": 0.863456130027771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40642112493515015, "epoch": 11.39, "learning_rate": 1.3381340012523483e-05, "loss": 0.4919, "step": 13471, "task_loss": 0.9533356428146362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5542224645614624, "epoch": 11.39, "learning_rate": 1.3378209142141515e-05, "loss": 0.6186, "step": 13472, "task_loss": 0.3777555823326111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3315945267677307, "epoch": 11.39, "learning_rate": 1.337507827175955e-05, "loss": 0.3905, "step": 13473, "task_loss": 0.3259689211845398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42900097370147705, "epoch": 11.39, "learning_rate": 1.3371947401377582e-05, "loss": 0.4802, "step": 13474, "task_loss": 2.150320291519165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3709228038787842, "epoch": 11.39, "learning_rate": 1.3368816530995617e-05, "loss": 0.5067, "step": 13475, "task_loss": 0.1489427387714386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35552260279655457, "epoch": 11.39, "learning_rate": 1.336568566061365e-05, "loss": 0.4802, "step": 13476, "task_loss": 1.1448328495025635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7506089806556702, "epoch": 11.39, "learning_rate": 1.3362554790231686e-05, "loss": 0.5929, "step": 13477, "task_loss": 0.6513552069664001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3635621666908264, "epoch": 11.39, "learning_rate": 1.335942391984972e-05, "loss": 0.4563, "step": 13478, "task_loss": 0.3204783499240875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5291178226470947, "epoch": 11.39, "learning_rate": 1.3356293049467753e-05, "loss": 0.4817, "step": 13479, "task_loss": 0.8866700530052185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43084555864334106, "epoch": 11.39, "learning_rate": 1.3353162179085788e-05, "loss": 0.5122, "step": 13480, "task_loss": 0.5674587488174438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3933725357055664, "epoch": 11.4, "learning_rate": 1.335003130870382e-05, "loss": 0.3252, "step": 13481, "task_loss": 0.025790400803089142 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5571322441101074, "epoch": 11.4, "learning_rate": 1.3346900438321855e-05, "loss": 0.5479, "step": 13482, "task_loss": 0.30859389901161194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2971702516078949, "epoch": 11.4, "learning_rate": 1.3343769567939887e-05, "loss": 0.3947, "step": 13483, "task_loss": 0.199801504611969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7566019296646118, "epoch": 11.4, "learning_rate": 1.3340638697557922e-05, "loss": 0.627, "step": 13484, "task_loss": 1.0852937698364258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5885090827941895, "epoch": 11.4, "learning_rate": 1.3337507827175955e-05, "loss": 0.5946, "step": 13485, "task_loss": 1.230541467666626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5352647304534912, "epoch": 11.4, "learning_rate": 1.333437695679399e-05, "loss": 0.4455, "step": 13486, "task_loss": 0.6573863625526428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44642752408981323, "epoch": 11.4, "learning_rate": 1.3331246086412022e-05, "loss": 0.5641, "step": 13487, "task_loss": 0.7030841112136841 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7347464561462402, "epoch": 11.4, "learning_rate": 1.3328115216030057e-05, "loss": 0.5299, "step": 13488, "task_loss": 0.9467954039573669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6187528967857361, "epoch": 11.4, "learning_rate": 1.3324984345648093e-05, "loss": 0.6187, "step": 13489, "task_loss": 1.266119122505188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38012802600860596, "epoch": 11.4, "learning_rate": 1.3321853475266124e-05, "loss": 0.51, "step": 13490, "task_loss": 0.22414803504943848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4983089566230774, "epoch": 11.4, "learning_rate": 1.331872260488416e-05, "loss": 0.3629, "step": 13491, "task_loss": 0.441162109375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42119261622428894, "epoch": 11.4, "learning_rate": 1.3315591734502191e-05, "loss": 0.4164, "step": 13492, "task_loss": 0.47416454553604126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5530366897583008, "epoch": 11.41, "learning_rate": 1.3312460864120227e-05, "loss": 0.5664, "step": 13493, "task_loss": 1.8995047807693481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6007485389709473, "epoch": 11.41, "learning_rate": 1.330932999373826e-05, "loss": 0.5588, "step": 13494, "task_loss": 1.316316843032837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43751686811447144, "epoch": 11.41, "learning_rate": 1.3306199123356295e-05, "loss": 0.5359, "step": 13495, "task_loss": 0.9713654518127441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6639513373374939, "epoch": 11.41, "learning_rate": 1.3303068252974327e-05, "loss": 0.4878, "step": 13496, "task_loss": 0.27489954233169556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38113003969192505, "epoch": 11.41, "learning_rate": 1.3299937382592362e-05, "loss": 0.474, "step": 13497, "task_loss": 0.38937392830848694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4935281276702881, "epoch": 11.41, "learning_rate": 1.3296806512210394e-05, "loss": 0.3707, "step": 13498, "task_loss": 0.6612992882728577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2878161668777466, "epoch": 11.41, "learning_rate": 1.3293675641828429e-05, "loss": 0.3924, "step": 13499, "task_loss": 0.27790069580078125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5652097463607788, "epoch": 11.41, "learning_rate": 1.329054477144646e-05, "loss": 0.623, "step": 13500, "task_loss": 0.3563183844089508 }, { "epoch": 11.41, "eval_accuracy": 0.9087524752475248, "eval_loss": 0.3352530300617218, "eval_runtime": 208.0406, "eval_samples_per_second": 121.371, "eval_steps_per_second": 0.952, "step": 13500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.465501606464386, "epoch": 11.41, "learning_rate": 1.3287413901064496e-05, "loss": 0.3948, "step": 13501, "task_loss": 0.735026478767395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37471097707748413, "epoch": 11.41, "learning_rate": 1.3284283030682531e-05, "loss": 0.4481, "step": 13502, "task_loss": 0.5957316756248474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6463758945465088, "epoch": 11.41, "learning_rate": 1.3281152160300565e-05, "loss": 0.6281, "step": 13503, "task_loss": 0.8868898153305054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46221762895584106, "epoch": 11.41, "learning_rate": 1.32780212899186e-05, "loss": 0.5087, "step": 13504, "task_loss": 0.6582521796226501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5101259350776672, "epoch": 11.42, "learning_rate": 1.3274890419536632e-05, "loss": 0.5422, "step": 13505, "task_loss": 0.8608101606369019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36429154872894287, "epoch": 11.42, "learning_rate": 1.3271759549154667e-05, "loss": 0.405, "step": 13506, "task_loss": 0.5104748010635376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7274377942085266, "epoch": 11.42, "learning_rate": 1.3268628678772699e-05, "loss": 0.6735, "step": 13507, "task_loss": 0.6605238318443298 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4454156458377838, "epoch": 11.42, "learning_rate": 1.3265497808390734e-05, "loss": 0.4043, "step": 13508, "task_loss": 0.737153172492981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3729313313961029, "epoch": 11.42, "learning_rate": 1.3262366938008766e-05, "loss": 0.4234, "step": 13509, "task_loss": 1.3114820718765259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3347185254096985, "epoch": 11.42, "learning_rate": 1.32592360676268e-05, "loss": 0.4625, "step": 13510, "task_loss": 0.8003670573234558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38057124614715576, "epoch": 11.42, "learning_rate": 1.3256105197244834e-05, "loss": 0.4816, "step": 13511, "task_loss": 0.612835168838501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3276286721229553, "epoch": 11.42, "learning_rate": 1.325297432686287e-05, "loss": 0.445, "step": 13512, "task_loss": 0.9936932921409607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40744730830192566, "epoch": 11.42, "learning_rate": 1.3249843456480901e-05, "loss": 0.5174, "step": 13513, "task_loss": 0.2156963050365448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5491023063659668, "epoch": 11.42, "learning_rate": 1.3246712586098936e-05, "loss": 0.6021, "step": 13514, "task_loss": 0.7237211465835571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35684919357299805, "epoch": 11.42, "learning_rate": 1.3243581715716971e-05, "loss": 0.4388, "step": 13515, "task_loss": 0.7996036410331726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4663187265396118, "epoch": 11.42, "learning_rate": 1.3240450845335003e-05, "loss": 0.5326, "step": 13516, "task_loss": 0.32694414258003235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36058488488197327, "epoch": 11.43, "learning_rate": 1.3237319974953038e-05, "loss": 0.4577, "step": 13517, "task_loss": 0.24273529648780823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5685744881629944, "epoch": 11.43, "learning_rate": 1.323418910457107e-05, "loss": 0.5303, "step": 13518, "task_loss": 0.9413823485374451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4420734941959381, "epoch": 11.43, "learning_rate": 1.3231058234189105e-05, "loss": 0.4536, "step": 13519, "task_loss": 0.14917747676372528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5207750797271729, "epoch": 11.43, "learning_rate": 1.3227927363807139e-05, "loss": 0.6566, "step": 13520, "task_loss": 0.8525029420852661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42725151777267456, "epoch": 11.43, "learning_rate": 1.3224796493425174e-05, "loss": 0.4842, "step": 13521, "task_loss": 0.6993181109428406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4563949406147003, "epoch": 11.43, "learning_rate": 1.3221665623043206e-05, "loss": 0.4776, "step": 13522, "task_loss": 0.5038756728172302 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5740674734115601, "epoch": 11.43, "learning_rate": 1.3218534752661241e-05, "loss": 0.5002, "step": 13523, "task_loss": 0.6791039109230042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6099991202354431, "epoch": 11.43, "learning_rate": 1.3215403882279273e-05, "loss": 0.4947, "step": 13524, "task_loss": 1.1884582042694092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3031008541584015, "epoch": 11.43, "learning_rate": 1.3212273011897308e-05, "loss": 0.4484, "step": 13525, "task_loss": 0.35861116647720337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5669224858283997, "epoch": 11.43, "learning_rate": 1.3209142141515343e-05, "loss": 0.499, "step": 13526, "task_loss": 1.0844275951385498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42665889859199524, "epoch": 11.43, "learning_rate": 1.3206011271133375e-05, "loss": 0.5141, "step": 13527, "task_loss": 0.33678507804870605 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3605293333530426, "epoch": 11.44, "learning_rate": 1.320288040075141e-05, "loss": 0.3204, "step": 13528, "task_loss": 0.2133532464504242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20931023359298706, "epoch": 11.44, "learning_rate": 1.3199749530369443e-05, "loss": 0.375, "step": 13529, "task_loss": 0.17649134993553162 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4781498312950134, "epoch": 11.44, "learning_rate": 1.3196618659987479e-05, "loss": 0.4149, "step": 13530, "task_loss": 0.6655266880989075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6147508025169373, "epoch": 11.44, "learning_rate": 1.319348778960551e-05, "loss": 0.5159, "step": 13531, "task_loss": 1.0065929889678955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5388458371162415, "epoch": 11.44, "learning_rate": 1.3190356919223546e-05, "loss": 0.5411, "step": 13532, "task_loss": 1.1024378538131714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.507765531539917, "epoch": 11.44, "learning_rate": 1.3187226048841577e-05, "loss": 0.4855, "step": 13533, "task_loss": 0.2183615118265152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3862976133823395, "epoch": 11.44, "learning_rate": 1.3184095178459613e-05, "loss": 0.4725, "step": 13534, "task_loss": 0.4217413365840912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3520740568637848, "epoch": 11.44, "learning_rate": 1.3180964308077644e-05, "loss": 0.3447, "step": 13535, "task_loss": 1.0696911811828613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3324918746948242, "epoch": 11.44, "learning_rate": 1.317783343769568e-05, "loss": 0.4358, "step": 13536, "task_loss": 0.43615442514419556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25392070412635803, "epoch": 11.44, "learning_rate": 1.3174702567313713e-05, "loss": 0.356, "step": 13537, "task_loss": 0.2642957270145416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17624101042747498, "epoch": 11.44, "learning_rate": 1.3171571696931748e-05, "loss": 0.4015, "step": 13538, "task_loss": 0.41774827241897583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3969062566757202, "epoch": 11.44, "learning_rate": 1.3168440826549783e-05, "loss": 0.4895, "step": 13539, "task_loss": 0.9249807000160217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2826523780822754, "epoch": 11.45, "learning_rate": 1.3165309956167815e-05, "loss": 0.3924, "step": 13540, "task_loss": 0.14629510045051575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4402908682823181, "epoch": 11.45, "learning_rate": 1.316217908578585e-05, "loss": 0.3758, "step": 13541, "task_loss": 0.25744032859802246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43253272771835327, "epoch": 11.45, "learning_rate": 1.3159048215403882e-05, "loss": 0.4695, "step": 13542, "task_loss": 1.3856626749038696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3785671591758728, "epoch": 11.45, "learning_rate": 1.3155917345021917e-05, "loss": 0.5143, "step": 13543, "task_loss": 0.592303991317749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28685781359672546, "epoch": 11.45, "learning_rate": 1.3152786474639949e-05, "loss": 0.5797, "step": 13544, "task_loss": 0.827608048915863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3646191358566284, "epoch": 11.45, "learning_rate": 1.3149655604257984e-05, "loss": 0.4383, "step": 13545, "task_loss": 0.42719709873199463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28489774465560913, "epoch": 11.45, "learning_rate": 1.3146524733876018e-05, "loss": 0.3913, "step": 13546, "task_loss": 0.47962093353271484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44862571358680725, "epoch": 11.45, "learning_rate": 1.3143393863494053e-05, "loss": 0.5069, "step": 13547, "task_loss": 0.1419134885072708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2937842011451721, "epoch": 11.45, "learning_rate": 1.3140262993112085e-05, "loss": 0.3535, "step": 13548, "task_loss": 0.5965250134468079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46713411808013916, "epoch": 11.45, "learning_rate": 1.313713212273012e-05, "loss": 0.4039, "step": 13549, "task_loss": 0.8947498798370361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5573399662971497, "epoch": 11.45, "learning_rate": 1.3134001252348152e-05, "loss": 0.5768, "step": 13550, "task_loss": 0.5630077719688416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5078802108764648, "epoch": 11.45, "learning_rate": 1.3130870381966187e-05, "loss": 0.5171, "step": 13551, "task_loss": 1.4530473947525024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4557645618915558, "epoch": 11.46, "learning_rate": 1.3127739511584222e-05, "loss": 0.4041, "step": 13552, "task_loss": 1.0645562410354614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47988152503967285, "epoch": 11.46, "learning_rate": 1.3124608641202255e-05, "loss": 0.4747, "step": 13553, "task_loss": 0.7960699796676636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4651888608932495, "epoch": 11.46, "learning_rate": 1.3121477770820289e-05, "loss": 0.5712, "step": 13554, "task_loss": 0.8426004648208618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5218020081520081, "epoch": 11.46, "learning_rate": 1.3118346900438322e-05, "loss": 0.5296, "step": 13555, "task_loss": 0.6149376034736633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6785230040550232, "epoch": 11.46, "learning_rate": 1.3115216030056357e-05, "loss": 0.4903, "step": 13556, "task_loss": 0.12935608625411987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4795346260070801, "epoch": 11.46, "learning_rate": 1.311208515967439e-05, "loss": 0.5207, "step": 13557, "task_loss": 0.919486939907074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26108771562576294, "epoch": 11.46, "learning_rate": 1.3108954289292424e-05, "loss": 0.5319, "step": 13558, "task_loss": 0.5815039277076721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3480111360549927, "epoch": 11.46, "learning_rate": 1.3105823418910456e-05, "loss": 0.4311, "step": 13559, "task_loss": 0.16610586643218994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18104463815689087, "epoch": 11.46, "learning_rate": 1.3102692548528491e-05, "loss": 0.4606, "step": 13560, "task_loss": 0.5945863127708435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4410724639892578, "epoch": 11.46, "learning_rate": 1.3099561678146525e-05, "loss": 0.5078, "step": 13561, "task_loss": 1.0294389724731445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5213173627853394, "epoch": 11.46, "learning_rate": 1.309643080776456e-05, "loss": 0.4514, "step": 13562, "task_loss": 0.1676766574382782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6341860294342041, "epoch": 11.46, "learning_rate": 1.3093299937382594e-05, "loss": 0.566, "step": 13563, "task_loss": 0.8248197436332703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4407084286212921, "epoch": 11.47, "learning_rate": 1.3090169067000627e-05, "loss": 0.5074, "step": 13564, "task_loss": 0.48778414726257324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6474162936210632, "epoch": 11.47, "learning_rate": 1.3087038196618662e-05, "loss": 0.4211, "step": 13565, "task_loss": 0.4227956235408783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32033079862594604, "epoch": 11.47, "learning_rate": 1.3083907326236694e-05, "loss": 0.387, "step": 13566, "task_loss": 0.08172602951526642 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5786766409873962, "epoch": 11.47, "learning_rate": 1.3080776455854729e-05, "loss": 0.5708, "step": 13567, "task_loss": 0.6628842353820801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6221537590026855, "epoch": 11.47, "learning_rate": 1.3077645585472761e-05, "loss": 0.4559, "step": 13568, "task_loss": 1.3806887865066528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2118602693080902, "epoch": 11.47, "learning_rate": 1.3074514715090796e-05, "loss": 0.364, "step": 13569, "task_loss": 0.2774501442909241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4657360315322876, "epoch": 11.47, "learning_rate": 1.307138384470883e-05, "loss": 0.4222, "step": 13570, "task_loss": 0.15399114787578583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.523737907409668, "epoch": 11.47, "learning_rate": 1.3068252974326865e-05, "loss": 0.4095, "step": 13571, "task_loss": 0.7474629878997803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25396502017974854, "epoch": 11.47, "learning_rate": 1.3065122103944896e-05, "loss": 0.3383, "step": 13572, "task_loss": 0.3071964383125305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5895223021507263, "epoch": 11.47, "learning_rate": 1.3061991233562932e-05, "loss": 0.4618, "step": 13573, "task_loss": 0.9642338752746582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.391086220741272, "epoch": 11.47, "learning_rate": 1.3058860363180963e-05, "loss": 0.3479, "step": 13574, "task_loss": 0.9100220203399658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3184642195701599, "epoch": 11.47, "learning_rate": 1.3055729492798999e-05, "loss": 0.3797, "step": 13575, "task_loss": 0.16384708881378174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19684599339962006, "epoch": 11.48, "learning_rate": 1.3052598622417034e-05, "loss": 0.3656, "step": 13576, "task_loss": 0.8581432104110718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3291710317134857, "epoch": 11.48, "learning_rate": 1.3049467752035066e-05, "loss": 0.4087, "step": 13577, "task_loss": 0.8241137266159058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47324931621551514, "epoch": 11.48, "learning_rate": 1.30463368816531e-05, "loss": 0.5087, "step": 13578, "task_loss": 0.7043988704681396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3561636805534363, "epoch": 11.48, "learning_rate": 1.3043206011271134e-05, "loss": 0.4529, "step": 13579, "task_loss": 0.27962324023246765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38667720556259155, "epoch": 11.48, "learning_rate": 1.304007514088917e-05, "loss": 0.4106, "step": 13580, "task_loss": 0.540160596370697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39725959300994873, "epoch": 11.48, "learning_rate": 1.3036944270507201e-05, "loss": 0.5114, "step": 13581, "task_loss": 0.6200377345085144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29235363006591797, "epoch": 11.48, "learning_rate": 1.3033813400125236e-05, "loss": 0.439, "step": 13582, "task_loss": 0.37806904315948486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3412091135978699, "epoch": 11.48, "learning_rate": 1.3030682529743268e-05, "loss": 0.3533, "step": 13583, "task_loss": 0.7568220496177673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5143647789955139, "epoch": 11.48, "learning_rate": 1.3027551659361303e-05, "loss": 0.4522, "step": 13584, "task_loss": 0.43480974435806274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47022730112075806, "epoch": 11.48, "learning_rate": 1.3024420788979335e-05, "loss": 0.4187, "step": 13585, "task_loss": 0.30889371037483215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5907959938049316, "epoch": 11.48, "learning_rate": 1.302128991859737e-05, "loss": 0.5221, "step": 13586, "task_loss": 0.46049240231513977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8478836417198181, "epoch": 11.48, "learning_rate": 1.3018159048215405e-05, "loss": 0.5549, "step": 13587, "task_loss": 1.0599287748336792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7298989295959473, "epoch": 11.49, "learning_rate": 1.3015028177833439e-05, "loss": 0.5784, "step": 13588, "task_loss": 0.8381431102752686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18815864622592926, "epoch": 11.49, "learning_rate": 1.3011897307451474e-05, "loss": 0.4328, "step": 13589, "task_loss": 0.3429357409477234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5841088891029358, "epoch": 11.49, "learning_rate": 1.3008766437069506e-05, "loss": 0.6141, "step": 13590, "task_loss": 0.6905454993247986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36471372842788696, "epoch": 11.49, "learning_rate": 1.3005635566687541e-05, "loss": 0.4877, "step": 13591, "task_loss": 0.38101258873939514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5781923532485962, "epoch": 11.49, "learning_rate": 1.3002504696305573e-05, "loss": 0.5609, "step": 13592, "task_loss": 1.117819905281067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2870767116546631, "epoch": 11.49, "learning_rate": 1.2999373825923608e-05, "loss": 0.4224, "step": 13593, "task_loss": 0.8072639107704163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3485253155231476, "epoch": 11.49, "learning_rate": 1.299624295554164e-05, "loss": 0.4384, "step": 13594, "task_loss": 1.1574794054031372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30147305130958557, "epoch": 11.49, "learning_rate": 1.2993112085159675e-05, "loss": 0.3955, "step": 13595, "task_loss": 0.5711506605148315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3548358976840973, "epoch": 11.49, "learning_rate": 1.2989981214777708e-05, "loss": 0.3869, "step": 13596, "task_loss": 0.7341586351394653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4958457946777344, "epoch": 11.49, "learning_rate": 1.2986850344395744e-05, "loss": 0.4686, "step": 13597, "task_loss": 0.3273211121559143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45378240942955017, "epoch": 11.49, "learning_rate": 1.2983719474013775e-05, "loss": 0.5909, "step": 13598, "task_loss": 1.0046473741531372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49878960847854614, "epoch": 11.5, "learning_rate": 1.298058860363181e-05, "loss": 0.5297, "step": 13599, "task_loss": 0.2711784243583679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38937580585479736, "epoch": 11.5, "learning_rate": 1.2977457733249846e-05, "loss": 0.4923, "step": 13600, "task_loss": 0.5963602662086487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.465900719165802, "epoch": 11.5, "learning_rate": 1.2974326862867877e-05, "loss": 0.4755, "step": 13601, "task_loss": 0.8341333866119385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36002957820892334, "epoch": 11.5, "learning_rate": 1.2971195992485913e-05, "loss": 0.4176, "step": 13602, "task_loss": 0.9990655183792114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5240424871444702, "epoch": 11.5, "learning_rate": 1.2968065122103944e-05, "loss": 0.5133, "step": 13603, "task_loss": 1.2146680355072021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37775978446006775, "epoch": 11.5, "learning_rate": 1.296493425172198e-05, "loss": 0.5153, "step": 13604, "task_loss": 0.28952640295028687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6384189128875732, "epoch": 11.5, "learning_rate": 1.2961803381340013e-05, "loss": 0.6843, "step": 13605, "task_loss": 1.2602962255477905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36190587282180786, "epoch": 11.5, "learning_rate": 1.2958672510958048e-05, "loss": 0.4499, "step": 13606, "task_loss": 0.46915069222450256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3443918824195862, "epoch": 11.5, "learning_rate": 1.295554164057608e-05, "loss": 0.4327, "step": 13607, "task_loss": 0.10162229090929031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26044440269470215, "epoch": 11.5, "learning_rate": 1.2952410770194115e-05, "loss": 0.4035, "step": 13608, "task_loss": 0.5630598068237305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40250250697135925, "epoch": 11.5, "learning_rate": 1.2949279899812147e-05, "loss": 0.5351, "step": 13609, "task_loss": 1.1042753458023071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43275901675224304, "epoch": 11.5, "learning_rate": 1.2946149029430182e-05, "loss": 0.5335, "step": 13610, "task_loss": 1.1269489526748657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3307715654373169, "epoch": 11.51, "learning_rate": 1.2943018159048214e-05, "loss": 0.4996, "step": 13611, "task_loss": 0.47766363620758057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3532070219516754, "epoch": 11.51, "learning_rate": 1.2939887288666249e-05, "loss": 0.5095, "step": 13612, "task_loss": 0.3226349353790283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44497227668762207, "epoch": 11.51, "learning_rate": 1.2936756418284284e-05, "loss": 0.6131, "step": 13613, "task_loss": 0.813949465751648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8838702440261841, "epoch": 11.51, "learning_rate": 1.2933625547902318e-05, "loss": 0.6828, "step": 13614, "task_loss": 1.2204327583312988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42079395055770874, "epoch": 11.51, "learning_rate": 1.2930494677520353e-05, "loss": 0.4046, "step": 13615, "task_loss": 0.124517060816288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32038795948028564, "epoch": 11.51, "learning_rate": 1.2927363807138385e-05, "loss": 0.4965, "step": 13616, "task_loss": 0.7868496775627136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4155932664871216, "epoch": 11.51, "learning_rate": 1.292423293675642e-05, "loss": 0.475, "step": 13617, "task_loss": 0.4771329164505005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6867243051528931, "epoch": 11.51, "learning_rate": 1.2921102066374452e-05, "loss": 0.5016, "step": 13618, "task_loss": 0.7922787070274353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40293359756469727, "epoch": 11.51, "learning_rate": 1.2917971195992487e-05, "loss": 0.4805, "step": 13619, "task_loss": 0.3431606590747833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33253973722457886, "epoch": 11.51, "learning_rate": 1.2914840325610519e-05, "loss": 0.4649, "step": 13620, "task_loss": 0.6566414833068848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49608272314071655, "epoch": 11.51, "learning_rate": 1.2911709455228554e-05, "loss": 0.4647, "step": 13621, "task_loss": 1.093940019607544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2844523787498474, "epoch": 11.51, "learning_rate": 1.2908578584846587e-05, "loss": 0.3676, "step": 13622, "task_loss": 0.1261158436536789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5023753046989441, "epoch": 11.52, "learning_rate": 1.2905447714464622e-05, "loss": 0.4612, "step": 13623, "task_loss": 0.5583555102348328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34431424736976624, "epoch": 11.52, "learning_rate": 1.2902316844082658e-05, "loss": 0.51, "step": 13624, "task_loss": 0.07092244178056717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38052207231521606, "epoch": 11.52, "learning_rate": 1.289918597370069e-05, "loss": 0.4005, "step": 13625, "task_loss": 0.5945107340812683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5684419274330139, "epoch": 11.52, "learning_rate": 1.2896055103318724e-05, "loss": 0.5271, "step": 13626, "task_loss": 0.4323827028274536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4163295030593872, "epoch": 11.52, "learning_rate": 1.2892924232936756e-05, "loss": 0.4309, "step": 13627, "task_loss": 0.7310483455657959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030007362365723, "epoch": 11.52, "learning_rate": 1.2889793362554791e-05, "loss": 0.5191, "step": 13628, "task_loss": 0.31837916374206543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8170035481452942, "epoch": 11.52, "learning_rate": 1.2886662492172825e-05, "loss": 0.5526, "step": 13629, "task_loss": 0.7971619963645935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22277000546455383, "epoch": 11.52, "learning_rate": 1.2883531621790858e-05, "loss": 0.3939, "step": 13630, "task_loss": 0.03981225937604904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29667067527770996, "epoch": 11.52, "learning_rate": 1.2880400751408892e-05, "loss": 0.5086, "step": 13631, "task_loss": 0.21904824674129486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27985236048698425, "epoch": 11.52, "learning_rate": 1.2877269881026927e-05, "loss": 0.523, "step": 13632, "task_loss": 0.8531167507171631 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.336548388004303, "epoch": 11.52, "learning_rate": 1.2874139010644959e-05, "loss": 0.3923, "step": 13633, "task_loss": 0.9702622890472412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36609846353530884, "epoch": 11.52, "learning_rate": 1.2871008140262994e-05, "loss": 0.3917, "step": 13634, "task_loss": 0.77543705701828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3643532693386078, "epoch": 11.53, "learning_rate": 1.2867877269881026e-05, "loss": 0.5918, "step": 13635, "task_loss": 0.4082566797733307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38345563411712646, "epoch": 11.53, "learning_rate": 1.2864746399499061e-05, "loss": 0.442, "step": 13636, "task_loss": 1.2656972408294678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5648877024650574, "epoch": 11.53, "learning_rate": 1.2861615529117096e-05, "loss": 0.4054, "step": 13637, "task_loss": 0.6289765238761902 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24408933520317078, "epoch": 11.53, "learning_rate": 1.285848465873513e-05, "loss": 0.5357, "step": 13638, "task_loss": 0.026631729677319527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7068628072738647, "epoch": 11.53, "learning_rate": 1.2855353788353163e-05, "loss": 0.4632, "step": 13639, "task_loss": 1.5097026824951172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6475425958633423, "epoch": 11.53, "learning_rate": 1.2852222917971197e-05, "loss": 0.4658, "step": 13640, "task_loss": 1.3294929265975952 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6214499473571777, "epoch": 11.53, "learning_rate": 1.2849092047589232e-05, "loss": 0.4447, "step": 13641, "task_loss": 1.3888760805130005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6308537721633911, "epoch": 11.53, "learning_rate": 1.2845961177207263e-05, "loss": 0.5698, "step": 13642, "task_loss": 0.9366759061813354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9143182039260864, "epoch": 11.53, "learning_rate": 1.2842830306825299e-05, "loss": 0.6274, "step": 13643, "task_loss": 1.5805721282958984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4190221428871155, "epoch": 11.53, "learning_rate": 1.283969943644333e-05, "loss": 0.4889, "step": 13644, "task_loss": 0.5715625286102295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35378363728523254, "epoch": 11.53, "learning_rate": 1.2836568566061366e-05, "loss": 0.4857, "step": 13645, "task_loss": 0.23176273703575134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6794815063476562, "epoch": 11.53, "learning_rate": 1.2833437695679399e-05, "loss": 0.566, "step": 13646, "task_loss": 0.8980302810668945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6193921566009521, "epoch": 11.54, "learning_rate": 1.2830306825297434e-05, "loss": 0.5407, "step": 13647, "task_loss": 1.0623375177383423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5728660821914673, "epoch": 11.54, "learning_rate": 1.2827175954915466e-05, "loss": 0.4706, "step": 13648, "task_loss": 0.2933824956417084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23068174719810486, "epoch": 11.54, "learning_rate": 1.2824045084533501e-05, "loss": 0.387, "step": 13649, "task_loss": 0.16132429242134094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2961544394493103, "epoch": 11.54, "learning_rate": 1.2820914214151536e-05, "loss": 0.3648, "step": 13650, "task_loss": 0.21025635302066803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34244248270988464, "epoch": 11.54, "learning_rate": 1.2817783343769568e-05, "loss": 0.4038, "step": 13651, "task_loss": 0.1415916234254837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6315490007400513, "epoch": 11.54, "learning_rate": 1.2814652473387603e-05, "loss": 0.4965, "step": 13652, "task_loss": 0.4853336811065674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.585257887840271, "epoch": 11.54, "learning_rate": 1.2811521603005635e-05, "loss": 0.5413, "step": 13653, "task_loss": 1.021960735321045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38229408860206604, "epoch": 11.54, "learning_rate": 1.280839073262367e-05, "loss": 0.3794, "step": 13654, "task_loss": 0.3004344701766968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7356055974960327, "epoch": 11.54, "learning_rate": 1.2805259862241704e-05, "loss": 0.4301, "step": 13655, "task_loss": 0.31877321004867554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3749920427799225, "epoch": 11.54, "learning_rate": 1.2802128991859739e-05, "loss": 0.4606, "step": 13656, "task_loss": 0.7383444309234619 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.625540554523468, "epoch": 11.54, "learning_rate": 1.279899812147777e-05, "loss": 0.5817, "step": 13657, "task_loss": 1.151963472366333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35258162021636963, "epoch": 11.54, "learning_rate": 1.2795867251095806e-05, "loss": 0.3497, "step": 13658, "task_loss": 0.3673621714115143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41779959201812744, "epoch": 11.55, "learning_rate": 1.2792736380713838e-05, "loss": 0.5388, "step": 13659, "task_loss": 0.8465784192085266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5204547643661499, "epoch": 11.55, "learning_rate": 1.2789605510331873e-05, "loss": 0.5332, "step": 13660, "task_loss": 0.6939333081245422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3753197491168976, "epoch": 11.55, "learning_rate": 1.2786474639949908e-05, "loss": 0.5342, "step": 13661, "task_loss": 0.07860612869262695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5699376463890076, "epoch": 11.55, "learning_rate": 1.278334376956794e-05, "loss": 0.5224, "step": 13662, "task_loss": 0.7324782609939575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6845011711120605, "epoch": 11.55, "learning_rate": 1.2780212899185975e-05, "loss": 0.5024, "step": 13663, "task_loss": 1.3177777528762817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5337499380111694, "epoch": 11.55, "learning_rate": 1.2777082028804008e-05, "loss": 0.5807, "step": 13664, "task_loss": 1.3023133277893066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7028467655181885, "epoch": 11.55, "learning_rate": 1.2773951158422044e-05, "loss": 0.5495, "step": 13665, "task_loss": 1.4393965005874634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45471301674842834, "epoch": 11.55, "learning_rate": 1.2770820288040075e-05, "loss": 0.462, "step": 13666, "task_loss": 1.0116595029830933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27742117643356323, "epoch": 11.55, "learning_rate": 1.276768941765811e-05, "loss": 0.5116, "step": 13667, "task_loss": 0.4238336682319641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3471364974975586, "epoch": 11.55, "learning_rate": 1.2764558547276142e-05, "loss": 0.4573, "step": 13668, "task_loss": 0.3930085301399231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6657384037971497, "epoch": 11.55, "learning_rate": 1.2761427676894177e-05, "loss": 0.5073, "step": 13669, "task_loss": 1.0652495622634888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5290519595146179, "epoch": 11.56, "learning_rate": 1.275829680651221e-05, "loss": 0.4582, "step": 13670, "task_loss": 1.1693178415298462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5231592059135437, "epoch": 11.56, "learning_rate": 1.2755165936130244e-05, "loss": 0.5278, "step": 13671, "task_loss": 1.5394930839538574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5507547855377197, "epoch": 11.56, "learning_rate": 1.2752035065748278e-05, "loss": 0.5585, "step": 13672, "task_loss": 0.20816175639629364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4213712811470032, "epoch": 11.56, "learning_rate": 1.2748904195366313e-05, "loss": 0.4665, "step": 13673, "task_loss": 0.7710381746292114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5959760546684265, "epoch": 11.56, "learning_rate": 1.2745773324984348e-05, "loss": 0.5428, "step": 13674, "task_loss": 0.45474734902381897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4029107093811035, "epoch": 11.56, "learning_rate": 1.274264245460238e-05, "loss": 0.3869, "step": 13675, "task_loss": 0.35561656951904297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3513185977935791, "epoch": 11.56, "learning_rate": 1.2739511584220415e-05, "loss": 0.5773, "step": 13676, "task_loss": 1.0004425048828125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21475434303283691, "epoch": 11.56, "learning_rate": 1.2736380713838447e-05, "loss": 0.4456, "step": 13677, "task_loss": 0.9826796650886536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41122812032699585, "epoch": 11.56, "learning_rate": 1.2733249843456482e-05, "loss": 0.4852, "step": 13678, "task_loss": 0.5267574787139893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5272624492645264, "epoch": 11.56, "learning_rate": 1.2730118973074514e-05, "loss": 0.4384, "step": 13679, "task_loss": 0.502621054649353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3725660741329193, "epoch": 11.56, "learning_rate": 1.2726988102692549e-05, "loss": 0.5013, "step": 13680, "task_loss": 0.6792399883270264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44723936915397644, "epoch": 11.56, "learning_rate": 1.2723857232310583e-05, "loss": 0.4796, "step": 13681, "task_loss": 0.27249401807785034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.390105277299881, "epoch": 11.57, "learning_rate": 1.2720726361928618e-05, "loss": 0.4117, "step": 13682, "task_loss": 0.7639283537864685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2725958228111267, "epoch": 11.57, "learning_rate": 1.271759549154665e-05, "loss": 0.4838, "step": 13683, "task_loss": 0.6965284943580627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.560053288936615, "epoch": 11.57, "learning_rate": 1.2714464621164685e-05, "loss": 0.5633, "step": 13684, "task_loss": 1.3182506561279297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44409623742103577, "epoch": 11.57, "learning_rate": 1.2711333750782716e-05, "loss": 0.3878, "step": 13685, "task_loss": 0.3006393313407898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.051825761795044, "epoch": 11.57, "learning_rate": 1.2708202880400752e-05, "loss": 0.7136, "step": 13686, "task_loss": 0.7673883438110352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40187978744506836, "epoch": 11.57, "learning_rate": 1.2705072010018787e-05, "loss": 0.5139, "step": 13687, "task_loss": 0.8714907765388489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34959861636161804, "epoch": 11.57, "learning_rate": 1.2701941139636819e-05, "loss": 0.4156, "step": 13688, "task_loss": 0.538514256477356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5595634579658508, "epoch": 11.57, "learning_rate": 1.2698810269254854e-05, "loss": 0.4428, "step": 13689, "task_loss": 0.49697110056877136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34711432456970215, "epoch": 11.57, "learning_rate": 1.2695679398872887e-05, "loss": 0.4805, "step": 13690, "task_loss": 0.5835899114608765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4557367265224457, "epoch": 11.57, "learning_rate": 1.2692548528490922e-05, "loss": 0.4236, "step": 13691, "task_loss": 0.3937794864177704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6470614075660706, "epoch": 11.57, "learning_rate": 1.2689417658108954e-05, "loss": 0.5638, "step": 13692, "task_loss": 0.8968009948730469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3590109944343567, "epoch": 11.57, "learning_rate": 1.268628678772699e-05, "loss": 0.3794, "step": 13693, "task_loss": 1.1595172882080078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42908361554145813, "epoch": 11.58, "learning_rate": 1.2683155917345021e-05, "loss": 0.3677, "step": 13694, "task_loss": 0.3257366418838501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40197446942329407, "epoch": 11.58, "learning_rate": 1.2680025046963056e-05, "loss": 0.4366, "step": 13695, "task_loss": 0.3186877965927124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.70206618309021, "epoch": 11.58, "learning_rate": 1.2676894176581088e-05, "loss": 0.479, "step": 13696, "task_loss": 1.1095941066741943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41027989983558655, "epoch": 11.58, "learning_rate": 1.2673763306199123e-05, "loss": 0.4243, "step": 13697, "task_loss": 1.1562209129333496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4673895835876465, "epoch": 11.58, "learning_rate": 1.2670632435817158e-05, "loss": 0.4863, "step": 13698, "task_loss": 0.10698428004980087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42072516679763794, "epoch": 11.58, "learning_rate": 1.2667501565435192e-05, "loss": 0.4901, "step": 13699, "task_loss": 1.1298396587371826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34735792875289917, "epoch": 11.58, "learning_rate": 1.2664370695053227e-05, "loss": 0.5623, "step": 13700, "task_loss": 0.2503602206707001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35101184248924255, "epoch": 11.58, "learning_rate": 1.2661239824671259e-05, "loss": 0.4361, "step": 13701, "task_loss": 0.2601656913757324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37387150526046753, "epoch": 11.58, "learning_rate": 1.2658108954289294e-05, "loss": 0.4555, "step": 13702, "task_loss": 1.3425347805023193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4736236035823822, "epoch": 11.58, "learning_rate": 1.2654978083907326e-05, "loss": 0.5838, "step": 13703, "task_loss": 0.7630059719085693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4524514377117157, "epoch": 11.58, "learning_rate": 1.2651847213525361e-05, "loss": 0.5528, "step": 13704, "task_loss": 0.8670541644096375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4179629385471344, "epoch": 11.58, "learning_rate": 1.2648716343143394e-05, "loss": 0.527, "step": 13705, "task_loss": 1.0152534246444702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3232167065143585, "epoch": 11.59, "learning_rate": 1.2645585472761428e-05, "loss": 0.3988, "step": 13706, "task_loss": 0.18080496788024902 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38011467456817627, "epoch": 11.59, "learning_rate": 1.2642454602379461e-05, "loss": 0.474, "step": 13707, "task_loss": 0.36689749360084534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39434200525283813, "epoch": 11.59, "learning_rate": 1.2639323731997497e-05, "loss": 0.3862, "step": 13708, "task_loss": 0.5536545515060425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8864916563034058, "epoch": 11.59, "learning_rate": 1.2636192861615528e-05, "loss": 0.6575, "step": 13709, "task_loss": 0.9639517664909363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3266053795814514, "epoch": 11.59, "learning_rate": 1.2633061991233563e-05, "loss": 0.3622, "step": 13710, "task_loss": 0.32397663593292236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5394694209098816, "epoch": 11.59, "learning_rate": 1.2629931120851599e-05, "loss": 0.6083, "step": 13711, "task_loss": 0.5402632355690002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38822782039642334, "epoch": 11.59, "learning_rate": 1.262680025046963e-05, "loss": 0.4682, "step": 13712, "task_loss": 0.680937647819519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8563036918640137, "epoch": 11.59, "learning_rate": 1.2623669380087666e-05, "loss": 0.6131, "step": 13713, "task_loss": 1.0878950357437134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5210243463516235, "epoch": 11.59, "learning_rate": 1.2620538509705699e-05, "loss": 0.4895, "step": 13714, "task_loss": 0.959894061088562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5470781922340393, "epoch": 11.59, "learning_rate": 1.2617407639323733e-05, "loss": 0.5456, "step": 13715, "task_loss": 0.8804739117622375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41541191935539246, "epoch": 11.59, "learning_rate": 1.2614276768941766e-05, "loss": 0.4369, "step": 13716, "task_loss": 0.5609145760536194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5331779718399048, "epoch": 11.59, "learning_rate": 1.2611145898559801e-05, "loss": 0.4858, "step": 13717, "task_loss": 0.9636666774749756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48515790700912476, "epoch": 11.6, "learning_rate": 1.2608015028177833e-05, "loss": 0.3934, "step": 13718, "task_loss": 0.41082993149757385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21295365691184998, "epoch": 11.6, "learning_rate": 1.2604884157795868e-05, "loss": 0.3523, "step": 13719, "task_loss": 0.1338648796081543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.612234354019165, "epoch": 11.6, "learning_rate": 1.26017532874139e-05, "loss": 0.5679, "step": 13720, "task_loss": 1.8202091455459595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7095921039581299, "epoch": 11.6, "learning_rate": 1.2598622417031935e-05, "loss": 0.5058, "step": 13721, "task_loss": 0.8892707824707031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6238918304443359, "epoch": 11.6, "learning_rate": 1.259549154664997e-05, "loss": 0.6293, "step": 13722, "task_loss": 0.8853986263275146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43231070041656494, "epoch": 11.6, "learning_rate": 1.2592360676268004e-05, "loss": 0.5105, "step": 13723, "task_loss": 0.7549663782119751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43143171072006226, "epoch": 11.6, "learning_rate": 1.2589229805886037e-05, "loss": 0.4762, "step": 13724, "task_loss": 0.4525502920150757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40028417110443115, "epoch": 11.6, "learning_rate": 1.258609893550407e-05, "loss": 0.5609, "step": 13725, "task_loss": 0.6316252946853638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.407858282327652, "epoch": 11.6, "learning_rate": 1.2582968065122106e-05, "loss": 0.5422, "step": 13726, "task_loss": 1.0339175462722778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5286489725112915, "epoch": 11.6, "learning_rate": 1.2579837194740138e-05, "loss": 0.4888, "step": 13727, "task_loss": 0.7282783389091492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4244403839111328, "epoch": 11.6, "learning_rate": 1.2576706324358173e-05, "loss": 0.5604, "step": 13728, "task_loss": 0.7822986245155334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6039859056472778, "epoch": 11.6, "learning_rate": 1.2573575453976205e-05, "loss": 0.4255, "step": 13729, "task_loss": 1.0468279123306274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40645432472229004, "epoch": 11.61, "learning_rate": 1.257044458359424e-05, "loss": 0.6261, "step": 13730, "task_loss": 0.24501779675483704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7171295881271362, "epoch": 11.61, "learning_rate": 1.2567313713212273e-05, "loss": 0.5673, "step": 13731, "task_loss": 1.0539225339889526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5020065307617188, "epoch": 11.61, "learning_rate": 1.2564182842830308e-05, "loss": 0.4475, "step": 13732, "task_loss": 0.5688334703445435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5253486633300781, "epoch": 11.61, "learning_rate": 1.256105197244834e-05, "loss": 0.4976, "step": 13733, "task_loss": 1.614999771118164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5429785251617432, "epoch": 11.61, "learning_rate": 1.2557921102066375e-05, "loss": 0.5236, "step": 13734, "task_loss": 0.2940812110900879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3892042934894562, "epoch": 11.61, "learning_rate": 1.255479023168441e-05, "loss": 0.3772, "step": 13735, "task_loss": 0.2974638342857361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4133490324020386, "epoch": 11.61, "learning_rate": 1.2551659361302442e-05, "loss": 0.4511, "step": 13736, "task_loss": 0.6415655016899109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3257046341896057, "epoch": 11.61, "learning_rate": 1.2548528490920477e-05, "loss": 0.4847, "step": 13737, "task_loss": 0.30878621339797974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5460668802261353, "epoch": 11.61, "learning_rate": 1.254539762053851e-05, "loss": 0.5341, "step": 13738, "task_loss": 0.6328739523887634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25418999791145325, "epoch": 11.61, "learning_rate": 1.2542266750156544e-05, "loss": 0.3748, "step": 13739, "task_loss": 0.05458153039216995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4451879858970642, "epoch": 11.61, "learning_rate": 1.2539135879774578e-05, "loss": 0.5612, "step": 13740, "task_loss": 1.0539852380752563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3057202994823456, "epoch": 11.61, "learning_rate": 1.2536005009392613e-05, "loss": 0.4211, "step": 13741, "task_loss": 0.7583096027374268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37336352467536926, "epoch": 11.62, "learning_rate": 1.2532874139010645e-05, "loss": 0.4657, "step": 13742, "task_loss": 0.4750801622867584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27916866540908813, "epoch": 11.62, "learning_rate": 1.252974326862868e-05, "loss": 0.4487, "step": 13743, "task_loss": 0.615877091884613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4428112804889679, "epoch": 11.62, "learning_rate": 1.2526612398246712e-05, "loss": 0.4212, "step": 13744, "task_loss": 1.0532232522964478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4440600872039795, "epoch": 11.62, "learning_rate": 1.2523481527864747e-05, "loss": 0.4586, "step": 13745, "task_loss": 0.23584559559822083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7942547798156738, "epoch": 11.62, "learning_rate": 1.2520350657482779e-05, "loss": 0.4928, "step": 13746, "task_loss": 0.48322007060050964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5989125967025757, "epoch": 11.62, "learning_rate": 1.2517219787100814e-05, "loss": 0.5158, "step": 13747, "task_loss": 0.9196664690971375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37095457315444946, "epoch": 11.62, "learning_rate": 1.2514088916718849e-05, "loss": 0.5103, "step": 13748, "task_loss": 0.6139047741889954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5283831357955933, "epoch": 11.62, "learning_rate": 1.2510958046336883e-05, "loss": 0.5876, "step": 13749, "task_loss": 0.9106743335723877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8065201640129089, "epoch": 11.62, "learning_rate": 1.2507827175954918e-05, "loss": 0.6579, "step": 13750, "task_loss": 0.49501317739486694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5149726867675781, "epoch": 11.62, "learning_rate": 1.250469630557295e-05, "loss": 0.4383, "step": 13751, "task_loss": 0.9655632376670837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44741344451904297, "epoch": 11.62, "learning_rate": 1.2501565435190985e-05, "loss": 0.592, "step": 13752, "task_loss": 0.40714332461357117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.264618456363678, "epoch": 11.63, "learning_rate": 1.2498434564809018e-05, "loss": 0.3133, "step": 13753, "task_loss": 0.07468917965888977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3652157783508301, "epoch": 11.63, "learning_rate": 1.2495303694427052e-05, "loss": 0.3639, "step": 13754, "task_loss": 0.5975590348243713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5319554805755615, "epoch": 11.63, "learning_rate": 1.2492172824045085e-05, "loss": 0.5531, "step": 13755, "task_loss": 0.9607937335968018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25033867359161377, "epoch": 11.63, "learning_rate": 1.2489041953663119e-05, "loss": 0.4125, "step": 13756, "task_loss": 0.1265423744916916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5228519439697266, "epoch": 11.63, "learning_rate": 1.2485911083281154e-05, "loss": 0.5701, "step": 13757, "task_loss": 1.2885515689849854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3058943748474121, "epoch": 11.63, "learning_rate": 1.2482780212899187e-05, "loss": 0.4938, "step": 13758, "task_loss": 0.8815776109695435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28757837414741516, "epoch": 11.63, "learning_rate": 1.247964934251722e-05, "loss": 0.6082, "step": 13759, "task_loss": 0.3945768475532532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5050127506256104, "epoch": 11.63, "learning_rate": 1.2476518472135254e-05, "loss": 0.5815, "step": 13760, "task_loss": 0.7021082043647766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4876416325569153, "epoch": 11.63, "learning_rate": 1.2473387601753288e-05, "loss": 0.5336, "step": 13761, "task_loss": 0.3285619020462036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18347090482711792, "epoch": 11.63, "learning_rate": 1.2470256731371321e-05, "loss": 0.4807, "step": 13762, "task_loss": 0.31503546237945557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43747732043266296, "epoch": 11.63, "learning_rate": 1.2467125860989355e-05, "loss": 0.4274, "step": 13763, "task_loss": 0.24492157995700836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5105479955673218, "epoch": 11.63, "learning_rate": 1.2463994990607388e-05, "loss": 0.4743, "step": 13764, "task_loss": 0.9211942553520203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17240948975086212, "epoch": 11.64, "learning_rate": 1.2460864120225423e-05, "loss": 0.4281, "step": 13765, "task_loss": 0.340116024017334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8573293089866638, "epoch": 11.64, "learning_rate": 1.2457733249843458e-05, "loss": 0.5594, "step": 13766, "task_loss": 0.5121303796768188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2982115149497986, "epoch": 11.64, "learning_rate": 1.2454602379461492e-05, "loss": 0.4394, "step": 13767, "task_loss": 0.3140920400619507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4447312355041504, "epoch": 11.64, "learning_rate": 1.2451471509079525e-05, "loss": 0.5693, "step": 13768, "task_loss": 0.8767495155334473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41510632634162903, "epoch": 11.64, "learning_rate": 1.2448340638697559e-05, "loss": 0.5095, "step": 13769, "task_loss": 0.38830533623695374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5311852693557739, "epoch": 11.64, "learning_rate": 1.2445209768315592e-05, "loss": 0.5415, "step": 13770, "task_loss": 1.127919316291809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28192514181137085, "epoch": 11.64, "learning_rate": 1.2442078897933626e-05, "loss": 0.5291, "step": 13771, "task_loss": 0.23582136631011963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41147398948669434, "epoch": 11.64, "learning_rate": 1.243894802755166e-05, "loss": 0.3709, "step": 13772, "task_loss": 1.0008090734481812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3954938054084778, "epoch": 11.64, "learning_rate": 1.2435817157169693e-05, "loss": 0.3818, "step": 13773, "task_loss": 0.16401752829551697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3946489095687866, "epoch": 11.64, "learning_rate": 1.2432686286787728e-05, "loss": 0.4389, "step": 13774, "task_loss": 0.43688708543777466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6172383427619934, "epoch": 11.64, "learning_rate": 1.2429555416405761e-05, "loss": 0.5515, "step": 13775, "task_loss": 0.5782727599143982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.516512393951416, "epoch": 11.64, "learning_rate": 1.2426424546023795e-05, "loss": 0.5927, "step": 13776, "task_loss": 0.15968118607997894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47924643754959106, "epoch": 11.65, "learning_rate": 1.2423293675641828e-05, "loss": 0.5316, "step": 13777, "task_loss": 0.49871405959129333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48993444442749023, "epoch": 11.65, "learning_rate": 1.2420162805259864e-05, "loss": 0.5441, "step": 13778, "task_loss": 0.725088894367218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5165560245513916, "epoch": 11.65, "learning_rate": 1.2417031934877897e-05, "loss": 0.4998, "step": 13779, "task_loss": 0.7782955169677734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17877525091171265, "epoch": 11.65, "learning_rate": 1.241390106449593e-05, "loss": 0.4775, "step": 13780, "task_loss": 0.1677992343902588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6417998671531677, "epoch": 11.65, "learning_rate": 1.2410770194113964e-05, "loss": 0.5576, "step": 13781, "task_loss": 0.9356915354728699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2664119601249695, "epoch": 11.65, "learning_rate": 1.2407639323731997e-05, "loss": 0.393, "step": 13782, "task_loss": 0.5726749300956726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3544803559780121, "epoch": 11.65, "learning_rate": 1.2404508453350033e-05, "loss": 0.4717, "step": 13783, "task_loss": 0.12915858626365662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2798641324043274, "epoch": 11.65, "learning_rate": 1.2401377582968066e-05, "loss": 0.4604, "step": 13784, "task_loss": 0.39576059579849243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3464524447917938, "epoch": 11.65, "learning_rate": 1.23982467125861e-05, "loss": 0.4117, "step": 13785, "task_loss": 0.34839656949043274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37567418813705444, "epoch": 11.65, "learning_rate": 1.2395115842204133e-05, "loss": 0.4099, "step": 13786, "task_loss": 0.07728774845600128 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4176000654697418, "epoch": 11.65, "learning_rate": 1.2391984971822167e-05, "loss": 0.4205, "step": 13787, "task_loss": 0.26376813650131226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41078102588653564, "epoch": 11.65, "learning_rate": 1.23888541014402e-05, "loss": 0.4378, "step": 13788, "task_loss": 1.122238278388977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24278204143047333, "epoch": 11.66, "learning_rate": 1.2385723231058233e-05, "loss": 0.4241, "step": 13789, "task_loss": 0.2806938886642456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6979601979255676, "epoch": 11.66, "learning_rate": 1.2382592360676269e-05, "loss": 0.6137, "step": 13790, "task_loss": 0.8490599989891052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.303165465593338, "epoch": 11.66, "learning_rate": 1.2379461490294302e-05, "loss": 0.4326, "step": 13791, "task_loss": 0.18321290612220764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5907576084136963, "epoch": 11.66, "learning_rate": 1.2376330619912337e-05, "loss": 0.6344, "step": 13792, "task_loss": 0.25189727544784546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5305290222167969, "epoch": 11.66, "learning_rate": 1.237319974953037e-05, "loss": 0.4476, "step": 13793, "task_loss": 0.8307930827140808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4824160039424896, "epoch": 11.66, "learning_rate": 1.2370068879148404e-05, "loss": 0.5382, "step": 13794, "task_loss": 1.616578221321106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3117449879646301, "epoch": 11.66, "learning_rate": 1.2366938008766438e-05, "loss": 0.4227, "step": 13795, "task_loss": 0.7621198892593384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45371025800704956, "epoch": 11.66, "learning_rate": 1.2363807138384471e-05, "loss": 0.6413, "step": 13796, "task_loss": 0.32861432433128357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5955427885055542, "epoch": 11.66, "learning_rate": 1.2360676268002505e-05, "loss": 0.5199, "step": 13797, "task_loss": 1.0553674697875977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40721726417541504, "epoch": 11.66, "learning_rate": 1.2357545397620538e-05, "loss": 0.3794, "step": 13798, "task_loss": 1.0848819017410278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5038737654685974, "epoch": 11.66, "learning_rate": 1.2354414527238573e-05, "loss": 0.4982, "step": 13799, "task_loss": 0.663417637348175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3632417321205139, "epoch": 11.66, "learning_rate": 1.2351283656856607e-05, "loss": 0.549, "step": 13800, "task_loss": 0.59310382604599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3582364022731781, "epoch": 11.67, "learning_rate": 1.234815278647464e-05, "loss": 0.4014, "step": 13801, "task_loss": 0.33013665676116943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47764065861701965, "epoch": 11.67, "learning_rate": 1.2345021916092675e-05, "loss": 0.4969, "step": 13802, "task_loss": 0.4467492401599884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3203350007534027, "epoch": 11.67, "learning_rate": 1.2341891045710709e-05, "loss": 0.4105, "step": 13803, "task_loss": 1.1633602380752563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6523016691207886, "epoch": 11.67, "learning_rate": 1.2338760175328742e-05, "loss": 0.615, "step": 13804, "task_loss": 0.9446948170661926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.345590204000473, "epoch": 11.67, "learning_rate": 1.2335629304946776e-05, "loss": 0.3287, "step": 13805, "task_loss": 0.7274593710899353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5391051769256592, "epoch": 11.67, "learning_rate": 1.233249843456481e-05, "loss": 0.5773, "step": 13806, "task_loss": 0.25381332635879517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.669338047504425, "epoch": 11.67, "learning_rate": 1.2329367564182843e-05, "loss": 0.3645, "step": 13807, "task_loss": 0.9288575053215027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3645652234554291, "epoch": 11.67, "learning_rate": 1.2326236693800878e-05, "loss": 0.4782, "step": 13808, "task_loss": 0.43509939312934875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6167148351669312, "epoch": 11.67, "learning_rate": 1.2323105823418911e-05, "loss": 0.4503, "step": 13809, "task_loss": 1.1950784921646118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2741347551345825, "epoch": 11.67, "learning_rate": 1.2319974953036945e-05, "loss": 0.4476, "step": 13810, "task_loss": 0.1760166883468628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37760019302368164, "epoch": 11.67, "learning_rate": 1.2316844082654978e-05, "loss": 0.645, "step": 13811, "task_loss": 0.5763120055198669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47324252128601074, "epoch": 11.67, "learning_rate": 1.2313713212273012e-05, "loss": 0.4957, "step": 13812, "task_loss": 0.4382353127002716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3341653048992157, "epoch": 11.68, "learning_rate": 1.2310582341891045e-05, "loss": 0.6112, "step": 13813, "task_loss": 0.48091453313827515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4452115297317505, "epoch": 11.68, "learning_rate": 1.230745147150908e-05, "loss": 0.4164, "step": 13814, "task_loss": 0.2768070697784424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3906286954879761, "epoch": 11.68, "learning_rate": 1.2304320601127114e-05, "loss": 0.4249, "step": 13815, "task_loss": 0.6501743793487549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6213551163673401, "epoch": 11.68, "learning_rate": 1.2301189730745147e-05, "loss": 0.4364, "step": 13816, "task_loss": 0.4338041841983795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28669536113739014, "epoch": 11.68, "learning_rate": 1.2298058860363183e-05, "loss": 0.4038, "step": 13817, "task_loss": 0.6845242381095886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46947911381721497, "epoch": 11.68, "learning_rate": 1.2294927989981216e-05, "loss": 0.6311, "step": 13818, "task_loss": 0.4467426538467407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19195236265659332, "epoch": 11.68, "learning_rate": 1.229179711959925e-05, "loss": 0.4934, "step": 13819, "task_loss": 1.317919135093689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3458535075187683, "epoch": 11.68, "learning_rate": 1.2288666249217283e-05, "loss": 0.4035, "step": 13820, "task_loss": 0.37994620203971863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3843117952346802, "epoch": 11.68, "learning_rate": 1.2285535378835317e-05, "loss": 0.5123, "step": 13821, "task_loss": 0.09649071842432022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4235343933105469, "epoch": 11.68, "learning_rate": 1.228240450845335e-05, "loss": 0.4524, "step": 13822, "task_loss": 0.9579771757125854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5496442914009094, "epoch": 11.68, "learning_rate": 1.2279273638071383e-05, "loss": 0.4703, "step": 13823, "task_loss": 1.200798749923706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48002123832702637, "epoch": 11.69, "learning_rate": 1.2276142767689417e-05, "loss": 0.4123, "step": 13824, "task_loss": 1.1022429466247559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3661094307899475, "epoch": 11.69, "learning_rate": 1.2273011897307452e-05, "loss": 0.5587, "step": 13825, "task_loss": 0.3020821809768677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2787158191204071, "epoch": 11.69, "learning_rate": 1.2269881026925486e-05, "loss": 0.4358, "step": 13826, "task_loss": 0.12760105729103088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4429776072502136, "epoch": 11.69, "learning_rate": 1.226675015654352e-05, "loss": 0.4359, "step": 13827, "task_loss": 0.9390592575073242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43697965145111084, "epoch": 11.69, "learning_rate": 1.2263619286161554e-05, "loss": 0.5852, "step": 13828, "task_loss": 0.7021316289901733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5650709867477417, "epoch": 11.69, "learning_rate": 1.2260488415779588e-05, "loss": 0.4785, "step": 13829, "task_loss": 0.6050969958305359 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9212340116500854, "epoch": 11.69, "learning_rate": 1.2257357545397621e-05, "loss": 0.5087, "step": 13830, "task_loss": 0.48537617921829224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2764546275138855, "epoch": 11.69, "learning_rate": 1.2254226675015655e-05, "loss": 0.3839, "step": 13831, "task_loss": 0.26941022276878357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4880419969558716, "epoch": 11.69, "learning_rate": 1.2251095804633688e-05, "loss": 0.4251, "step": 13832, "task_loss": 0.5273005962371826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5885966420173645, "epoch": 11.69, "learning_rate": 1.2247964934251723e-05, "loss": 0.4253, "step": 13833, "task_loss": 0.7910374402999878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3442428708076477, "epoch": 11.69, "learning_rate": 1.2244834063869757e-05, "loss": 0.3904, "step": 13834, "task_loss": 0.24800081551074982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4312483072280884, "epoch": 11.69, "learning_rate": 1.224170319348779e-05, "loss": 0.558, "step": 13835, "task_loss": 0.3841242492198944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4484248757362366, "epoch": 11.7, "learning_rate": 1.2238572323105824e-05, "loss": 0.4342, "step": 13836, "task_loss": 1.3116505146026611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5497550964355469, "epoch": 11.7, "learning_rate": 1.2235441452723857e-05, "loss": 0.5364, "step": 13837, "task_loss": 1.5675450563430786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6128219366073608, "epoch": 11.7, "learning_rate": 1.223231058234189e-05, "loss": 0.6145, "step": 13838, "task_loss": 1.38715660572052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3168964684009552, "epoch": 11.7, "learning_rate": 1.2229179711959926e-05, "loss": 0.3299, "step": 13839, "task_loss": 0.6353919506072998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18186262249946594, "epoch": 11.7, "learning_rate": 1.222604884157796e-05, "loss": 0.4179, "step": 13840, "task_loss": 0.10946784913539886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40236276388168335, "epoch": 11.7, "learning_rate": 1.2222917971195993e-05, "loss": 0.5233, "step": 13841, "task_loss": 0.18518279492855072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.338421106338501, "epoch": 11.7, "learning_rate": 1.2219787100814028e-05, "loss": 0.3384, "step": 13842, "task_loss": 0.9385035037994385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6382277607917786, "epoch": 11.7, "learning_rate": 1.2216656230432061e-05, "loss": 0.4728, "step": 13843, "task_loss": 1.5686824321746826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3927868604660034, "epoch": 11.7, "learning_rate": 1.2213525360050095e-05, "loss": 0.4446, "step": 13844, "task_loss": 1.0716830492019653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5095841884613037, "epoch": 11.7, "learning_rate": 1.2210394489668128e-05, "loss": 0.4356, "step": 13845, "task_loss": 0.5720797181129456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34315788745880127, "epoch": 11.7, "learning_rate": 1.2207263619286162e-05, "loss": 0.4454, "step": 13846, "task_loss": 0.4894440174102783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.520668089389801, "epoch": 11.7, "learning_rate": 1.2204132748904195e-05, "loss": 0.4848, "step": 13847, "task_loss": 0.7416315674781799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5370751023292542, "epoch": 11.71, "learning_rate": 1.2201001878522229e-05, "loss": 0.4959, "step": 13848, "task_loss": 0.4330871105194092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4263536334037781, "epoch": 11.71, "learning_rate": 1.2197871008140262e-05, "loss": 0.5095, "step": 13849, "task_loss": 0.6180939674377441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5560932159423828, "epoch": 11.71, "learning_rate": 1.2194740137758297e-05, "loss": 0.4047, "step": 13850, "task_loss": 0.7519540786743164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4532046616077423, "epoch": 11.71, "learning_rate": 1.2191609267376333e-05, "loss": 0.6113, "step": 13851, "task_loss": 0.039424341171979904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36168038845062256, "epoch": 11.71, "learning_rate": 1.2188478396994366e-05, "loss": 0.5064, "step": 13852, "task_loss": 0.2566092014312744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18694233894348145, "epoch": 11.71, "learning_rate": 1.21853475266124e-05, "loss": 0.2917, "step": 13853, "task_loss": 0.12006955593824387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22784747183322906, "epoch": 11.71, "learning_rate": 1.2182216656230433e-05, "loss": 0.3949, "step": 13854, "task_loss": 0.2265239953994751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3856375515460968, "epoch": 11.71, "learning_rate": 1.2179085785848467e-05, "loss": 0.5483, "step": 13855, "task_loss": 1.058974266052246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3827498257160187, "epoch": 11.71, "learning_rate": 1.21759549154665e-05, "loss": 0.4092, "step": 13856, "task_loss": 0.916482150554657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3845829665660858, "epoch": 11.71, "learning_rate": 1.2172824045084533e-05, "loss": 0.4895, "step": 13857, "task_loss": 1.0446637868881226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2607157826423645, "epoch": 11.71, "learning_rate": 1.2169693174702567e-05, "loss": 0.4261, "step": 13858, "task_loss": 0.7747445702552795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4244752526283264, "epoch": 11.71, "learning_rate": 1.2166562304320602e-05, "loss": 0.4826, "step": 13859, "task_loss": 0.4488050639629364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41294237971305847, "epoch": 11.72, "learning_rate": 1.2163431433938636e-05, "loss": 0.6543, "step": 13860, "task_loss": 0.4131245017051697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4035091996192932, "epoch": 11.72, "learning_rate": 1.2160300563556669e-05, "loss": 0.4609, "step": 13861, "task_loss": 0.691680371761322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2581302523612976, "epoch": 11.72, "learning_rate": 1.2157169693174703e-05, "loss": 0.5011, "step": 13862, "task_loss": 0.5820341110229492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2300643026828766, "epoch": 11.72, "learning_rate": 1.2154038822792738e-05, "loss": 0.2949, "step": 13863, "task_loss": 0.04785415530204773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43561768531799316, "epoch": 11.72, "learning_rate": 1.2150907952410771e-05, "loss": 0.6235, "step": 13864, "task_loss": 0.6403368711471558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7227959632873535, "epoch": 11.72, "learning_rate": 1.2147777082028805e-05, "loss": 0.526, "step": 13865, "task_loss": 0.6579365134239197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45373713970184326, "epoch": 11.72, "learning_rate": 1.2144646211646838e-05, "loss": 0.3906, "step": 13866, "task_loss": 0.9900370836257935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4382374584674835, "epoch": 11.72, "learning_rate": 1.2141515341264872e-05, "loss": 0.4438, "step": 13867, "task_loss": 0.2287328541278839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33591368794441223, "epoch": 11.72, "learning_rate": 1.2138384470882907e-05, "loss": 0.4126, "step": 13868, "task_loss": 0.6931079030036926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4817171096801758, "epoch": 11.72, "learning_rate": 1.213525360050094e-05, "loss": 0.4698, "step": 13869, "task_loss": 0.527594804763794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40770357847213745, "epoch": 11.72, "learning_rate": 1.2132122730118974e-05, "loss": 0.365, "step": 13870, "task_loss": 1.319311261177063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5092493295669556, "epoch": 11.72, "learning_rate": 1.2128991859737007e-05, "loss": 0.4243, "step": 13871, "task_loss": 0.8180999755859375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3063533306121826, "epoch": 11.73, "learning_rate": 1.212586098935504e-05, "loss": 0.2881, "step": 13872, "task_loss": 0.0993763878941536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29896026849746704, "epoch": 11.73, "learning_rate": 1.2122730118973074e-05, "loss": 0.5291, "step": 13873, "task_loss": 0.2798925042152405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21853424608707428, "epoch": 11.73, "learning_rate": 1.2119599248591108e-05, "loss": 0.3403, "step": 13874, "task_loss": 0.021210990846157074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5338377356529236, "epoch": 11.73, "learning_rate": 1.2116468378209143e-05, "loss": 0.4658, "step": 13875, "task_loss": 0.8526037931442261 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.597014844417572, "epoch": 11.73, "learning_rate": 1.2113337507827176e-05, "loss": 0.4913, "step": 13876, "task_loss": 0.36987918615341187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42443040013313293, "epoch": 11.73, "learning_rate": 1.2110206637445211e-05, "loss": 0.3595, "step": 13877, "task_loss": 0.27027183771133423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43069687485694885, "epoch": 11.73, "learning_rate": 1.2107075767063245e-05, "loss": 0.4265, "step": 13878, "task_loss": 0.22988125681877136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41738182306289673, "epoch": 11.73, "learning_rate": 1.2103944896681278e-05, "loss": 0.4329, "step": 13879, "task_loss": 1.1797202825546265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6630719900131226, "epoch": 11.73, "learning_rate": 1.2100814026299312e-05, "loss": 0.438, "step": 13880, "task_loss": 0.49728259444236755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5223828554153442, "epoch": 11.73, "learning_rate": 1.2097683155917345e-05, "loss": 0.6017, "step": 13881, "task_loss": 0.8540897369384766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5686228275299072, "epoch": 11.73, "learning_rate": 1.2094552285535379e-05, "loss": 0.4873, "step": 13882, "task_loss": 0.7001005411148071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4876648485660553, "epoch": 11.73, "learning_rate": 1.2091421415153412e-05, "loss": 0.4007, "step": 13883, "task_loss": 0.39440834522247314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4159260392189026, "epoch": 11.74, "learning_rate": 1.2088290544771447e-05, "loss": 0.6267, "step": 13884, "task_loss": 1.179342269897461 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41571056842803955, "epoch": 11.74, "learning_rate": 1.2085159674389481e-05, "loss": 0.411, "step": 13885, "task_loss": 0.5380693674087524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6702549457550049, "epoch": 11.74, "learning_rate": 1.2082028804007514e-05, "loss": 0.4997, "step": 13886, "task_loss": 1.028802752494812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6487961411476135, "epoch": 11.74, "learning_rate": 1.2078897933625548e-05, "loss": 0.5145, "step": 13887, "task_loss": 0.8573175668716431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5972914099693298, "epoch": 11.74, "learning_rate": 1.2075767063243583e-05, "loss": 0.555, "step": 13888, "task_loss": 0.5808204412460327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4997016191482544, "epoch": 11.74, "learning_rate": 1.2072636192861617e-05, "loss": 0.5401, "step": 13889, "task_loss": 0.6227388978004456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5588910579681396, "epoch": 11.74, "learning_rate": 1.206950532247965e-05, "loss": 0.4862, "step": 13890, "task_loss": 0.9969218373298645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42395341396331787, "epoch": 11.74, "learning_rate": 1.2066374452097684e-05, "loss": 0.4981, "step": 13891, "task_loss": 0.3305291533470154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3777437210083008, "epoch": 11.74, "learning_rate": 1.2063243581715717e-05, "loss": 0.3685, "step": 13892, "task_loss": 0.17135664820671082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28818026185035706, "epoch": 11.74, "learning_rate": 1.2060112711333752e-05, "loss": 0.4458, "step": 13893, "task_loss": 0.1307392120361328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24682079255580902, "epoch": 11.74, "learning_rate": 1.2056981840951786e-05, "loss": 0.3803, "step": 13894, "task_loss": 0.7261713743209839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8117759227752686, "epoch": 11.75, "learning_rate": 1.2053850970569819e-05, "loss": 0.5259, "step": 13895, "task_loss": 1.265139102935791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47725221514701843, "epoch": 11.75, "learning_rate": 1.2050720100187853e-05, "loss": 0.3803, "step": 13896, "task_loss": 0.8386959433555603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8365530967712402, "epoch": 11.75, "learning_rate": 1.2047589229805886e-05, "loss": 0.6026, "step": 13897, "task_loss": 1.0320807695388794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34286612272262573, "epoch": 11.75, "learning_rate": 1.204445835942392e-05, "loss": 0.4421, "step": 13898, "task_loss": 0.9136995673179626 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45925605297088623, "epoch": 11.75, "learning_rate": 1.2041327489041953e-05, "loss": 0.6769, "step": 13899, "task_loss": 0.9197911024093628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5562574863433838, "epoch": 11.75, "learning_rate": 1.2038196618659988e-05, "loss": 0.538, "step": 13900, "task_loss": 1.74964439868927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5318293571472168, "epoch": 11.75, "learning_rate": 1.2035065748278022e-05, "loss": 0.4549, "step": 13901, "task_loss": 0.5083686113357544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4743778109550476, "epoch": 11.75, "learning_rate": 1.2031934877896057e-05, "loss": 0.4877, "step": 13902, "task_loss": 0.824667751789093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5597249269485474, "epoch": 11.75, "learning_rate": 1.202880400751409e-05, "loss": 0.4296, "step": 13903, "task_loss": 0.11213166266679764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5159629583358765, "epoch": 11.75, "learning_rate": 1.2025673137132124e-05, "loss": 0.5193, "step": 13904, "task_loss": 0.28777608275413513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34535402059555054, "epoch": 11.75, "learning_rate": 1.2022542266750157e-05, "loss": 0.4365, "step": 13905, "task_loss": 0.0393868125975132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3955748677253723, "epoch": 11.75, "learning_rate": 1.201941139636819e-05, "loss": 0.3539, "step": 13906, "task_loss": 0.23713675141334534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31871840357780457, "epoch": 11.76, "learning_rate": 1.2016280525986224e-05, "loss": 0.3265, "step": 13907, "task_loss": 0.06253989040851593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5991958379745483, "epoch": 11.76, "learning_rate": 1.2013149655604258e-05, "loss": 0.4331, "step": 13908, "task_loss": 0.2855033874511719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38452762365341187, "epoch": 11.76, "learning_rate": 1.2010018785222293e-05, "loss": 0.4674, "step": 13909, "task_loss": 0.5138195753097534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6102333068847656, "epoch": 11.76, "learning_rate": 1.2006887914840326e-05, "loss": 0.6032, "step": 13910, "task_loss": 1.1887003183364868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5503163933753967, "epoch": 11.76, "learning_rate": 1.200375704445836e-05, "loss": 0.4585, "step": 13911, "task_loss": 1.5210247039794922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35403549671173096, "epoch": 11.76, "learning_rate": 1.2000626174076395e-05, "loss": 0.4718, "step": 13912, "task_loss": 0.19018006324768066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4327998757362366, "epoch": 11.76, "learning_rate": 1.1997495303694428e-05, "loss": 0.5517, "step": 13913, "task_loss": 0.6924914121627808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3092600405216217, "epoch": 11.76, "learning_rate": 1.1994364433312462e-05, "loss": 0.4704, "step": 13914, "task_loss": 0.2582423686981201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4062391221523285, "epoch": 11.76, "learning_rate": 1.1991233562930495e-05, "loss": 0.6338, "step": 13915, "task_loss": 0.8345068097114563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4299857020378113, "epoch": 11.76, "learning_rate": 1.1988102692548529e-05, "loss": 0.4095, "step": 13916, "task_loss": 0.38023754954338074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6186404824256897, "epoch": 11.76, "learning_rate": 1.1984971822166562e-05, "loss": 0.5427, "step": 13917, "task_loss": 0.8247162699699402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34560903906822205, "epoch": 11.76, "learning_rate": 1.1981840951784598e-05, "loss": 0.4784, "step": 13918, "task_loss": 0.8315048217773438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4292099177837372, "epoch": 11.77, "learning_rate": 1.1978710081402631e-05, "loss": 0.4671, "step": 13919, "task_loss": 0.3724275827407837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7199612259864807, "epoch": 11.77, "learning_rate": 1.1975579211020664e-05, "loss": 0.4681, "step": 13920, "task_loss": 0.7798829078674316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29945626854896545, "epoch": 11.77, "learning_rate": 1.1972448340638698e-05, "loss": 0.5669, "step": 13921, "task_loss": 0.689639687538147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30643463134765625, "epoch": 11.77, "learning_rate": 1.1969317470256731e-05, "loss": 0.4925, "step": 13922, "task_loss": 0.777299165725708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38155579566955566, "epoch": 11.77, "learning_rate": 1.1966186599874765e-05, "loss": 0.4769, "step": 13923, "task_loss": 0.9853657484054565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7773825526237488, "epoch": 11.77, "learning_rate": 1.1963055729492798e-05, "loss": 0.5649, "step": 13924, "task_loss": 1.4552476406097412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35829004645347595, "epoch": 11.77, "learning_rate": 1.1959924859110834e-05, "loss": 0.4218, "step": 13925, "task_loss": 0.3208982050418854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4433589279651642, "epoch": 11.77, "learning_rate": 1.1956793988728867e-05, "loss": 0.5135, "step": 13926, "task_loss": 1.1635481119155884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4143015444278717, "epoch": 11.77, "learning_rate": 1.1953663118346902e-05, "loss": 0.3739, "step": 13927, "task_loss": 0.26161107420921326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5448620319366455, "epoch": 11.77, "learning_rate": 1.1950532247964936e-05, "loss": 0.5655, "step": 13928, "task_loss": 0.6255455017089844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2765699326992035, "epoch": 11.77, "learning_rate": 1.1947401377582969e-05, "loss": 0.519, "step": 13929, "task_loss": 0.4527423083782196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44019198417663574, "epoch": 11.77, "learning_rate": 1.1944270507201003e-05, "loss": 0.527, "step": 13930, "task_loss": 1.0196959972381592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3673984706401825, "epoch": 11.78, "learning_rate": 1.1941139636819036e-05, "loss": 0.521, "step": 13931, "task_loss": 0.3832041025161743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48783382773399353, "epoch": 11.78, "learning_rate": 1.193800876643707e-05, "loss": 0.5036, "step": 13932, "task_loss": 0.3114665150642395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5148577690124512, "epoch": 11.78, "learning_rate": 1.1934877896055103e-05, "loss": 0.4795, "step": 13933, "task_loss": 0.27301663160324097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5588634610176086, "epoch": 11.78, "learning_rate": 1.1931747025673137e-05, "loss": 0.5924, "step": 13934, "task_loss": 0.6382935643196106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6912541389465332, "epoch": 11.78, "learning_rate": 1.1928616155291172e-05, "loss": 0.5509, "step": 13935, "task_loss": 1.2222201824188232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35264843702316284, "epoch": 11.78, "learning_rate": 1.1925485284909205e-05, "loss": 0.4359, "step": 13936, "task_loss": 0.6652204394340515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4856625497341156, "epoch": 11.78, "learning_rate": 1.192235441452724e-05, "loss": 0.5912, "step": 13937, "task_loss": 0.28888100385665894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2990935444831848, "epoch": 11.78, "learning_rate": 1.1919223544145274e-05, "loss": 0.4758, "step": 13938, "task_loss": 0.8539230227470398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31274494528770447, "epoch": 11.78, "learning_rate": 1.1916092673763307e-05, "loss": 0.43, "step": 13939, "task_loss": 0.9009300470352173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.59138423204422, "epoch": 11.78, "learning_rate": 1.191296180338134e-05, "loss": 0.3766, "step": 13940, "task_loss": 0.9478199481964111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.297981321811676, "epoch": 11.78, "learning_rate": 1.1909830932999374e-05, "loss": 0.437, "step": 13941, "task_loss": 0.4698110520839691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.307521253824234, "epoch": 11.78, "learning_rate": 1.1906700062617408e-05, "loss": 0.4254, "step": 13942, "task_loss": 1.1273869276046753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4318426847457886, "epoch": 11.79, "learning_rate": 1.1903569192235441e-05, "loss": 0.3699, "step": 13943, "task_loss": 0.46588990092277527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45096760988235474, "epoch": 11.79, "learning_rate": 1.1900438321853476e-05, "loss": 0.5934, "step": 13944, "task_loss": 1.568630337715149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6491484045982361, "epoch": 11.79, "learning_rate": 1.189730745147151e-05, "loss": 0.5202, "step": 13945, "task_loss": 0.9707975387573242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2859615683555603, "epoch": 11.79, "learning_rate": 1.1894176581089543e-05, "loss": 0.3647, "step": 13946, "task_loss": 1.473409652709961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7490450143814087, "epoch": 11.79, "learning_rate": 1.1891045710707577e-05, "loss": 0.6943, "step": 13947, "task_loss": 0.7653852105140686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45613202452659607, "epoch": 11.79, "learning_rate": 1.188791484032561e-05, "loss": 0.4883, "step": 13948, "task_loss": 1.1889101266860962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27323949337005615, "epoch": 11.79, "learning_rate": 1.1884783969943645e-05, "loss": 0.3905, "step": 13949, "task_loss": 0.4686676859855652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3274701237678528, "epoch": 11.79, "learning_rate": 1.1881653099561679e-05, "loss": 0.532, "step": 13950, "task_loss": 0.3747076094150543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7757526636123657, "epoch": 11.79, "learning_rate": 1.1878522229179712e-05, "loss": 0.6148, "step": 13951, "task_loss": 0.7370638847351074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5251644849777222, "epoch": 11.79, "learning_rate": 1.1875391358797746e-05, "loss": 0.6205, "step": 13952, "task_loss": 0.9291985034942627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24128352105617523, "epoch": 11.79, "learning_rate": 1.1872260488415781e-05, "loss": 0.4322, "step": 13953, "task_loss": 0.03339387848973274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3582671582698822, "epoch": 11.79, "learning_rate": 1.1869129618033814e-05, "loss": 0.4241, "step": 13954, "task_loss": 0.15625078976154327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41452616453170776, "epoch": 11.8, "learning_rate": 1.1865998747651848e-05, "loss": 0.4432, "step": 13955, "task_loss": 0.4338703751564026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42857441306114197, "epoch": 11.8, "learning_rate": 1.1862867877269881e-05, "loss": 0.5498, "step": 13956, "task_loss": 0.6705812811851501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5009704828262329, "epoch": 11.8, "learning_rate": 1.1859737006887915e-05, "loss": 0.456, "step": 13957, "task_loss": 0.4653222858905792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25550609827041626, "epoch": 11.8, "learning_rate": 1.1856606136505948e-05, "loss": 0.3472, "step": 13958, "task_loss": 0.07959931343793869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32638683915138245, "epoch": 11.8, "learning_rate": 1.1853475266123982e-05, "loss": 0.4134, "step": 13959, "task_loss": 0.5895489454269409 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.722497820854187, "epoch": 11.8, "learning_rate": 1.1850344395742017e-05, "loss": 0.5042, "step": 13960, "task_loss": 1.0834909677505493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28093060851097107, "epoch": 11.8, "learning_rate": 1.184721352536005e-05, "loss": 0.3939, "step": 13961, "task_loss": 0.6349290013313293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49037885665893555, "epoch": 11.8, "learning_rate": 1.1844082654978086e-05, "loss": 0.4823, "step": 13962, "task_loss": 0.9877067804336548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24340464174747467, "epoch": 11.8, "learning_rate": 1.1840951784596119e-05, "loss": 0.3907, "step": 13963, "task_loss": 0.6519643664360046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3399897813796997, "epoch": 11.8, "learning_rate": 1.1837820914214153e-05, "loss": 0.5869, "step": 13964, "task_loss": 1.7274887561798096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5704013109207153, "epoch": 11.8, "learning_rate": 1.1834690043832186e-05, "loss": 0.6337, "step": 13965, "task_loss": 0.2406664490699768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4421725869178772, "epoch": 11.81, "learning_rate": 1.183155917345022e-05, "loss": 0.4998, "step": 13966, "task_loss": 0.6302717328071594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30418914556503296, "epoch": 11.81, "learning_rate": 1.1828428303068253e-05, "loss": 0.4325, "step": 13967, "task_loss": 0.6394633650779724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40227261185646057, "epoch": 11.81, "learning_rate": 1.1825297432686287e-05, "loss": 0.484, "step": 13968, "task_loss": 0.469608336687088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6458826065063477, "epoch": 11.81, "learning_rate": 1.1822166562304322e-05, "loss": 0.4375, "step": 13969, "task_loss": 0.9548981785774231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20289722084999084, "epoch": 11.81, "learning_rate": 1.1819035691922355e-05, "loss": 0.3956, "step": 13970, "task_loss": 0.036941055208444595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3144903779029846, "epoch": 11.81, "learning_rate": 1.1815904821540389e-05, "loss": 0.4563, "step": 13971, "task_loss": 1.0453948974609375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4973679482936859, "epoch": 11.81, "learning_rate": 1.1812773951158422e-05, "loss": 0.4547, "step": 13972, "task_loss": 0.4114377796649933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38808587193489075, "epoch": 11.81, "learning_rate": 1.1809643080776456e-05, "loss": 0.362, "step": 13973, "task_loss": 0.16717009246349335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31845739483833313, "epoch": 11.81, "learning_rate": 1.180651221039449e-05, "loss": 0.5585, "step": 13974, "task_loss": 0.37266355752944946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4974393844604492, "epoch": 11.81, "learning_rate": 1.1803381340012524e-05, "loss": 0.4254, "step": 13975, "task_loss": 0.7900385856628418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5425271391868591, "epoch": 11.81, "learning_rate": 1.1800250469630558e-05, "loss": 0.4985, "step": 13976, "task_loss": 1.0963484048843384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31990891695022583, "epoch": 11.81, "learning_rate": 1.1797119599248591e-05, "loss": 0.395, "step": 13977, "task_loss": 0.3066284656524658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45746299624443054, "epoch": 11.82, "learning_rate": 1.1793988728866626e-05, "loss": 0.5664, "step": 13978, "task_loss": 0.5430468916893005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4387495517730713, "epoch": 11.82, "learning_rate": 1.179085785848466e-05, "loss": 0.4245, "step": 13979, "task_loss": 0.4662114679813385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24414831399917603, "epoch": 11.82, "learning_rate": 1.1787726988102693e-05, "loss": 0.5182, "step": 13980, "task_loss": 0.10932820290327072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6469959020614624, "epoch": 11.82, "learning_rate": 1.1784596117720727e-05, "loss": 0.6643, "step": 13981, "task_loss": 0.5737144351005554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3455977141857147, "epoch": 11.82, "learning_rate": 1.178146524733876e-05, "loss": 0.3911, "step": 13982, "task_loss": 0.6300783157348633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6631691455841064, "epoch": 11.82, "learning_rate": 1.1778334376956794e-05, "loss": 0.5564, "step": 13983, "task_loss": 0.6978567838668823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.795154333114624, "epoch": 11.82, "learning_rate": 1.1775203506574827e-05, "loss": 0.4698, "step": 13984, "task_loss": 0.9801595211029053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5166429281234741, "epoch": 11.82, "learning_rate": 1.1772072636192862e-05, "loss": 0.4414, "step": 13985, "task_loss": 0.40820232033729553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4005280137062073, "epoch": 11.82, "learning_rate": 1.1768941765810896e-05, "loss": 0.5462, "step": 13986, "task_loss": 0.6596273183822632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15090125799179077, "epoch": 11.82, "learning_rate": 1.1765810895428931e-05, "loss": 0.3371, "step": 13987, "task_loss": 0.14372165501117706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45632338523864746, "epoch": 11.82, "learning_rate": 1.1762680025046964e-05, "loss": 0.4961, "step": 13988, "task_loss": 0.40311405062675476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36816900968551636, "epoch": 11.82, "learning_rate": 1.1759549154664998e-05, "loss": 0.4057, "step": 13989, "task_loss": 1.3861839771270752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4582993984222412, "epoch": 11.83, "learning_rate": 1.1756418284283031e-05, "loss": 0.5795, "step": 13990, "task_loss": 0.6080071926116943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2042461633682251, "epoch": 11.83, "learning_rate": 1.1753287413901065e-05, "loss": 0.336, "step": 13991, "task_loss": 0.11398528516292572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36488282680511475, "epoch": 11.83, "learning_rate": 1.1750156543519098e-05, "loss": 0.3687, "step": 13992, "task_loss": 1.0075286626815796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2521618604660034, "epoch": 11.83, "learning_rate": 1.1747025673137132e-05, "loss": 0.4582, "step": 13993, "task_loss": 0.48524606227874756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.51270592212677, "epoch": 11.83, "learning_rate": 1.1743894802755167e-05, "loss": 0.5503, "step": 13994, "task_loss": 0.7299681901931763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26144111156463623, "epoch": 11.83, "learning_rate": 1.17407639323732e-05, "loss": 0.3119, "step": 13995, "task_loss": 0.1628972589969635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5413733124732971, "epoch": 11.83, "learning_rate": 1.1737633061991234e-05, "loss": 0.625, "step": 13996, "task_loss": 1.1651688814163208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31762173771858215, "epoch": 11.83, "learning_rate": 1.1734502191609267e-05, "loss": 0.6004, "step": 13997, "task_loss": 0.17223963141441345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3131852447986603, "epoch": 11.83, "learning_rate": 1.1731371321227303e-05, "loss": 0.3852, "step": 13998, "task_loss": 0.30565887689590454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3040185272693634, "epoch": 11.83, "learning_rate": 1.1728240450845336e-05, "loss": 0.4121, "step": 13999, "task_loss": 0.6262569427490234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49200451374053955, "epoch": 11.83, "learning_rate": 1.172510958046337e-05, "loss": 0.6012, "step": 14000, "task_loss": 1.0382100343704224 }, { "epoch": 11.83, "eval_accuracy": 0.9097821782178218, "eval_loss": 0.3291562497615814, "eval_runtime": 207.2706, "eval_samples_per_second": 121.821, "eval_steps_per_second": 0.955, "step": 14000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36988604068756104, "epoch": 11.83, "learning_rate": 1.1721978710081403e-05, "loss": 0.4126, "step": 14001, "task_loss": 0.4264497458934784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5047324299812317, "epoch": 11.84, "learning_rate": 1.1718847839699437e-05, "loss": 0.539, "step": 14002, "task_loss": 0.6658588647842407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7970152497291565, "epoch": 11.84, "learning_rate": 1.1715716969317472e-05, "loss": 0.6447, "step": 14003, "task_loss": 1.0266387462615967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36648714542388916, "epoch": 11.84, "learning_rate": 1.1712586098935505e-05, "loss": 0.4795, "step": 14004, "task_loss": 0.5952311158180237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3915730118751526, "epoch": 11.84, "learning_rate": 1.1709455228553539e-05, "loss": 0.4973, "step": 14005, "task_loss": 1.0708293914794922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3757440149784088, "epoch": 11.84, "learning_rate": 1.1706324358171572e-05, "loss": 0.3742, "step": 14006, "task_loss": 0.2606291174888611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48679596185684204, "epoch": 11.84, "learning_rate": 1.1703193487789606e-05, "loss": 0.4822, "step": 14007, "task_loss": 0.5096094012260437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5356853008270264, "epoch": 11.84, "learning_rate": 1.1700062617407639e-05, "loss": 0.5712, "step": 14008, "task_loss": 0.7086843252182007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22284510731697083, "epoch": 11.84, "learning_rate": 1.1696931747025673e-05, "loss": 0.2887, "step": 14009, "task_loss": 0.2751060128211975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4761136770248413, "epoch": 11.84, "learning_rate": 1.1693800876643706e-05, "loss": 0.3357, "step": 14010, "task_loss": 0.4624575078487396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7031955122947693, "epoch": 11.84, "learning_rate": 1.1690670006261741e-05, "loss": 0.5791, "step": 14011, "task_loss": 0.5384601354598999 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5501011610031128, "epoch": 11.84, "learning_rate": 1.1687539135879776e-05, "loss": 0.5065, "step": 14012, "task_loss": 0.2626993656158447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2349933683872223, "epoch": 11.84, "learning_rate": 1.168440826549781e-05, "loss": 0.3209, "step": 14013, "task_loss": 0.056062664836645126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45518815517425537, "epoch": 11.85, "learning_rate": 1.1681277395115843e-05, "loss": 0.393, "step": 14014, "task_loss": 0.31936323642730713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4743463397026062, "epoch": 11.85, "learning_rate": 1.1678146524733877e-05, "loss": 0.4533, "step": 14015, "task_loss": 0.3311441242694855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3786928057670593, "epoch": 11.85, "learning_rate": 1.167501565435191e-05, "loss": 0.5042, "step": 14016, "task_loss": 0.46500298380851746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.373269259929657, "epoch": 11.85, "learning_rate": 1.1671884783969944e-05, "loss": 0.4352, "step": 14017, "task_loss": 0.499997079372406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2642863988876343, "epoch": 11.85, "learning_rate": 1.1668753913587977e-05, "loss": 0.4346, "step": 14018, "task_loss": 0.039312850683927536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2834063768386841, "epoch": 11.85, "learning_rate": 1.166562304320601e-05, "loss": 0.44, "step": 14019, "task_loss": 0.433743417263031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2561429738998413, "epoch": 11.85, "learning_rate": 1.1662492172824046e-05, "loss": 0.5052, "step": 14020, "task_loss": 0.23510313034057617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6617315411567688, "epoch": 11.85, "learning_rate": 1.165936130244208e-05, "loss": 0.6123, "step": 14021, "task_loss": 1.214451551437378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5458118319511414, "epoch": 11.85, "learning_rate": 1.1656230432060113e-05, "loss": 0.494, "step": 14022, "task_loss": 0.5987029075622559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4722503423690796, "epoch": 11.85, "learning_rate": 1.1653099561678148e-05, "loss": 0.6452, "step": 14023, "task_loss": 0.6522718667984009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18496039509773254, "epoch": 11.85, "learning_rate": 1.1649968691296181e-05, "loss": 0.4464, "step": 14024, "task_loss": 0.6499797105789185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3858136534690857, "epoch": 11.85, "learning_rate": 1.1646837820914215e-05, "loss": 0.4073, "step": 14025, "task_loss": 0.9590620994567871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.650253176689148, "epoch": 11.86, "learning_rate": 1.1643706950532248e-05, "loss": 0.5611, "step": 14026, "task_loss": 1.1371558904647827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49508360028266907, "epoch": 11.86, "learning_rate": 1.1640576080150282e-05, "loss": 0.4656, "step": 14027, "task_loss": 0.9843572974205017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38444602489471436, "epoch": 11.86, "learning_rate": 1.1637445209768315e-05, "loss": 0.4159, "step": 14028, "task_loss": 0.42650097608566284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4180159568786621, "epoch": 11.86, "learning_rate": 1.163431433938635e-05, "loss": 0.5403, "step": 14029, "task_loss": 0.4297528564929962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4049866795539856, "epoch": 11.86, "learning_rate": 1.1631183469004384e-05, "loss": 0.3773, "step": 14030, "task_loss": 0.1919286698102951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.856109619140625, "epoch": 11.86, "learning_rate": 1.1628052598622417e-05, "loss": 0.4593, "step": 14031, "task_loss": 0.9286243915557861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3103421926498413, "epoch": 11.86, "learning_rate": 1.1624921728240451e-05, "loss": 0.4097, "step": 14032, "task_loss": 0.4595620036125183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28251123428344727, "epoch": 11.86, "learning_rate": 1.1621790857858484e-05, "loss": 0.3904, "step": 14033, "task_loss": 0.3389819264411926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5093796253204346, "epoch": 11.86, "learning_rate": 1.1618659987476518e-05, "loss": 0.4014, "step": 14034, "task_loss": 0.5741698741912842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49712705612182617, "epoch": 11.86, "learning_rate": 1.1615529117094553e-05, "loss": 0.4912, "step": 14035, "task_loss": 1.1505131721496582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4721943438053131, "epoch": 11.86, "learning_rate": 1.1612398246712587e-05, "loss": 0.4194, "step": 14036, "task_loss": 0.9146241545677185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6987203359603882, "epoch": 11.87, "learning_rate": 1.1609267376330622e-05, "loss": 0.5273, "step": 14037, "task_loss": 1.193362832069397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.273027241230011, "epoch": 11.87, "learning_rate": 1.1606136505948655e-05, "loss": 0.4405, "step": 14038, "task_loss": 0.42888548970222473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4962064027786255, "epoch": 11.87, "learning_rate": 1.1603005635566689e-05, "loss": 0.3191, "step": 14039, "task_loss": 0.9126311540603638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3754996061325073, "epoch": 11.87, "learning_rate": 1.1599874765184722e-05, "loss": 0.542, "step": 14040, "task_loss": 0.18579426407814026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8182848691940308, "epoch": 11.87, "learning_rate": 1.1596743894802756e-05, "loss": 0.5494, "step": 14041, "task_loss": 0.8670841455459595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45521676540374756, "epoch": 11.87, "learning_rate": 1.1593613024420789e-05, "loss": 0.5056, "step": 14042, "task_loss": 0.6519986391067505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4083189368247986, "epoch": 11.87, "learning_rate": 1.1590482154038823e-05, "loss": 0.4148, "step": 14043, "task_loss": 0.659054160118103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5607189536094666, "epoch": 11.87, "learning_rate": 1.1587351283656856e-05, "loss": 0.5363, "step": 14044, "task_loss": 0.6876751184463501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8408393263816833, "epoch": 11.87, "learning_rate": 1.1584220413274891e-05, "loss": 0.5465, "step": 14045, "task_loss": 1.0970438718795776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3333359956741333, "epoch": 11.87, "learning_rate": 1.1581089542892925e-05, "loss": 0.4971, "step": 14046, "task_loss": 0.8457790017127991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3314821124076843, "epoch": 11.87, "learning_rate": 1.157795867251096e-05, "loss": 0.4817, "step": 14047, "task_loss": 0.41477149724960327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7074357271194458, "epoch": 11.87, "learning_rate": 1.1574827802128993e-05, "loss": 0.4894, "step": 14048, "task_loss": 0.7824748158454895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4945370554924011, "epoch": 11.88, "learning_rate": 1.1571696931747027e-05, "loss": 0.4537, "step": 14049, "task_loss": 0.7856977581977844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4940529465675354, "epoch": 11.88, "learning_rate": 1.156856606136506e-05, "loss": 0.5635, "step": 14050, "task_loss": 0.4830300807952881 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26314932107925415, "epoch": 11.88, "learning_rate": 1.1565435190983094e-05, "loss": 0.4782, "step": 14051, "task_loss": 0.41673988103866577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38853520154953003, "epoch": 11.88, "learning_rate": 1.1562304320601127e-05, "loss": 0.4645, "step": 14052, "task_loss": 0.9197804927825928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6032255291938782, "epoch": 11.88, "learning_rate": 1.155917345021916e-05, "loss": 0.4814, "step": 14053, "task_loss": 0.8620644807815552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45145073533058167, "epoch": 11.88, "learning_rate": 1.1556042579837196e-05, "loss": 0.4785, "step": 14054, "task_loss": 0.8316320776939392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33597099781036377, "epoch": 11.88, "learning_rate": 1.155291170945523e-05, "loss": 0.6109, "step": 14055, "task_loss": 1.0436949729919434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45589157938957214, "epoch": 11.88, "learning_rate": 1.1549780839073263e-05, "loss": 0.4049, "step": 14056, "task_loss": 1.1274335384368896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6043757796287537, "epoch": 11.88, "learning_rate": 1.1546649968691296e-05, "loss": 0.5285, "step": 14057, "task_loss": 0.8345929980278015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4830585718154907, "epoch": 11.88, "learning_rate": 1.154351909830933e-05, "loss": 0.593, "step": 14058, "task_loss": 0.5890297889709473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5638519525527954, "epoch": 11.88, "learning_rate": 1.1540388227927363e-05, "loss": 0.4838, "step": 14059, "task_loss": 0.3602108359336853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7073821425437927, "epoch": 11.88, "learning_rate": 1.1537257357545398e-05, "loss": 0.6065, "step": 14060, "task_loss": 0.7667819857597351 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.511428952217102, "epoch": 11.89, "learning_rate": 1.1534126487163432e-05, "loss": 0.4431, "step": 14061, "task_loss": 0.7173548340797424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41919466853141785, "epoch": 11.89, "learning_rate": 1.1530995616781465e-05, "loss": 0.3824, "step": 14062, "task_loss": 0.5416041612625122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4081040024757385, "epoch": 11.89, "learning_rate": 1.15278647463995e-05, "loss": 0.4347, "step": 14063, "task_loss": 0.32869601249694824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6633526086807251, "epoch": 11.89, "learning_rate": 1.1524733876017534e-05, "loss": 0.4353, "step": 14064, "task_loss": 1.0592859983444214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3784102201461792, "epoch": 11.89, "learning_rate": 1.1521603005635568e-05, "loss": 0.4879, "step": 14065, "task_loss": 1.0409761667251587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5289937257766724, "epoch": 11.89, "learning_rate": 1.1518472135253601e-05, "loss": 0.4288, "step": 14066, "task_loss": 0.5802410244941711 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21067865192890167, "epoch": 11.89, "learning_rate": 1.1515341264871634e-05, "loss": 0.4978, "step": 14067, "task_loss": 0.04271572455763817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5309053659439087, "epoch": 11.89, "learning_rate": 1.1512210394489668e-05, "loss": 0.4668, "step": 14068, "task_loss": 0.8288323283195496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7960895299911499, "epoch": 11.89, "learning_rate": 1.1509079524107701e-05, "loss": 0.6668, "step": 14069, "task_loss": 1.0565202236175537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6170222163200378, "epoch": 11.89, "learning_rate": 1.1505948653725737e-05, "loss": 0.4917, "step": 14070, "task_loss": 0.31876814365386963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5530741214752197, "epoch": 11.89, "learning_rate": 1.150281778334377e-05, "loss": 0.4669, "step": 14071, "task_loss": 0.5652998089790344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5675452351570129, "epoch": 11.89, "learning_rate": 1.1499686912961805e-05, "loss": 0.4378, "step": 14072, "task_loss": 0.9236382842063904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6296908259391785, "epoch": 11.9, "learning_rate": 1.1496556042579839e-05, "loss": 0.501, "step": 14073, "task_loss": 1.4537047147750854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5580959916114807, "epoch": 11.9, "learning_rate": 1.1493425172197872e-05, "loss": 0.5425, "step": 14074, "task_loss": 0.7652212977409363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2111906111240387, "epoch": 11.9, "learning_rate": 1.1490294301815906e-05, "loss": 0.3853, "step": 14075, "task_loss": 0.4568442106246948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5270494222640991, "epoch": 11.9, "learning_rate": 1.1487163431433939e-05, "loss": 0.6097, "step": 14076, "task_loss": 1.0886327028274536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40952521562576294, "epoch": 11.9, "learning_rate": 1.1484032561051973e-05, "loss": 0.4987, "step": 14077, "task_loss": 0.7285313606262207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5464845895767212, "epoch": 11.9, "learning_rate": 1.1480901690670006e-05, "loss": 0.5292, "step": 14078, "task_loss": 1.3797963857650757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7627694606781006, "epoch": 11.9, "learning_rate": 1.1477770820288041e-05, "loss": 0.6032, "step": 14079, "task_loss": 0.8047756552696228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7051277160644531, "epoch": 11.9, "learning_rate": 1.1474639949906075e-05, "loss": 0.6448, "step": 14080, "task_loss": 0.7615541815757751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5622076392173767, "epoch": 11.9, "learning_rate": 1.1471509079524108e-05, "loss": 0.4618, "step": 14081, "task_loss": 0.3282923102378845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6902161836624146, "epoch": 11.9, "learning_rate": 1.1468378209142142e-05, "loss": 0.5626, "step": 14082, "task_loss": 0.5215178728103638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6048520803451538, "epoch": 11.9, "learning_rate": 1.1465247338760175e-05, "loss": 0.444, "step": 14083, "task_loss": 0.7467309236526489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5383458137512207, "epoch": 11.9, "learning_rate": 1.146211646837821e-05, "loss": 0.5181, "step": 14084, "task_loss": 0.11858366429805756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2856748402118683, "epoch": 11.91, "learning_rate": 1.1458985597996244e-05, "loss": 0.3507, "step": 14085, "task_loss": 1.2130849361419678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5324110984802246, "epoch": 11.91, "learning_rate": 1.1455854727614277e-05, "loss": 0.5735, "step": 14086, "task_loss": 0.4011542499065399 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38787928223609924, "epoch": 11.91, "learning_rate": 1.145272385723231e-05, "loss": 0.5037, "step": 14087, "task_loss": 1.2993351221084595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6229820847511292, "epoch": 11.91, "learning_rate": 1.1449592986850346e-05, "loss": 0.5853, "step": 14088, "task_loss": 0.8061084747314453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2606625556945801, "epoch": 11.91, "learning_rate": 1.144646211646838e-05, "loss": 0.4773, "step": 14089, "task_loss": 0.07424473762512207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42359814047813416, "epoch": 11.91, "learning_rate": 1.1443331246086413e-05, "loss": 0.501, "step": 14090, "task_loss": 0.13800688087940216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3048666715621948, "epoch": 11.91, "learning_rate": 1.1440200375704446e-05, "loss": 0.3337, "step": 14091, "task_loss": 0.8555279970169067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46800732612609863, "epoch": 11.91, "learning_rate": 1.143706950532248e-05, "loss": 0.5763, "step": 14092, "task_loss": 1.6677497625350952 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5930626392364502, "epoch": 11.91, "learning_rate": 1.1433938634940513e-05, "loss": 0.4529, "step": 14093, "task_loss": 1.2000788450241089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33532434701919556, "epoch": 11.91, "learning_rate": 1.1430807764558547e-05, "loss": 0.4408, "step": 14094, "task_loss": 0.42725637555122375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2988957166671753, "epoch": 11.91, "learning_rate": 1.142767689417658e-05, "loss": 0.4382, "step": 14095, "task_loss": 0.18539078533649445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5585975050926208, "epoch": 11.91, "learning_rate": 1.1424546023794615e-05, "loss": 0.4056, "step": 14096, "task_loss": 0.6409057378768921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7001733779907227, "epoch": 11.92, "learning_rate": 1.142141515341265e-05, "loss": 0.7904, "step": 14097, "task_loss": 0.5791674256324768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5364475846290588, "epoch": 11.92, "learning_rate": 1.1418284283030684e-05, "loss": 0.4435, "step": 14098, "task_loss": 0.7109691500663757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6491963863372803, "epoch": 11.92, "learning_rate": 1.1415153412648718e-05, "loss": 0.539, "step": 14099, "task_loss": 0.7367326021194458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5911850333213806, "epoch": 11.92, "learning_rate": 1.1412022542266751e-05, "loss": 0.406, "step": 14100, "task_loss": 0.5156306028366089 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49337947368621826, "epoch": 11.92, "learning_rate": 1.1408891671884784e-05, "loss": 0.4616, "step": 14101, "task_loss": 0.7487035989761353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43597257137298584, "epoch": 11.92, "learning_rate": 1.1405760801502818e-05, "loss": 0.3469, "step": 14102, "task_loss": 0.31235241889953613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.330547958612442, "epoch": 11.92, "learning_rate": 1.1402629931120851e-05, "loss": 0.3958, "step": 14103, "task_loss": 1.0005849599838257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3485320806503296, "epoch": 11.92, "learning_rate": 1.1399499060738885e-05, "loss": 0.4812, "step": 14104, "task_loss": 0.5088470578193665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2830621898174286, "epoch": 11.92, "learning_rate": 1.139636819035692e-05, "loss": 0.4577, "step": 14105, "task_loss": 0.2867435812950134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5078609585762024, "epoch": 11.92, "learning_rate": 1.1393237319974954e-05, "loss": 0.5068, "step": 14106, "task_loss": 0.38532981276512146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5647274255752563, "epoch": 11.92, "learning_rate": 1.1390106449592987e-05, "loss": 0.4902, "step": 14107, "task_loss": 0.5578709244728088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5091791749000549, "epoch": 11.93, "learning_rate": 1.138697557921102e-05, "loss": 0.5267, "step": 14108, "task_loss": 0.8869589567184448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.685789942741394, "epoch": 11.93, "learning_rate": 1.1383844708829056e-05, "loss": 0.438, "step": 14109, "task_loss": 0.31375905871391296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42736518383026123, "epoch": 11.93, "learning_rate": 1.1380713838447089e-05, "loss": 0.3974, "step": 14110, "task_loss": 0.5390063524246216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4880772829055786, "epoch": 11.93, "learning_rate": 1.1377582968065123e-05, "loss": 0.4982, "step": 14111, "task_loss": 0.2691528797149658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5196737051010132, "epoch": 11.93, "learning_rate": 1.1374452097683156e-05, "loss": 0.3591, "step": 14112, "task_loss": 0.31510722637176514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29379692673683167, "epoch": 11.93, "learning_rate": 1.1371321227301191e-05, "loss": 0.4442, "step": 14113, "task_loss": 0.6849684715270996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4686240255832672, "epoch": 11.93, "learning_rate": 1.1368190356919225e-05, "loss": 0.5037, "step": 14114, "task_loss": 0.39656272530555725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28596025705337524, "epoch": 11.93, "learning_rate": 1.1365059486537258e-05, "loss": 0.4516, "step": 14115, "task_loss": 0.6415674686431885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3501743674278259, "epoch": 11.93, "learning_rate": 1.1361928616155292e-05, "loss": 0.4452, "step": 14116, "task_loss": 0.11694861948490143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34857651591300964, "epoch": 11.93, "learning_rate": 1.1358797745773325e-05, "loss": 0.6226, "step": 14117, "task_loss": 0.08919744938611984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.654544472694397, "epoch": 11.93, "learning_rate": 1.1355666875391359e-05, "loss": 0.5002, "step": 14118, "task_loss": 1.076307773590088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4778223931789398, "epoch": 11.93, "learning_rate": 1.1352536005009392e-05, "loss": 0.4477, "step": 14119, "task_loss": 0.483521431684494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3070007562637329, "epoch": 11.94, "learning_rate": 1.1349405134627426e-05, "loss": 0.425, "step": 14120, "task_loss": 0.5958427786827087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4029087424278259, "epoch": 11.94, "learning_rate": 1.134627426424546e-05, "loss": 0.3786, "step": 14121, "task_loss": 0.1012999415397644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21922729909420013, "epoch": 11.94, "learning_rate": 1.1343143393863496e-05, "loss": 0.3873, "step": 14122, "task_loss": 0.44507160782814026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48073649406433105, "epoch": 11.94, "learning_rate": 1.134001252348153e-05, "loss": 0.4792, "step": 14123, "task_loss": 0.6179108619689941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6126859188079834, "epoch": 11.94, "learning_rate": 1.1336881653099563e-05, "loss": 0.5577, "step": 14124, "task_loss": 0.40132203698158264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5327641367912292, "epoch": 11.94, "learning_rate": 1.1333750782717596e-05, "loss": 0.5641, "step": 14125, "task_loss": 0.5882784128189087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4086203873157501, "epoch": 11.94, "learning_rate": 1.133061991233563e-05, "loss": 0.3972, "step": 14126, "task_loss": 0.19627012312412262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34759262204170227, "epoch": 11.94, "learning_rate": 1.1327489041953663e-05, "loss": 0.4045, "step": 14127, "task_loss": 0.5331326127052307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4908328056335449, "epoch": 11.94, "learning_rate": 1.1324358171571697e-05, "loss": 0.4977, "step": 14128, "task_loss": 1.0184108018875122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8306559324264526, "epoch": 11.94, "learning_rate": 1.132122730118973e-05, "loss": 0.5929, "step": 14129, "task_loss": 0.6208680272102356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3476760685443878, "epoch": 11.94, "learning_rate": 1.1318096430807765e-05, "loss": 0.4368, "step": 14130, "task_loss": 0.6090332865715027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24943187832832336, "epoch": 11.94, "learning_rate": 1.1314965560425799e-05, "loss": 0.357, "step": 14131, "task_loss": 0.1685677319765091 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5146100521087646, "epoch": 11.95, "learning_rate": 1.1311834690043832e-05, "loss": 0.3426, "step": 14132, "task_loss": 0.7761980295181274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30009204149246216, "epoch": 11.95, "learning_rate": 1.1308703819661868e-05, "loss": 0.5083, "step": 14133, "task_loss": 0.5420651435852051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6317131519317627, "epoch": 11.95, "learning_rate": 1.1305572949279901e-05, "loss": 0.5305, "step": 14134, "task_loss": 0.22156094014644623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6958804726600647, "epoch": 11.95, "learning_rate": 1.1302442078897934e-05, "loss": 0.4744, "step": 14135, "task_loss": 0.6222507953643799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3346426486968994, "epoch": 11.95, "learning_rate": 1.1299311208515968e-05, "loss": 0.4191, "step": 14136, "task_loss": 0.6121584177017212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3797474503517151, "epoch": 11.95, "learning_rate": 1.1296180338134001e-05, "loss": 0.4797, "step": 14137, "task_loss": 0.367134690284729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.551867663860321, "epoch": 11.95, "learning_rate": 1.1293049467752035e-05, "loss": 0.5414, "step": 14138, "task_loss": 0.9826722741127014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45689529180526733, "epoch": 11.95, "learning_rate": 1.128991859737007e-05, "loss": 0.4077, "step": 14139, "task_loss": 0.5176029205322266 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5585969686508179, "epoch": 11.95, "learning_rate": 1.1286787726988104e-05, "loss": 0.4843, "step": 14140, "task_loss": 1.7756552696228027 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5632709860801697, "epoch": 11.95, "learning_rate": 1.1283656856606137e-05, "loss": 0.4686, "step": 14141, "task_loss": 0.26255109906196594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.437765896320343, "epoch": 11.95, "learning_rate": 1.128052598622417e-05, "loss": 0.6168, "step": 14142, "task_loss": 0.528969943523407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41433876752853394, "epoch": 11.95, "learning_rate": 1.1277395115842204e-05, "loss": 0.4801, "step": 14143, "task_loss": 0.34797897934913635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5485647916793823, "epoch": 11.96, "learning_rate": 1.1274264245460237e-05, "loss": 0.4473, "step": 14144, "task_loss": 0.40979883074760437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5869049429893494, "epoch": 11.96, "learning_rate": 1.1271133375078271e-05, "loss": 0.4179, "step": 14145, "task_loss": 0.6739272475242615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5571236610412598, "epoch": 11.96, "learning_rate": 1.1268002504696306e-05, "loss": 0.5035, "step": 14146, "task_loss": 0.40978139638900757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30242735147476196, "epoch": 11.96, "learning_rate": 1.126487163431434e-05, "loss": 0.3561, "step": 14147, "task_loss": 0.16240473091602325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5501389503479004, "epoch": 11.96, "learning_rate": 1.1261740763932375e-05, "loss": 0.4749, "step": 14148, "task_loss": 1.0502768754959106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4084145426750183, "epoch": 11.96, "learning_rate": 1.1258609893550408e-05, "loss": 0.3452, "step": 14149, "task_loss": 0.4511638283729553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3953613042831421, "epoch": 11.96, "learning_rate": 1.1255479023168442e-05, "loss": 0.384, "step": 14150, "task_loss": 1.201457142829895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2616533637046814, "epoch": 11.96, "learning_rate": 1.1252348152786475e-05, "loss": 0.3595, "step": 14151, "task_loss": 0.13949108123779297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5346006751060486, "epoch": 11.96, "learning_rate": 1.1249217282404509e-05, "loss": 0.5591, "step": 14152, "task_loss": 0.9566660523414612 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0577802658081055, "epoch": 11.96, "learning_rate": 1.1246086412022542e-05, "loss": 0.6228, "step": 14153, "task_loss": 0.6702136993408203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4979538917541504, "epoch": 11.96, "learning_rate": 1.1242955541640576e-05, "loss": 0.5088, "step": 14154, "task_loss": 0.6704326868057251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4179975390434265, "epoch": 11.96, "learning_rate": 1.123982467125861e-05, "loss": 0.5259, "step": 14155, "task_loss": 0.23027609288692474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5619025230407715, "epoch": 11.97, "learning_rate": 1.1236693800876644e-05, "loss": 0.4937, "step": 14156, "task_loss": 0.7528891563415527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8830852508544922, "epoch": 11.97, "learning_rate": 1.1233562930494678e-05, "loss": 0.6394, "step": 14157, "task_loss": 0.7761714458465576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49472489953041077, "epoch": 11.97, "learning_rate": 1.1230432060112713e-05, "loss": 0.4333, "step": 14158, "task_loss": 0.402646005153656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34189867973327637, "epoch": 11.97, "learning_rate": 1.1227301189730746e-05, "loss": 0.4843, "step": 14159, "task_loss": 0.8466321229934692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36767059564590454, "epoch": 11.97, "learning_rate": 1.122417031934878e-05, "loss": 0.3864, "step": 14160, "task_loss": 0.2211342751979828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5131309032440186, "epoch": 11.97, "learning_rate": 1.1221039448966813e-05, "loss": 0.4377, "step": 14161, "task_loss": 1.5922114849090576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3733234703540802, "epoch": 11.97, "learning_rate": 1.1217908578584847e-05, "loss": 0.5026, "step": 14162, "task_loss": 0.6322461366653442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8497600555419922, "epoch": 11.97, "learning_rate": 1.121477770820288e-05, "loss": 0.4805, "step": 14163, "task_loss": 0.6220279335975647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33336928486824036, "epoch": 11.97, "learning_rate": 1.1211646837820915e-05, "loss": 0.4357, "step": 14164, "task_loss": 1.4859349727630615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7664539813995361, "epoch": 11.97, "learning_rate": 1.1208515967438949e-05, "loss": 0.627, "step": 14165, "task_loss": 0.5557132959365845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37607890367507935, "epoch": 11.97, "learning_rate": 1.1205385097056982e-05, "loss": 0.5064, "step": 14166, "task_loss": 1.0498559474945068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.501200258731842, "epoch": 11.97, "learning_rate": 1.1202254226675016e-05, "loss": 0.3752, "step": 14167, "task_loss": 0.5042843818664551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2790474593639374, "epoch": 11.98, "learning_rate": 1.119912335629305e-05, "loss": 0.3516, "step": 14168, "task_loss": 0.8728694915771484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4573880732059479, "epoch": 11.98, "learning_rate": 1.1195992485911083e-05, "loss": 0.4232, "step": 14169, "task_loss": 0.8858767151832581 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43039435148239136, "epoch": 11.98, "learning_rate": 1.1192861615529118e-05, "loss": 0.4146, "step": 14170, "task_loss": 0.27731919288635254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3306029438972473, "epoch": 11.98, "learning_rate": 1.1189730745147151e-05, "loss": 0.3626, "step": 14171, "task_loss": 0.4907364547252655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40307170152664185, "epoch": 11.98, "learning_rate": 1.1186599874765185e-05, "loss": 0.4418, "step": 14172, "task_loss": 0.813574492931366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27920714020729065, "epoch": 11.98, "learning_rate": 1.118346900438322e-05, "loss": 0.3924, "step": 14173, "task_loss": 0.19262337684631348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45210689306259155, "epoch": 11.98, "learning_rate": 1.1180338134001254e-05, "loss": 0.5583, "step": 14174, "task_loss": 0.25380030274391174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.348021924495697, "epoch": 11.98, "learning_rate": 1.1177207263619287e-05, "loss": 0.3892, "step": 14175, "task_loss": 0.888351321220398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39710021018981934, "epoch": 11.98, "learning_rate": 1.117407639323732e-05, "loss": 0.3624, "step": 14176, "task_loss": 0.28361454606056213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43794405460357666, "epoch": 11.98, "learning_rate": 1.1170945522855354e-05, "loss": 0.6463, "step": 14177, "task_loss": 0.9212139248847961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3844129741191864, "epoch": 11.98, "learning_rate": 1.1167814652473387e-05, "loss": 0.3635, "step": 14178, "task_loss": 0.30763182044029236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3024844527244568, "epoch": 11.99, "learning_rate": 1.1164683782091421e-05, "loss": 0.4412, "step": 14179, "task_loss": 0.10105758160352707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15265776216983795, "epoch": 11.99, "learning_rate": 1.1161552911709454e-05, "loss": 0.4674, "step": 14180, "task_loss": 0.18544183671474457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5587944388389587, "epoch": 11.99, "learning_rate": 1.115842204132749e-05, "loss": 0.5647, "step": 14181, "task_loss": 0.7205062508583069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4327270984649658, "epoch": 11.99, "learning_rate": 1.1155291170945525e-05, "loss": 0.4592, "step": 14182, "task_loss": 0.250076025724411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4514155685901642, "epoch": 11.99, "learning_rate": 1.1152160300563558e-05, "loss": 0.5151, "step": 14183, "task_loss": 0.969456672668457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3812023401260376, "epoch": 11.99, "learning_rate": 1.1149029430181592e-05, "loss": 0.4463, "step": 14184, "task_loss": 0.292491614818573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2548711895942688, "epoch": 11.99, "learning_rate": 1.1145898559799625e-05, "loss": 0.4257, "step": 14185, "task_loss": 0.30036109685897827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3774791657924652, "epoch": 11.99, "learning_rate": 1.1142767689417659e-05, "loss": 0.4723, "step": 14186, "task_loss": 0.07497508823871613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27018439769744873, "epoch": 11.99, "learning_rate": 1.1139636819035692e-05, "loss": 0.4292, "step": 14187, "task_loss": 0.3746761381626129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4645305275917053, "epoch": 11.99, "learning_rate": 1.1136505948653726e-05, "loss": 0.5486, "step": 14188, "task_loss": 0.5372028946876526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29389113187789917, "epoch": 11.99, "learning_rate": 1.113337507827176e-05, "loss": 0.4172, "step": 14189, "task_loss": 0.8306737542152405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22406159341335297, "epoch": 11.99, "learning_rate": 1.1130244207889794e-05, "loss": 0.4499, "step": 14190, "task_loss": 0.41328105330467224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5417648553848267, "epoch": 12.0, "learning_rate": 1.1127113337507828e-05, "loss": 0.4135, "step": 14191, "task_loss": 0.8563370704650879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43708065152168274, "epoch": 12.0, "learning_rate": 1.1123982467125861e-05, "loss": 0.4537, "step": 14192, "task_loss": 0.7050898671150208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3205282986164093, "epoch": 12.0, "learning_rate": 1.1120851596743895e-05, "loss": 0.508, "step": 14193, "task_loss": 0.9095473885536194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4731139540672302, "epoch": 12.0, "learning_rate": 1.1117720726361928e-05, "loss": 0.4433, "step": 14194, "task_loss": 0.25788235664367676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2677261233329773, "epoch": 12.0, "learning_rate": 1.1114589855979963e-05, "loss": 0.4301, "step": 14195, "task_loss": 0.19115126132965088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3660587668418884, "epoch": 12.0, "learning_rate": 1.1111458985597997e-05, "loss": 0.427, "step": 14196, "task_loss": 0.42662206292152405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46143245697021484, "epoch": 12.0, "learning_rate": 1.110832811521603e-05, "loss": 1.001, "step": 14197, "task_loss": 1.2178093194961548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44622349739074707, "epoch": 12.0, "learning_rate": 1.1105197244834065e-05, "loss": 0.5728, "step": 14198, "task_loss": 0.7843007445335388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4885210692882538, "epoch": 12.0, "learning_rate": 1.1102066374452099e-05, "loss": 0.3999, "step": 14199, "task_loss": 0.28947150707244873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4270339608192444, "epoch": 12.0, "learning_rate": 1.1098935504070132e-05, "loss": 0.4577, "step": 14200, "task_loss": 0.95258629322052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2998239994049072, "epoch": 12.0, "learning_rate": 1.1095804633688166e-05, "loss": 0.3592, "step": 14201, "task_loss": 0.6210178136825562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4757399559020996, "epoch": 12.01, "learning_rate": 1.10926737633062e-05, "loss": 0.5112, "step": 14202, "task_loss": 1.1321845054626465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19947952032089233, "epoch": 12.01, "learning_rate": 1.1089542892924233e-05, "loss": 0.5531, "step": 14203, "task_loss": 0.35529401898384094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5281033515930176, "epoch": 12.01, "learning_rate": 1.1086412022542266e-05, "loss": 0.379, "step": 14204, "task_loss": 0.6861400008201599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4210466742515564, "epoch": 12.01, "learning_rate": 1.10832811521603e-05, "loss": 0.5983, "step": 14205, "task_loss": 0.27876660227775574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2725776731967926, "epoch": 12.01, "learning_rate": 1.1080150281778335e-05, "loss": 0.4203, "step": 14206, "task_loss": 0.4947453737258911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23578771948814392, "epoch": 12.01, "learning_rate": 1.107701941139637e-05, "loss": 0.3923, "step": 14207, "task_loss": 0.8739140033721924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6395652294158936, "epoch": 12.01, "learning_rate": 1.1073888541014404e-05, "loss": 0.497, "step": 14208, "task_loss": 0.3625721335411072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.50965416431427, "epoch": 12.01, "learning_rate": 1.1070757670632437e-05, "loss": 0.4012, "step": 14209, "task_loss": 1.0040594339370728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.505757749080658, "epoch": 12.01, "learning_rate": 1.106762680025047e-05, "loss": 0.3497, "step": 14210, "task_loss": 0.9642729759216309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4411538243293762, "epoch": 12.01, "learning_rate": 1.1064495929868504e-05, "loss": 0.4643, "step": 14211, "task_loss": 0.5487423539161682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7081422805786133, "epoch": 12.01, "learning_rate": 1.1061365059486537e-05, "loss": 0.5527, "step": 14212, "task_loss": 0.781170666217804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5413951277732849, "epoch": 12.01, "learning_rate": 1.1058234189104571e-05, "loss": 0.5206, "step": 14213, "task_loss": 1.083329439163208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3317584991455078, "epoch": 12.02, "learning_rate": 1.1055103318722604e-05, "loss": 0.5067, "step": 14214, "task_loss": 0.24890664219856262 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5413247346878052, "epoch": 12.02, "learning_rate": 1.105197244834064e-05, "loss": 0.5267, "step": 14215, "task_loss": 1.3765324354171753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3127034902572632, "epoch": 12.02, "learning_rate": 1.1048841577958673e-05, "loss": 0.4396, "step": 14216, "task_loss": 0.0790405198931694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5648461580276489, "epoch": 12.02, "learning_rate": 1.1045710707576707e-05, "loss": 0.414, "step": 14217, "task_loss": 0.6009764075279236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45241838693618774, "epoch": 12.02, "learning_rate": 1.104257983719474e-05, "loss": 0.4041, "step": 14218, "task_loss": 0.08922677487134933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4369577169418335, "epoch": 12.02, "learning_rate": 1.1039448966812775e-05, "loss": 0.547, "step": 14219, "task_loss": 0.24126578867435455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6147549152374268, "epoch": 12.02, "learning_rate": 1.1036318096430809e-05, "loss": 0.4835, "step": 14220, "task_loss": 0.5615859031677246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3159685432910919, "epoch": 12.02, "learning_rate": 1.1033187226048842e-05, "loss": 0.3737, "step": 14221, "task_loss": 0.6526374220848083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31359145045280457, "epoch": 12.02, "learning_rate": 1.1030056355666876e-05, "loss": 0.4578, "step": 14222, "task_loss": 0.37861669063568115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3326275646686554, "epoch": 12.02, "learning_rate": 1.1026925485284909e-05, "loss": 0.3379, "step": 14223, "task_loss": 0.7693266868591309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8344820141792297, "epoch": 12.02, "learning_rate": 1.1023794614902944e-05, "loss": 0.4885, "step": 14224, "task_loss": 1.4750574827194214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7007253766059875, "epoch": 12.02, "learning_rate": 1.1020663744520978e-05, "loss": 0.5546, "step": 14225, "task_loss": 0.8010768890380859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2694801390171051, "epoch": 12.03, "learning_rate": 1.1017532874139011e-05, "loss": 0.3895, "step": 14226, "task_loss": 0.7884275317192078 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6044503450393677, "epoch": 12.03, "learning_rate": 1.1014402003757045e-05, "loss": 0.4407, "step": 14227, "task_loss": 1.2006651163101196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23918114602565765, "epoch": 12.03, "learning_rate": 1.1011271133375078e-05, "loss": 0.3195, "step": 14228, "task_loss": 0.5358859896659851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5844542980194092, "epoch": 12.03, "learning_rate": 1.1008140262993112e-05, "loss": 0.5705, "step": 14229, "task_loss": 0.4277006685733795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6004002094268799, "epoch": 12.03, "learning_rate": 1.1005009392611145e-05, "loss": 0.4484, "step": 14230, "task_loss": 0.6471752524375916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3404162526130676, "epoch": 12.03, "learning_rate": 1.100187852222918e-05, "loss": 0.5192, "step": 14231, "task_loss": 0.3209749758243561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5395246744155884, "epoch": 12.03, "learning_rate": 1.0998747651847214e-05, "loss": 0.4967, "step": 14232, "task_loss": 0.9758056402206421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3660319447517395, "epoch": 12.03, "learning_rate": 1.0995616781465249e-05, "loss": 0.4473, "step": 14233, "task_loss": 0.5816670060157776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7154187560081482, "epoch": 12.03, "learning_rate": 1.0992485911083282e-05, "loss": 0.5353, "step": 14234, "task_loss": 0.7811766862869263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32804369926452637, "epoch": 12.03, "learning_rate": 1.0989355040701316e-05, "loss": 0.5241, "step": 14235, "task_loss": 1.4386229515075684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45010650157928467, "epoch": 12.03, "learning_rate": 1.098622417031935e-05, "loss": 0.3643, "step": 14236, "task_loss": 0.5852500200271606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4467324912548065, "epoch": 12.03, "learning_rate": 1.0983093299937383e-05, "loss": 0.3611, "step": 14237, "task_loss": 0.5976445078849792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7055814862251282, "epoch": 12.04, "learning_rate": 1.0979962429555416e-05, "loss": 0.5114, "step": 14238, "task_loss": 1.5821348428726196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42524588108062744, "epoch": 12.04, "learning_rate": 1.097683155917345e-05, "loss": 0.3576, "step": 14239, "task_loss": 0.3623480498790741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5565593838691711, "epoch": 12.04, "learning_rate": 1.0973700688791485e-05, "loss": 0.6081, "step": 14240, "task_loss": 0.4083707332611084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0115470886230469, "epoch": 12.04, "learning_rate": 1.0970569818409518e-05, "loss": 0.7285, "step": 14241, "task_loss": 0.45061665773391724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3080175817012787, "epoch": 12.04, "learning_rate": 1.0967438948027552e-05, "loss": 0.4029, "step": 14242, "task_loss": 0.5636217594146729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7094666957855225, "epoch": 12.04, "learning_rate": 1.0964308077645585e-05, "loss": 0.5111, "step": 14243, "task_loss": 1.2744858264923096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47482830286026, "epoch": 12.04, "learning_rate": 1.096117720726362e-05, "loss": 0.487, "step": 14244, "task_loss": 0.31760817766189575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5039408206939697, "epoch": 12.04, "learning_rate": 1.0958046336881654e-05, "loss": 0.5399, "step": 14245, "task_loss": 0.3828793168067932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35840359330177307, "epoch": 12.04, "learning_rate": 1.0954915466499688e-05, "loss": 0.3867, "step": 14246, "task_loss": 0.07382551580667496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6817078590393066, "epoch": 12.04, "learning_rate": 1.0951784596117721e-05, "loss": 0.4502, "step": 14247, "task_loss": 0.7224633097648621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.629581093788147, "epoch": 12.04, "learning_rate": 1.0948653725735754e-05, "loss": 0.4649, "step": 14248, "task_loss": 0.4215214252471924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6054502725601196, "epoch": 12.04, "learning_rate": 1.094552285535379e-05, "loss": 0.4495, "step": 14249, "task_loss": 1.521333932876587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42624807357788086, "epoch": 12.05, "learning_rate": 1.0942391984971823e-05, "loss": 0.4531, "step": 14250, "task_loss": 0.4235699474811554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6088417768478394, "epoch": 12.05, "learning_rate": 1.0939261114589857e-05, "loss": 0.5225, "step": 14251, "task_loss": 1.1338647603988647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3476814329624176, "epoch": 12.05, "learning_rate": 1.093613024420789e-05, "loss": 0.491, "step": 14252, "task_loss": 0.34116607904434204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5123651027679443, "epoch": 12.05, "learning_rate": 1.0932999373825924e-05, "loss": 0.4621, "step": 14253, "task_loss": 1.141829490661621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44851160049438477, "epoch": 12.05, "learning_rate": 1.0929868503443957e-05, "loss": 0.4547, "step": 14254, "task_loss": 0.3347153663635254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28749024868011475, "epoch": 12.05, "learning_rate": 1.092673763306199e-05, "loss": 0.436, "step": 14255, "task_loss": 0.767422616481781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5188666582107544, "epoch": 12.05, "learning_rate": 1.0923606762680026e-05, "loss": 0.4395, "step": 14256, "task_loss": 1.103185772895813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41140660643577576, "epoch": 12.05, "learning_rate": 1.0920475892298059e-05, "loss": 0.4418, "step": 14257, "task_loss": 0.2961469888687134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2966407835483551, "epoch": 12.05, "learning_rate": 1.0917345021916094e-05, "loss": 0.3165, "step": 14258, "task_loss": 0.42115673422813416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6173520684242249, "epoch": 12.05, "learning_rate": 1.0914214151534128e-05, "loss": 0.4118, "step": 14259, "task_loss": 0.4481413960456848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30559107661247253, "epoch": 12.05, "learning_rate": 1.0911083281152161e-05, "loss": 0.4725, "step": 14260, "task_loss": 0.46276265382766724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4384281039237976, "epoch": 12.05, "learning_rate": 1.0907952410770195e-05, "loss": 0.3607, "step": 14261, "task_loss": 0.5512164831161499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7060815095901489, "epoch": 12.06, "learning_rate": 1.0904821540388228e-05, "loss": 0.4778, "step": 14262, "task_loss": 0.5307072401046753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38463109731674194, "epoch": 12.06, "learning_rate": 1.0901690670006262e-05, "loss": 0.4349, "step": 14263, "task_loss": 0.9843781590461731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5911178588867188, "epoch": 12.06, "learning_rate": 1.0898559799624295e-05, "loss": 0.5374, "step": 14264, "task_loss": 0.9166448712348938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3865102231502533, "epoch": 12.06, "learning_rate": 1.089542892924233e-05, "loss": 0.4683, "step": 14265, "task_loss": 0.30350130796432495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6467983722686768, "epoch": 12.06, "learning_rate": 1.0892298058860364e-05, "loss": 0.4592, "step": 14266, "task_loss": 0.6108946800231934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3300192654132843, "epoch": 12.06, "learning_rate": 1.0889167188478397e-05, "loss": 0.3935, "step": 14267, "task_loss": 0.2950390577316284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3907513916492462, "epoch": 12.06, "learning_rate": 1.0886036318096432e-05, "loss": 0.4281, "step": 14268, "task_loss": 0.09810105711221695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5121525526046753, "epoch": 12.06, "learning_rate": 1.0882905447714466e-05, "loss": 0.4423, "step": 14269, "task_loss": 0.9449127316474915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4978410601615906, "epoch": 12.06, "learning_rate": 1.08797745773325e-05, "loss": 0.5242, "step": 14270, "task_loss": 0.5769362449645996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28581905364990234, "epoch": 12.06, "learning_rate": 1.0876643706950533e-05, "loss": 0.3733, "step": 14271, "task_loss": 0.21184402704238892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24548637866973877, "epoch": 12.06, "learning_rate": 1.0873512836568566e-05, "loss": 0.4755, "step": 14272, "task_loss": 0.40707966685295105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6306463479995728, "epoch": 12.07, "learning_rate": 1.08703819661866e-05, "loss": 0.4695, "step": 14273, "task_loss": 0.24166853725910187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5560914278030396, "epoch": 12.07, "learning_rate": 1.0867251095804635e-05, "loss": 0.4575, "step": 14274, "task_loss": 1.642582893371582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7031424045562744, "epoch": 12.07, "learning_rate": 1.0864120225422668e-05, "loss": 0.525, "step": 14275, "task_loss": 0.8397018313407898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4573254883289337, "epoch": 12.07, "learning_rate": 1.0860989355040702e-05, "loss": 0.6095, "step": 14276, "task_loss": 1.9757424592971802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4128069579601288, "epoch": 12.07, "learning_rate": 1.0857858484658735e-05, "loss": 0.472, "step": 14277, "task_loss": 0.5330651998519897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5584495067596436, "epoch": 12.07, "learning_rate": 1.0854727614276769e-05, "loss": 0.4735, "step": 14278, "task_loss": 1.3222459554672241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4494805932044983, "epoch": 12.07, "learning_rate": 1.0851596743894802e-05, "loss": 0.541, "step": 14279, "task_loss": 1.2702590227127075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3833054304122925, "epoch": 12.07, "learning_rate": 1.0848465873512838e-05, "loss": 0.476, "step": 14280, "task_loss": 0.6063472032546997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4197247624397278, "epoch": 12.07, "learning_rate": 1.0845335003130871e-05, "loss": 0.4073, "step": 14281, "task_loss": 0.8339488506317139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4904690384864807, "epoch": 12.07, "learning_rate": 1.0842204132748904e-05, "loss": 0.6244, "step": 14282, "task_loss": 0.5087635517120361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5838115811347961, "epoch": 12.07, "learning_rate": 1.083907326236694e-05, "loss": 0.4954, "step": 14283, "task_loss": 1.194712519645691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46960610151290894, "epoch": 12.07, "learning_rate": 1.0835942391984973e-05, "loss": 0.4771, "step": 14284, "task_loss": 0.9346461296081543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5106005072593689, "epoch": 12.08, "learning_rate": 1.0832811521603007e-05, "loss": 0.4374, "step": 14285, "task_loss": 1.2684897184371948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3431893289089203, "epoch": 12.08, "learning_rate": 1.082968065122104e-05, "loss": 0.5451, "step": 14286, "task_loss": 1.1898401975631714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3114050626754761, "epoch": 12.08, "learning_rate": 1.0826549780839074e-05, "loss": 0.3674, "step": 14287, "task_loss": 0.2262502759695053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32455191016197205, "epoch": 12.08, "learning_rate": 1.0823418910457107e-05, "loss": 0.2573, "step": 14288, "task_loss": 0.12605667114257812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.4302022457122803, "epoch": 12.08, "learning_rate": 1.082028804007514e-05, "loss": 0.7394, "step": 14289, "task_loss": 1.8735564947128296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24501660466194153, "epoch": 12.08, "learning_rate": 1.0817157169693174e-05, "loss": 0.4585, "step": 14290, "task_loss": 0.4140072166919708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30826854705810547, "epoch": 12.08, "learning_rate": 1.0814026299311209e-05, "loss": 0.4401, "step": 14291, "task_loss": 0.4043033719062805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.381248414516449, "epoch": 12.08, "learning_rate": 1.0810895428929243e-05, "loss": 0.3533, "step": 14292, "task_loss": 0.5500754714012146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5700349807739258, "epoch": 12.08, "learning_rate": 1.0807764558547278e-05, "loss": 0.4283, "step": 14293, "task_loss": 1.1878700256347656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45606687664985657, "epoch": 12.08, "learning_rate": 1.0804633688165311e-05, "loss": 0.3845, "step": 14294, "task_loss": 0.6546815037727356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3108525276184082, "epoch": 12.08, "learning_rate": 1.0801502817783345e-05, "loss": 0.3706, "step": 14295, "task_loss": 0.46056726574897766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5195808410644531, "epoch": 12.08, "learning_rate": 1.0798371947401378e-05, "loss": 0.3805, "step": 14296, "task_loss": 0.3188391327857971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.516746997833252, "epoch": 12.09, "learning_rate": 1.0795241077019412e-05, "loss": 0.4503, "step": 14297, "task_loss": 0.9458262920379639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41447335481643677, "epoch": 12.09, "learning_rate": 1.0792110206637445e-05, "loss": 0.4743, "step": 14298, "task_loss": 0.21779371798038483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.377832293510437, "epoch": 12.09, "learning_rate": 1.0788979336255479e-05, "loss": 0.4524, "step": 14299, "task_loss": 0.6515744924545288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47972288727760315, "epoch": 12.09, "learning_rate": 1.0785848465873514e-05, "loss": 0.4477, "step": 14300, "task_loss": 0.8766074776649475 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4689141511917114, "epoch": 12.09, "learning_rate": 1.0782717595491547e-05, "loss": 0.5331, "step": 14301, "task_loss": 0.5298905968666077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27496352791786194, "epoch": 12.09, "learning_rate": 1.077958672510958e-05, "loss": 0.3061, "step": 14302, "task_loss": 0.05126400291919708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32177743315696716, "epoch": 12.09, "learning_rate": 1.0776455854727614e-05, "loss": 0.4423, "step": 14303, "task_loss": 0.17065125703811646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35393935441970825, "epoch": 12.09, "learning_rate": 1.0773324984345648e-05, "loss": 0.443, "step": 14304, "task_loss": 0.28802990913391113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.09740079939365387, "epoch": 12.09, "learning_rate": 1.0770194113963683e-05, "loss": 0.4218, "step": 14305, "task_loss": 0.004234001040458679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5315406322479248, "epoch": 12.09, "learning_rate": 1.0767063243581716e-05, "loss": 0.4673, "step": 14306, "task_loss": 1.1205406188964844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29441601037979126, "epoch": 12.09, "learning_rate": 1.076393237319975e-05, "loss": 0.5164, "step": 14307, "task_loss": 0.19648346304893494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.526677131652832, "epoch": 12.09, "learning_rate": 1.0760801502817783e-05, "loss": 0.4301, "step": 14308, "task_loss": 0.7646129727363586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36331820487976074, "epoch": 12.1, "learning_rate": 1.0757670632435818e-05, "loss": 0.361, "step": 14309, "task_loss": 0.9324651956558228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3327568769454956, "epoch": 12.1, "learning_rate": 1.0754539762053852e-05, "loss": 0.4555, "step": 14310, "task_loss": 0.637464165687561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22174501419067383, "epoch": 12.1, "learning_rate": 1.0751408891671885e-05, "loss": 0.3564, "step": 14311, "task_loss": 0.36637696623802185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5162281394004822, "epoch": 12.1, "learning_rate": 1.0748278021289919e-05, "loss": 0.5035, "step": 14312, "task_loss": 0.1608704775571823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5740127563476562, "epoch": 12.1, "learning_rate": 1.0745147150907952e-05, "loss": 0.5064, "step": 14313, "task_loss": 0.4751388132572174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28744786977767944, "epoch": 12.1, "learning_rate": 1.0742016280525986e-05, "loss": 0.3977, "step": 14314, "task_loss": 0.3980337679386139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6065167188644409, "epoch": 12.1, "learning_rate": 1.073888541014402e-05, "loss": 0.5846, "step": 14315, "task_loss": 1.0426876544952393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5903807878494263, "epoch": 12.1, "learning_rate": 1.0735754539762055e-05, "loss": 0.431, "step": 14316, "task_loss": 0.3804337680339813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5221355557441711, "epoch": 12.1, "learning_rate": 1.073262366938009e-05, "loss": 0.4535, "step": 14317, "task_loss": 1.2695155143737793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4769640564918518, "epoch": 12.1, "learning_rate": 1.0729492798998123e-05, "loss": 0.5827, "step": 14318, "task_loss": 0.2655911147594452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3285027742385864, "epoch": 12.1, "learning_rate": 1.0726361928616157e-05, "loss": 0.3557, "step": 14319, "task_loss": 0.8144038319587708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2877611517906189, "epoch": 12.1, "learning_rate": 1.072323105823419e-05, "loss": 0.4061, "step": 14320, "task_loss": 0.5053623914718628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5474313497543335, "epoch": 12.11, "learning_rate": 1.0720100187852224e-05, "loss": 0.4885, "step": 14321, "task_loss": 0.9696460962295532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5281748175621033, "epoch": 12.11, "learning_rate": 1.0716969317470257e-05, "loss": 0.5679, "step": 14322, "task_loss": 1.0944161415100098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39178144931793213, "epoch": 12.11, "learning_rate": 1.071383844708829e-05, "loss": 0.5145, "step": 14323, "task_loss": 0.5011078119277954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27589333057403564, "epoch": 12.11, "learning_rate": 1.0710707576706324e-05, "loss": 0.3972, "step": 14324, "task_loss": 0.6269230842590332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24151262640953064, "epoch": 12.11, "learning_rate": 1.070757670632436e-05, "loss": 0.34, "step": 14325, "task_loss": 0.3392014801502228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5283783674240112, "epoch": 12.11, "learning_rate": 1.0704445835942393e-05, "loss": 0.5247, "step": 14326, "task_loss": 0.8659906387329102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4650988280773163, "epoch": 12.11, "learning_rate": 1.0701314965560426e-05, "loss": 0.5887, "step": 14327, "task_loss": 1.0284438133239746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28769245743751526, "epoch": 12.11, "learning_rate": 1.069818409517846e-05, "loss": 0.3589, "step": 14328, "task_loss": 0.1659041941165924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18874487280845642, "epoch": 12.11, "learning_rate": 1.0695053224796495e-05, "loss": 0.3016, "step": 14329, "task_loss": 0.3652135133743286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8771164417266846, "epoch": 12.11, "learning_rate": 1.0691922354414528e-05, "loss": 0.5271, "step": 14330, "task_loss": 0.5827243328094482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4464137554168701, "epoch": 12.11, "learning_rate": 1.0688791484032562e-05, "loss": 0.4503, "step": 14331, "task_loss": 0.4676007330417633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24739885330200195, "epoch": 12.11, "learning_rate": 1.0685660613650595e-05, "loss": 0.3492, "step": 14332, "task_loss": 0.14592798054218292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.605164110660553, "epoch": 12.12, "learning_rate": 1.0682529743268629e-05, "loss": 0.6102, "step": 14333, "task_loss": 0.49635419249534607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4590044617652893, "epoch": 12.12, "learning_rate": 1.0679398872886664e-05, "loss": 0.4853, "step": 14334, "task_loss": 0.05645718425512314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5362765192985535, "epoch": 12.12, "learning_rate": 1.0676268002504697e-05, "loss": 0.6482, "step": 14335, "task_loss": 1.4737306833267212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4526323080062866, "epoch": 12.12, "learning_rate": 1.067313713212273e-05, "loss": 0.3027, "step": 14336, "task_loss": 0.2732606530189514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4897040128707886, "epoch": 12.12, "learning_rate": 1.0670006261740764e-05, "loss": 0.4027, "step": 14337, "task_loss": 0.47824376821517944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29795992374420166, "epoch": 12.12, "learning_rate": 1.0666875391358798e-05, "loss": 0.389, "step": 14338, "task_loss": 0.34079065918922424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46705082058906555, "epoch": 12.12, "learning_rate": 1.0663744520976831e-05, "loss": 0.5157, "step": 14339, "task_loss": 0.5222864151000977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3598610758781433, "epoch": 12.12, "learning_rate": 1.0660613650594865e-05, "loss": 0.5716, "step": 14340, "task_loss": 0.12275747209787369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38575804233551025, "epoch": 12.12, "learning_rate": 1.06574827802129e-05, "loss": 0.4274, "step": 14341, "task_loss": 1.2229626178741455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24070696532726288, "epoch": 12.12, "learning_rate": 1.0654351909830933e-05, "loss": 0.3255, "step": 14342, "task_loss": 0.06829378008842468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5357925295829773, "epoch": 12.12, "learning_rate": 1.0651221039448969e-05, "loss": 0.4305, "step": 14343, "task_loss": 0.6143598556518555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3870159387588501, "epoch": 12.13, "learning_rate": 1.0648090169067002e-05, "loss": 0.5452, "step": 14344, "task_loss": 0.46229368448257446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6079191565513611, "epoch": 12.13, "learning_rate": 1.0644959298685035e-05, "loss": 0.564, "step": 14345, "task_loss": 0.7564048171043396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4514954686164856, "epoch": 12.13, "learning_rate": 1.0641828428303069e-05, "loss": 0.4368, "step": 14346, "task_loss": 0.42757710814476013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36132046580314636, "epoch": 12.13, "learning_rate": 1.0638697557921102e-05, "loss": 0.5186, "step": 14347, "task_loss": 0.72565096616745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5947951078414917, "epoch": 12.13, "learning_rate": 1.0635566687539136e-05, "loss": 0.5169, "step": 14348, "task_loss": 1.034671664237976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32464849948883057, "epoch": 12.13, "learning_rate": 1.063243581715717e-05, "loss": 0.4886, "step": 14349, "task_loss": 0.503749430179596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7351672649383545, "epoch": 12.13, "learning_rate": 1.0629304946775205e-05, "loss": 0.6231, "step": 14350, "task_loss": 0.6374776363372803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45760685205459595, "epoch": 12.13, "learning_rate": 1.0626174076393238e-05, "loss": 0.4587, "step": 14351, "task_loss": 0.17940083146095276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4162447452545166, "epoch": 12.13, "learning_rate": 1.0623043206011271e-05, "loss": 0.4031, "step": 14352, "task_loss": 0.5281674265861511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27064162492752075, "epoch": 12.13, "learning_rate": 1.0619912335629305e-05, "loss": 0.5011, "step": 14353, "task_loss": 0.09629122167825699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49322351813316345, "epoch": 12.13, "learning_rate": 1.061678146524734e-05, "loss": 0.4013, "step": 14354, "task_loss": 1.065155267715454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5707390308380127, "epoch": 12.13, "learning_rate": 1.0613650594865374e-05, "loss": 0.4506, "step": 14355, "task_loss": 1.1495569944381714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32197800278663635, "epoch": 12.14, "learning_rate": 1.0610519724483407e-05, "loss": 0.369, "step": 14356, "task_loss": 0.4552195072174072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36918556690216064, "epoch": 12.14, "learning_rate": 1.060738885410144e-05, "loss": 0.4555, "step": 14357, "task_loss": 0.17733369767665863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43694567680358887, "epoch": 12.14, "learning_rate": 1.0604257983719474e-05, "loss": 0.702, "step": 14358, "task_loss": 1.0117051601409912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3128105700016022, "epoch": 12.14, "learning_rate": 1.060112711333751e-05, "loss": 0.47, "step": 14359, "task_loss": 0.555210292339325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5245579481124878, "epoch": 12.14, "learning_rate": 1.0597996242955543e-05, "loss": 0.5273, "step": 14360, "task_loss": 0.34215468168258667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5041369199752808, "epoch": 12.14, "learning_rate": 1.0594865372573576e-05, "loss": 0.4619, "step": 14361, "task_loss": 0.9245482087135315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030362606048584, "epoch": 12.14, "learning_rate": 1.059173450219161e-05, "loss": 0.4444, "step": 14362, "task_loss": 0.5341368317604065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39705395698547363, "epoch": 12.14, "learning_rate": 1.0588603631809643e-05, "loss": 0.5823, "step": 14363, "task_loss": 1.4101688861846924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4200749397277832, "epoch": 12.14, "learning_rate": 1.0585472761427677e-05, "loss": 0.5351, "step": 14364, "task_loss": 0.34908124804496765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4344818592071533, "epoch": 12.14, "learning_rate": 1.058234189104571e-05, "loss": 0.4765, "step": 14365, "task_loss": 0.9257472157478333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3961338996887207, "epoch": 12.14, "learning_rate": 1.0579211020663745e-05, "loss": 0.3881, "step": 14366, "task_loss": 1.5195409059524536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4841802716255188, "epoch": 12.14, "learning_rate": 1.0576080150281779e-05, "loss": 0.4542, "step": 14367, "task_loss": 0.4519163966178894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21970084309577942, "epoch": 12.15, "learning_rate": 1.0572949279899814e-05, "loss": 0.4601, "step": 14368, "task_loss": 0.25875866413116455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45513150095939636, "epoch": 12.15, "learning_rate": 1.0569818409517847e-05, "loss": 0.4271, "step": 14369, "task_loss": 0.5981289744377136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7893266677856445, "epoch": 12.15, "learning_rate": 1.056668753913588e-05, "loss": 0.6339, "step": 14370, "task_loss": 0.44379526376724243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40752536058425903, "epoch": 12.15, "learning_rate": 1.0563556668753914e-05, "loss": 0.5088, "step": 14371, "task_loss": 1.31508469581604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22788561880588531, "epoch": 12.15, "learning_rate": 1.0560425798371948e-05, "loss": 0.4118, "step": 14372, "task_loss": 0.6536576151847839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3483720123767853, "epoch": 12.15, "learning_rate": 1.0557294927989981e-05, "loss": 0.4358, "step": 14373, "task_loss": 0.2789776921272278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3046489357948303, "epoch": 12.15, "learning_rate": 1.0554164057608015e-05, "loss": 0.4685, "step": 14374, "task_loss": 0.598371148109436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.300214946269989, "epoch": 12.15, "learning_rate": 1.0551033187226048e-05, "loss": 0.444, "step": 14375, "task_loss": 0.49128979444503784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5397158265113831, "epoch": 12.15, "learning_rate": 1.0547902316844083e-05, "loss": 0.3875, "step": 14376, "task_loss": 0.8464384078979492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5021793842315674, "epoch": 12.15, "learning_rate": 1.0544771446462117e-05, "loss": 0.4205, "step": 14377, "task_loss": 0.23266632854938507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4545818567276001, "epoch": 12.15, "learning_rate": 1.054164057608015e-05, "loss": 0.5472, "step": 14378, "task_loss": 0.5537594556808472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.365557461977005, "epoch": 12.15, "learning_rate": 1.0538509705698185e-05, "loss": 0.3033, "step": 14379, "task_loss": 0.7085545063018799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4609684944152832, "epoch": 12.16, "learning_rate": 1.0535378835316219e-05, "loss": 0.5536, "step": 14380, "task_loss": 0.8451536893844604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32743537425994873, "epoch": 12.16, "learning_rate": 1.0532247964934252e-05, "loss": 0.4045, "step": 14381, "task_loss": 0.2836667597293854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3480837941169739, "epoch": 12.16, "learning_rate": 1.0529117094552286e-05, "loss": 0.3901, "step": 14382, "task_loss": 0.578345000743866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3884063959121704, "epoch": 12.16, "learning_rate": 1.052598622417032e-05, "loss": 0.4583, "step": 14383, "task_loss": 0.1885230839252472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5062903165817261, "epoch": 12.16, "learning_rate": 1.0522855353788353e-05, "loss": 0.4379, "step": 14384, "task_loss": 0.8300375938415527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44222140312194824, "epoch": 12.16, "learning_rate": 1.0519724483406388e-05, "loss": 0.4627, "step": 14385, "task_loss": 0.6230456829071045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33393535017967224, "epoch": 12.16, "learning_rate": 1.0516593613024421e-05, "loss": 0.3356, "step": 14386, "task_loss": 0.1344321221113205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3238874673843384, "epoch": 12.16, "learning_rate": 1.0513462742642455e-05, "loss": 0.426, "step": 14387, "task_loss": 0.672087550163269 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4814194440841675, "epoch": 12.16, "learning_rate": 1.0510331872260488e-05, "loss": 0.4728, "step": 14388, "task_loss": 0.6512962579727173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7607449889183044, "epoch": 12.16, "learning_rate": 1.0507201001878522e-05, "loss": 0.5312, "step": 14389, "task_loss": 0.9123091697692871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35607463121414185, "epoch": 12.16, "learning_rate": 1.0504070131496555e-05, "loss": 0.5937, "step": 14390, "task_loss": 0.7389654517173767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5289707183837891, "epoch": 12.16, "learning_rate": 1.050093926111459e-05, "loss": 0.6296, "step": 14391, "task_loss": 0.9860467314720154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3435077667236328, "epoch": 12.17, "learning_rate": 1.0497808390732624e-05, "loss": 0.4408, "step": 14392, "task_loss": 0.7003581523895264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2566436231136322, "epoch": 12.17, "learning_rate": 1.049467752035066e-05, "loss": 0.2118, "step": 14393, "task_loss": 0.15457573533058167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39128342270851135, "epoch": 12.17, "learning_rate": 1.0491546649968693e-05, "loss": 0.4746, "step": 14394, "task_loss": 0.8048794865608215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6222230195999146, "epoch": 12.17, "learning_rate": 1.0488415779586726e-05, "loss": 0.4665, "step": 14395, "task_loss": 0.7310325503349304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4979274868965149, "epoch": 12.17, "learning_rate": 1.048528490920476e-05, "loss": 0.6651, "step": 14396, "task_loss": 0.6782494187355042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3838594853878021, "epoch": 12.17, "learning_rate": 1.0482154038822793e-05, "loss": 0.5141, "step": 14397, "task_loss": 1.0308129787445068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6183660626411438, "epoch": 12.17, "learning_rate": 1.0479023168440827e-05, "loss": 0.4724, "step": 14398, "task_loss": 0.5613218545913696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28680115938186646, "epoch": 12.17, "learning_rate": 1.047589229805886e-05, "loss": 0.4948, "step": 14399, "task_loss": 0.15128111839294434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4225577414035797, "epoch": 12.17, "learning_rate": 1.0472761427676894e-05, "loss": 0.42, "step": 14400, "task_loss": 0.279283732175827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3566906452178955, "epoch": 12.17, "learning_rate": 1.0469630557294929e-05, "loss": 0.4476, "step": 14401, "task_loss": 1.2375667095184326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37281373143196106, "epoch": 12.17, "learning_rate": 1.0466499686912962e-05, "loss": 0.4644, "step": 14402, "task_loss": 0.8519936800003052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4349334239959717, "epoch": 12.17, "learning_rate": 1.0463368816530997e-05, "loss": 0.3353, "step": 14403, "task_loss": 0.6405240297317505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4354105591773987, "epoch": 12.18, "learning_rate": 1.046023794614903e-05, "loss": 0.378, "step": 14404, "task_loss": 1.1605160236358643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4261544346809387, "epoch": 12.18, "learning_rate": 1.0457107075767064e-05, "loss": 0.501, "step": 14405, "task_loss": 0.39255887269973755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2636262774467468, "epoch": 12.18, "learning_rate": 1.0453976205385098e-05, "loss": 0.4251, "step": 14406, "task_loss": 0.665710985660553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33147701621055603, "epoch": 12.18, "learning_rate": 1.0450845335003131e-05, "loss": 0.2957, "step": 14407, "task_loss": 0.581792414188385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39237579703330994, "epoch": 12.18, "learning_rate": 1.0447714464621165e-05, "loss": 0.5528, "step": 14408, "task_loss": 0.75770103931427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5021493434906006, "epoch": 12.18, "learning_rate": 1.0444583594239198e-05, "loss": 0.4694, "step": 14409, "task_loss": 1.0850639343261719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41866785287857056, "epoch": 12.18, "learning_rate": 1.0441452723857233e-05, "loss": 0.4007, "step": 14410, "task_loss": 0.30870091915130615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46387138962745667, "epoch": 12.18, "learning_rate": 1.0438321853475267e-05, "loss": 0.4722, "step": 14411, "task_loss": 0.15239259600639343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8393765091896057, "epoch": 12.18, "learning_rate": 1.04351909830933e-05, "loss": 0.6262, "step": 14412, "task_loss": 0.6006639003753662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6553009748458862, "epoch": 12.18, "learning_rate": 1.0432060112711334e-05, "loss": 0.6277, "step": 14413, "task_loss": 0.2514353394508362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0921969413757324, "epoch": 12.18, "learning_rate": 1.0428929242329367e-05, "loss": 0.7588, "step": 14414, "task_loss": 0.8854677677154541 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4458732604980469, "epoch": 12.19, "learning_rate": 1.0425798371947402e-05, "loss": 0.3486, "step": 14415, "task_loss": 0.6304497718811035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36313730478286743, "epoch": 12.19, "learning_rate": 1.0422667501565436e-05, "loss": 0.492, "step": 14416, "task_loss": 1.4621870517730713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4968961179256439, "epoch": 12.19, "learning_rate": 1.041953663118347e-05, "loss": 0.4283, "step": 14417, "task_loss": 0.4673575758934021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5730973482131958, "epoch": 12.19, "learning_rate": 1.0416405760801503e-05, "loss": 0.6016, "step": 14418, "task_loss": 1.0135385990142822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4741082191467285, "epoch": 12.19, "learning_rate": 1.0413274890419538e-05, "loss": 0.3755, "step": 14419, "task_loss": 0.45772960782051086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3423520624637604, "epoch": 12.19, "learning_rate": 1.0410144020037572e-05, "loss": 0.4917, "step": 14420, "task_loss": 0.6133259534835815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.364510715007782, "epoch": 12.19, "learning_rate": 1.0407013149655605e-05, "loss": 0.3812, "step": 14421, "task_loss": 0.4161166548728943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4240012764930725, "epoch": 12.19, "learning_rate": 1.0403882279273638e-05, "loss": 0.4043, "step": 14422, "task_loss": 0.7641737461090088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44077232480049133, "epoch": 12.19, "learning_rate": 1.0400751408891672e-05, "loss": 0.4159, "step": 14423, "task_loss": 0.37722450494766235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4951172471046448, "epoch": 12.19, "learning_rate": 1.0397620538509705e-05, "loss": 0.6154, "step": 14424, "task_loss": 0.36513155698776245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.427529901266098, "epoch": 12.19, "learning_rate": 1.0394489668127739e-05, "loss": 0.3653, "step": 14425, "task_loss": 0.5739886164665222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5879110097885132, "epoch": 12.19, "learning_rate": 1.0391358797745774e-05, "loss": 0.497, "step": 14426, "task_loss": 0.022833621129393578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4721629023551941, "epoch": 12.2, "learning_rate": 1.0388227927363808e-05, "loss": 0.4277, "step": 14427, "task_loss": 1.0874279737472534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7417623400688171, "epoch": 12.2, "learning_rate": 1.0385097056981843e-05, "loss": 0.5856, "step": 14428, "task_loss": 1.193292498588562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6387940645217896, "epoch": 12.2, "learning_rate": 1.0381966186599876e-05, "loss": 0.4568, "step": 14429, "task_loss": 0.37833914160728455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8463840484619141, "epoch": 12.2, "learning_rate": 1.037883531621791e-05, "loss": 0.5878, "step": 14430, "task_loss": 0.8928154110908508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6879289150238037, "epoch": 12.2, "learning_rate": 1.0375704445835943e-05, "loss": 0.5606, "step": 14431, "task_loss": 1.822443962097168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36211714148521423, "epoch": 12.2, "learning_rate": 1.0372573575453977e-05, "loss": 0.3596, "step": 14432, "task_loss": 0.2147509753704071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35199254751205444, "epoch": 12.2, "learning_rate": 1.036944270507201e-05, "loss": 0.4398, "step": 14433, "task_loss": 0.21253979206085205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28367722034454346, "epoch": 12.2, "learning_rate": 1.0366311834690044e-05, "loss": 0.4001, "step": 14434, "task_loss": 0.08705613017082214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3618050217628479, "epoch": 12.2, "learning_rate": 1.0363180964308079e-05, "loss": 0.3575, "step": 14435, "task_loss": 1.080910325050354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45351630449295044, "epoch": 12.2, "learning_rate": 1.0360050093926112e-05, "loss": 0.3833, "step": 14436, "task_loss": 0.8203758001327515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8436499834060669, "epoch": 12.2, "learning_rate": 1.0356919223544146e-05, "loss": 0.5461, "step": 14437, "task_loss": 1.403106927871704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6357681155204773, "epoch": 12.2, "learning_rate": 1.0353788353162179e-05, "loss": 0.549, "step": 14438, "task_loss": 1.7529962062835693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2131883203983307, "epoch": 12.21, "learning_rate": 1.0350657482780213e-05, "loss": 0.3929, "step": 14439, "task_loss": 0.14342816174030304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46421951055526733, "epoch": 12.21, "learning_rate": 1.0347526612398248e-05, "loss": 0.3891, "step": 14440, "task_loss": 1.053742527961731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40772315859794617, "epoch": 12.21, "learning_rate": 1.0344395742016281e-05, "loss": 0.531, "step": 14441, "task_loss": 0.33124926686286926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40094950795173645, "epoch": 12.21, "learning_rate": 1.0341264871634315e-05, "loss": 0.32, "step": 14442, "task_loss": 0.37712255120277405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30377310514450073, "epoch": 12.21, "learning_rate": 1.0338134001252348e-05, "loss": 0.2972, "step": 14443, "task_loss": 1.2096781730651855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7174099087715149, "epoch": 12.21, "learning_rate": 1.0335003130870383e-05, "loss": 0.496, "step": 14444, "task_loss": 1.5609568357467651 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39331623911857605, "epoch": 12.21, "learning_rate": 1.0331872260488417e-05, "loss": 0.5037, "step": 14445, "task_loss": 1.032792568206787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5183265209197998, "epoch": 12.21, "learning_rate": 1.032874139010645e-05, "loss": 0.4854, "step": 14446, "task_loss": 1.049506425857544 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47514277696609497, "epoch": 12.21, "learning_rate": 1.0325610519724484e-05, "loss": 0.6691, "step": 14447, "task_loss": 0.5313239097595215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47613683342933655, "epoch": 12.21, "learning_rate": 1.0322479649342517e-05, "loss": 0.4473, "step": 14448, "task_loss": 0.7659183144569397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4101937413215637, "epoch": 12.21, "learning_rate": 1.031934877896055e-05, "loss": 0.5418, "step": 14449, "task_loss": 0.7467595934867859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0098963975906372, "epoch": 12.21, "learning_rate": 1.0316217908578584e-05, "loss": 0.5889, "step": 14450, "task_loss": 1.2725436687469482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4754630923271179, "epoch": 12.22, "learning_rate": 1.0313087038196618e-05, "loss": 0.4132, "step": 14451, "task_loss": 0.6962178945541382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28643181920051575, "epoch": 12.22, "learning_rate": 1.0309956167814653e-05, "loss": 0.4898, "step": 14452, "task_loss": 0.5916834473609924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44634881615638733, "epoch": 12.22, "learning_rate": 1.0306825297432688e-05, "loss": 0.405, "step": 14453, "task_loss": 0.5861638784408569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4780642092227936, "epoch": 12.22, "learning_rate": 1.0303694427050722e-05, "loss": 0.4447, "step": 14454, "task_loss": 0.35493382811546326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28228139877319336, "epoch": 12.22, "learning_rate": 1.0300563556668755e-05, "loss": 0.3488, "step": 14455, "task_loss": 0.6362370848655701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.243242084980011, "epoch": 12.22, "learning_rate": 1.0297432686286788e-05, "loss": 0.4024, "step": 14456, "task_loss": 0.7034878134727478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3152490556240082, "epoch": 12.22, "learning_rate": 1.0294301815904822e-05, "loss": 0.3787, "step": 14457, "task_loss": 0.27997326850891113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49571770429611206, "epoch": 12.22, "learning_rate": 1.0291170945522855e-05, "loss": 0.4362, "step": 14458, "task_loss": 0.8413154482841492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43845507502555847, "epoch": 12.22, "learning_rate": 1.0288040075140889e-05, "loss": 0.4176, "step": 14459, "task_loss": 0.7176029682159424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30476614832878113, "epoch": 12.22, "learning_rate": 1.0284909204758922e-05, "loss": 0.3924, "step": 14460, "task_loss": 0.7703454494476318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2649594247341156, "epoch": 12.22, "learning_rate": 1.0281778334376958e-05, "loss": 0.3718, "step": 14461, "task_loss": 0.438680499792099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6323792934417725, "epoch": 12.22, "learning_rate": 1.0278647463994991e-05, "loss": 0.521, "step": 14462, "task_loss": 0.7026801109313965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34872913360595703, "epoch": 12.23, "learning_rate": 1.0275516593613024e-05, "loss": 0.4694, "step": 14463, "task_loss": 0.5214009881019592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4487060010433197, "epoch": 12.23, "learning_rate": 1.027238572323106e-05, "loss": 0.4283, "step": 14464, "task_loss": 0.4911840856075287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6098967790603638, "epoch": 12.23, "learning_rate": 1.0269254852849093e-05, "loss": 0.5077, "step": 14465, "task_loss": 0.9063931107521057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.611375093460083, "epoch": 12.23, "learning_rate": 1.0266123982467127e-05, "loss": 0.6204, "step": 14466, "task_loss": 1.1654434204101562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6288065910339355, "epoch": 12.23, "learning_rate": 1.026299311208516e-05, "loss": 0.498, "step": 14467, "task_loss": 0.7282651662826538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3574977517127991, "epoch": 12.23, "learning_rate": 1.0259862241703194e-05, "loss": 0.3804, "step": 14468, "task_loss": 0.4534858167171478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8085570335388184, "epoch": 12.23, "learning_rate": 1.0256731371321229e-05, "loss": 0.6034, "step": 14469, "task_loss": 1.3755818605422974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3932071626186371, "epoch": 12.23, "learning_rate": 1.0253600500939262e-05, "loss": 0.347, "step": 14470, "task_loss": 0.2860782742500305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3250789940357208, "epoch": 12.23, "learning_rate": 1.0250469630557296e-05, "loss": 0.3952, "step": 14471, "task_loss": 1.1748497486114502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3216533362865448, "epoch": 12.23, "learning_rate": 1.024733876017533e-05, "loss": 0.4293, "step": 14472, "task_loss": 0.4709009826183319 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2494492530822754, "epoch": 12.23, "learning_rate": 1.0244207889793363e-05, "loss": 0.3922, "step": 14473, "task_loss": 0.16961602866649628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9228334426879883, "epoch": 12.23, "learning_rate": 1.0241077019411396e-05, "loss": 0.5976, "step": 14474, "task_loss": 0.5621400475502014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5778192281723022, "epoch": 12.24, "learning_rate": 1.023794614902943e-05, "loss": 0.4593, "step": 14475, "task_loss": 1.3145537376403809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20948417484760284, "epoch": 12.24, "learning_rate": 1.0234815278647463e-05, "loss": 0.4539, "step": 14476, "task_loss": 0.057169362902641296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5281665921211243, "epoch": 12.24, "learning_rate": 1.0231684408265498e-05, "loss": 0.5946, "step": 14477, "task_loss": 0.4867236316204071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4313843250274658, "epoch": 12.24, "learning_rate": 1.0228553537883533e-05, "loss": 0.4112, "step": 14478, "task_loss": 0.570667564868927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5530379414558411, "epoch": 12.24, "learning_rate": 1.0225422667501567e-05, "loss": 0.6289, "step": 14479, "task_loss": 0.07699833065271378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3982776999473572, "epoch": 12.24, "learning_rate": 1.02222917971196e-05, "loss": 0.5931, "step": 14480, "task_loss": 1.4724922180175781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5062803626060486, "epoch": 12.24, "learning_rate": 1.0219160926737634e-05, "loss": 0.4135, "step": 14481, "task_loss": 0.23322251439094543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43579232692718506, "epoch": 12.24, "learning_rate": 1.0216030056355667e-05, "loss": 0.383, "step": 14482, "task_loss": 0.1957930326461792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31929194927215576, "epoch": 12.24, "learning_rate": 1.02128991859737e-05, "loss": 0.4158, "step": 14483, "task_loss": 0.2037956714630127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31927573680877686, "epoch": 12.24, "learning_rate": 1.0209768315591734e-05, "loss": 0.4806, "step": 14484, "task_loss": 0.9961643218994141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38961637020111084, "epoch": 12.24, "learning_rate": 1.0206637445209768e-05, "loss": 0.4147, "step": 14485, "task_loss": 1.4935146570205688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.643234372138977, "epoch": 12.24, "learning_rate": 1.0203506574827803e-05, "loss": 0.506, "step": 14486, "task_loss": 0.8576134443283081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.284318745136261, "epoch": 12.25, "learning_rate": 1.0200375704445836e-05, "loss": 0.3256, "step": 14487, "task_loss": 0.528914213180542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46164485812187195, "epoch": 12.25, "learning_rate": 1.019724483406387e-05, "loss": 0.4254, "step": 14488, "task_loss": 1.0491230487823486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6016682982444763, "epoch": 12.25, "learning_rate": 1.0194113963681905e-05, "loss": 0.6015, "step": 14489, "task_loss": 0.3544567823410034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48786869645118713, "epoch": 12.25, "learning_rate": 1.0190983093299938e-05, "loss": 0.4944, "step": 14490, "task_loss": 1.2955312728881836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4367653429508209, "epoch": 12.25, "learning_rate": 1.0187852222917972e-05, "loss": 0.4584, "step": 14491, "task_loss": 0.7390338182449341 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2372315675020218, "epoch": 12.25, "learning_rate": 1.0184721352536005e-05, "loss": 0.3915, "step": 14492, "task_loss": 0.38172993063926697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45365989208221436, "epoch": 12.25, "learning_rate": 1.0181590482154039e-05, "loss": 0.4072, "step": 14493, "task_loss": 0.9413063526153564 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4093652367591858, "epoch": 12.25, "learning_rate": 1.0178459611772072e-05, "loss": 0.4154, "step": 14494, "task_loss": 0.7636248469352722 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2721865177154541, "epoch": 12.25, "learning_rate": 1.0175328741390108e-05, "loss": 0.3386, "step": 14495, "task_loss": 0.5567446351051331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29463934898376465, "epoch": 12.25, "learning_rate": 1.0172197871008141e-05, "loss": 0.4281, "step": 14496, "task_loss": 0.21087147295475006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4380183517932892, "epoch": 12.25, "learning_rate": 1.0169067000626175e-05, "loss": 0.5474, "step": 14497, "task_loss": 0.3512466549873352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.521388828754425, "epoch": 12.26, "learning_rate": 1.0165936130244208e-05, "loss": 0.3964, "step": 14498, "task_loss": 0.41047433018684387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4540198743343353, "epoch": 12.26, "learning_rate": 1.0162805259862241e-05, "loss": 0.4496, "step": 14499, "task_loss": 0.5711773633956909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26952749490737915, "epoch": 12.26, "learning_rate": 1.0159674389480275e-05, "loss": 0.4984, "step": 14500, "task_loss": 0.2640037536621094 }, { "epoch": 12.26, "eval_accuracy": 0.9112079207920792, "eval_loss": 0.3230462074279785, "eval_runtime": 208.2584, "eval_samples_per_second": 121.244, "eval_steps_per_second": 0.951, "step": 14500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4598454236984253, "epoch": 12.26, "learning_rate": 1.015654351909831e-05, "loss": 0.4741, "step": 14501, "task_loss": 1.5084513425827026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26540422439575195, "epoch": 12.26, "learning_rate": 1.0153412648716344e-05, "loss": 0.5379, "step": 14502, "task_loss": 0.1507975459098816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5766352415084839, "epoch": 12.26, "learning_rate": 1.0150281778334377e-05, "loss": 0.4105, "step": 14503, "task_loss": 0.9161078333854675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5222874879837036, "epoch": 12.26, "learning_rate": 1.0147150907952412e-05, "loss": 0.4917, "step": 14504, "task_loss": 0.809251606464386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20015108585357666, "epoch": 12.26, "learning_rate": 1.0144020037570446e-05, "loss": 0.3659, "step": 14505, "task_loss": 0.3396831154823303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4076424539089203, "epoch": 12.26, "learning_rate": 1.014088916718848e-05, "loss": 0.4445, "step": 14506, "task_loss": 0.5808071494102478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29095131158828735, "epoch": 12.26, "learning_rate": 1.0137758296806513e-05, "loss": 0.478, "step": 14507, "task_loss": 0.2254061996936798 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42585915327072144, "epoch": 12.26, "learning_rate": 1.0134627426424546e-05, "loss": 0.5877, "step": 14508, "task_loss": 0.8663202524185181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46065250039100647, "epoch": 12.26, "learning_rate": 1.013149655604258e-05, "loss": 0.4709, "step": 14509, "task_loss": 0.3306494355201721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41879332065582275, "epoch": 12.27, "learning_rate": 1.0128365685660613e-05, "loss": 0.4667, "step": 14510, "task_loss": 1.2878303527832031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9931045770645142, "epoch": 12.27, "learning_rate": 1.0125234815278648e-05, "loss": 0.5188, "step": 14511, "task_loss": 1.2908979654312134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4835801422595978, "epoch": 12.27, "learning_rate": 1.0122103944896682e-05, "loss": 0.6321, "step": 14512, "task_loss": 1.3245409727096558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7034839391708374, "epoch": 12.27, "learning_rate": 1.0118973074514717e-05, "loss": 0.4971, "step": 14513, "task_loss": 0.9991678595542908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7045981884002686, "epoch": 12.27, "learning_rate": 1.011584220413275e-05, "loss": 0.555, "step": 14514, "task_loss": 0.40556567907333374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46041518449783325, "epoch": 12.27, "learning_rate": 1.0112711333750784e-05, "loss": 0.5213, "step": 14515, "task_loss": 0.09152089059352875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39552241563796997, "epoch": 12.27, "learning_rate": 1.0109580463368817e-05, "loss": 0.414, "step": 14516, "task_loss": 0.2587932050228119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47272050380706787, "epoch": 12.27, "learning_rate": 1.010644959298685e-05, "loss": 0.4904, "step": 14517, "task_loss": 0.14127735793590546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2721254229545593, "epoch": 12.27, "learning_rate": 1.0103318722604884e-05, "loss": 0.4853, "step": 14518, "task_loss": 0.2894459664821625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4117918014526367, "epoch": 12.27, "learning_rate": 1.0100187852222918e-05, "loss": 0.3917, "step": 14519, "task_loss": 0.8198657631874084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4675142765045166, "epoch": 12.27, "learning_rate": 1.0097056981840953e-05, "loss": 0.3824, "step": 14520, "task_loss": 0.3673561215400696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4718136191368103, "epoch": 12.27, "learning_rate": 1.0093926111458986e-05, "loss": 0.5931, "step": 14521, "task_loss": 0.6957215666770935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.559131383895874, "epoch": 12.28, "learning_rate": 1.009079524107702e-05, "loss": 0.499, "step": 14522, "task_loss": 0.38375797867774963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28818973898887634, "epoch": 12.28, "learning_rate": 1.0087664370695053e-05, "loss": 0.3856, "step": 14523, "task_loss": 0.7117280960083008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32036256790161133, "epoch": 12.28, "learning_rate": 1.0084533500313087e-05, "loss": 0.3988, "step": 14524, "task_loss": 0.42993807792663574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.323172390460968, "epoch": 12.28, "learning_rate": 1.008140262993112e-05, "loss": 0.4493, "step": 14525, "task_loss": 0.8093869686126709 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2789170742034912, "epoch": 12.28, "learning_rate": 1.0078271759549155e-05, "loss": 0.4002, "step": 14526, "task_loss": 0.7599226832389832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2773168087005615, "epoch": 12.28, "learning_rate": 1.0075140889167189e-05, "loss": 0.3809, "step": 14527, "task_loss": 0.376405268907547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5843798518180847, "epoch": 12.28, "learning_rate": 1.0072010018785222e-05, "loss": 0.6145, "step": 14528, "task_loss": 0.15840435028076172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4041118919849396, "epoch": 12.28, "learning_rate": 1.0068879148403258e-05, "loss": 0.3651, "step": 14529, "task_loss": 0.5558462738990784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21639998257160187, "epoch": 12.28, "learning_rate": 1.0065748278021291e-05, "loss": 0.4349, "step": 14530, "task_loss": 0.21574071049690247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3545726537704468, "epoch": 12.28, "learning_rate": 1.0062617407639325e-05, "loss": 0.5112, "step": 14531, "task_loss": 0.604895830154419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4155365228652954, "epoch": 12.28, "learning_rate": 1.0059486537257358e-05, "loss": 0.4905, "step": 14532, "task_loss": 1.0379302501678467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43529796600341797, "epoch": 12.28, "learning_rate": 1.0056355666875391e-05, "loss": 0.378, "step": 14533, "task_loss": 0.29578596353530884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5206825733184814, "epoch": 12.29, "learning_rate": 1.0053224796493425e-05, "loss": 0.3813, "step": 14534, "task_loss": 0.13575197756290436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42034369707107544, "epoch": 12.29, "learning_rate": 1.0050093926111458e-05, "loss": 0.383, "step": 14535, "task_loss": 0.7625797390937805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22533157467842102, "epoch": 12.29, "learning_rate": 1.0046963055729492e-05, "loss": 0.3299, "step": 14536, "task_loss": 0.4419572055339813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3773445188999176, "epoch": 12.29, "learning_rate": 1.0043832185347527e-05, "loss": 0.4431, "step": 14537, "task_loss": 0.5341012477874756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49701938033103943, "epoch": 12.29, "learning_rate": 1.0040701314965562e-05, "loss": 0.3759, "step": 14538, "task_loss": 0.23090395331382751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4756737947463989, "epoch": 12.29, "learning_rate": 1.0037570444583596e-05, "loss": 0.4973, "step": 14539, "task_loss": 0.4227042496204376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.383512020111084, "epoch": 12.29, "learning_rate": 1.003443957420163e-05, "loss": 0.4834, "step": 14540, "task_loss": 1.7165206670761108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37886977195739746, "epoch": 12.29, "learning_rate": 1.0031308703819663e-05, "loss": 0.4667, "step": 14541, "task_loss": 0.3031805455684662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4949953556060791, "epoch": 12.29, "learning_rate": 1.0028177833437696e-05, "loss": 0.4329, "step": 14542, "task_loss": 0.8752137422561646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.597128689289093, "epoch": 12.29, "learning_rate": 1.002504696305573e-05, "loss": 0.4312, "step": 14543, "task_loss": 1.1711857318878174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47389185428619385, "epoch": 12.29, "learning_rate": 1.0021916092673763e-05, "loss": 0.4419, "step": 14544, "task_loss": 0.5139250159263611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6570245623588562, "epoch": 12.29, "learning_rate": 1.0018785222291798e-05, "loss": 0.3866, "step": 14545, "task_loss": 0.9047538638114929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.10478439927101135, "epoch": 12.3, "learning_rate": 1.0015654351909832e-05, "loss": 0.393, "step": 14546, "task_loss": 0.01945769041776657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6854181289672852, "epoch": 12.3, "learning_rate": 1.0012523481527865e-05, "loss": 0.5266, "step": 14547, "task_loss": 1.1572860479354858 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3705092668533325, "epoch": 12.3, "learning_rate": 1.0009392611145899e-05, "loss": 0.4064, "step": 14548, "task_loss": 0.480216383934021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22455236315727234, "epoch": 12.3, "learning_rate": 1.0006261740763932e-05, "loss": 0.3449, "step": 14549, "task_loss": 0.3555169105529785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6217678785324097, "epoch": 12.3, "learning_rate": 1.0003130870381967e-05, "loss": 0.5418, "step": 14550, "task_loss": 0.5852357149124146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7339543104171753, "epoch": 12.3, "learning_rate": 1e-05, "loss": 0.5388, "step": 14551, "task_loss": 0.4038277566432953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25761958956718445, "epoch": 12.3, "learning_rate": 9.996869129618034e-06, "loss": 0.3845, "step": 14552, "task_loss": 0.524608850479126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5937647819519043, "epoch": 12.3, "learning_rate": 9.993738259236068e-06, "loss": 0.4743, "step": 14553, "task_loss": 0.4629884958267212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33379894495010376, "epoch": 12.3, "learning_rate": 9.990607388854103e-06, "loss": 0.5206, "step": 14554, "task_loss": 0.6865701675415039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5873088240623474, "epoch": 12.3, "learning_rate": 9.987476518472136e-06, "loss": 0.4887, "step": 14555, "task_loss": 0.2607617676258087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5039251446723938, "epoch": 12.3, "learning_rate": 9.98434564809017e-06, "loss": 0.4543, "step": 14556, "task_loss": 0.8605638742446899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3878421485424042, "epoch": 12.3, "learning_rate": 9.981214777708203e-06, "loss": 0.4731, "step": 14557, "task_loss": 0.15139126777648926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5210468769073486, "epoch": 12.31, "learning_rate": 9.978083907326237e-06, "loss": 0.5839, "step": 14558, "task_loss": 1.6660962104797363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3442641496658325, "epoch": 12.31, "learning_rate": 9.97495303694427e-06, "loss": 0.4902, "step": 14559, "task_loss": 1.0411878824234009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3128277361392975, "epoch": 12.31, "learning_rate": 9.971822166562304e-06, "loss": 0.413, "step": 14560, "task_loss": 0.7111147046089172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.11822091042995453, "epoch": 12.31, "learning_rate": 9.968691296180337e-06, "loss": 0.2784, "step": 14561, "task_loss": 0.0031592899467796087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.356578528881073, "epoch": 12.31, "learning_rate": 9.965560425798372e-06, "loss": 0.4729, "step": 14562, "task_loss": 0.054770659655332565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7153518199920654, "epoch": 12.31, "learning_rate": 9.962429555416408e-06, "loss": 0.4567, "step": 14563, "task_loss": 0.29447317123413086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4403454661369324, "epoch": 12.31, "learning_rate": 9.959298685034441e-06, "loss": 0.402, "step": 14564, "task_loss": 0.2333456128835678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4258449077606201, "epoch": 12.31, "learning_rate": 9.956167814652475e-06, "loss": 0.4475, "step": 14565, "task_loss": 0.7003014087677002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47005489468574524, "epoch": 12.31, "learning_rate": 9.953036944270508e-06, "loss": 0.3921, "step": 14566, "task_loss": 0.863681435585022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37751859426498413, "epoch": 12.31, "learning_rate": 9.949906073888542e-06, "loss": 0.5719, "step": 14567, "task_loss": 0.7533006072044373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3518124222755432, "epoch": 12.31, "learning_rate": 9.946775203506575e-06, "loss": 0.4626, "step": 14568, "task_loss": 0.23398379981517792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44701969623565674, "epoch": 12.32, "learning_rate": 9.943644333124608e-06, "loss": 0.3892, "step": 14569, "task_loss": 0.5345175266265869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5175586938858032, "epoch": 12.32, "learning_rate": 9.940513462742642e-06, "loss": 0.3743, "step": 14570, "task_loss": 0.4404710531234741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3737456202507019, "epoch": 12.32, "learning_rate": 9.937382592360677e-06, "loss": 0.3649, "step": 14571, "task_loss": 0.9470277428627014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33431363105773926, "epoch": 12.32, "learning_rate": 9.93425172197871e-06, "loss": 0.5677, "step": 14572, "task_loss": 0.8039200901985168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4713023006916046, "epoch": 12.32, "learning_rate": 9.931120851596744e-06, "loss": 0.536, "step": 14573, "task_loss": 0.9152107834815979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27370917797088623, "epoch": 12.32, "learning_rate": 9.927989981214778e-06, "loss": 0.3755, "step": 14574, "task_loss": 0.44634807109832764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4038134813308716, "epoch": 12.32, "learning_rate": 9.924859110832813e-06, "loss": 0.3457, "step": 14575, "task_loss": 0.5954149961471558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5326566100120544, "epoch": 12.32, "learning_rate": 9.921728240450846e-06, "loss": 0.5033, "step": 14576, "task_loss": 0.2707122564315796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8373572826385498, "epoch": 12.32, "learning_rate": 9.91859737006888e-06, "loss": 0.5419, "step": 14577, "task_loss": 0.4740775227546692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42720848321914673, "epoch": 12.32, "learning_rate": 9.915466499686913e-06, "loss": 0.4782, "step": 14578, "task_loss": 0.7127529382705688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2603372037410736, "epoch": 12.32, "learning_rate": 9.912335629304947e-06, "loss": 0.5209, "step": 14579, "task_loss": 0.6596167683601379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5580909848213196, "epoch": 12.32, "learning_rate": 9.909204758922982e-06, "loss": 0.3527, "step": 14580, "task_loss": 0.779138445854187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6649438142776489, "epoch": 12.33, "learning_rate": 9.906073888541015e-06, "loss": 0.588, "step": 14581, "task_loss": 0.2990189790725708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41873130202293396, "epoch": 12.33, "learning_rate": 9.902943018159049e-06, "loss": 0.5219, "step": 14582, "task_loss": 1.1253660917282104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6432522535324097, "epoch": 12.33, "learning_rate": 9.899812147777082e-06, "loss": 0.3857, "step": 14583, "task_loss": 1.027683973312378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6593071818351746, "epoch": 12.33, "learning_rate": 9.896681277395116e-06, "loss": 0.4572, "step": 14584, "task_loss": 0.4214004576206207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7980611324310303, "epoch": 12.33, "learning_rate": 9.893550407013149e-06, "loss": 0.5813, "step": 14585, "task_loss": 1.0225154161453247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.377458393573761, "epoch": 12.33, "learning_rate": 9.890419536631183e-06, "loss": 0.4359, "step": 14586, "task_loss": 0.3142402470111847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4010908901691437, "epoch": 12.33, "learning_rate": 9.887288666249218e-06, "loss": 0.5282, "step": 14587, "task_loss": 1.1308616399765015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4228366017341614, "epoch": 12.33, "learning_rate": 9.884157795867251e-06, "loss": 0.4432, "step": 14588, "task_loss": 0.6235272884368896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2651147246360779, "epoch": 12.33, "learning_rate": 9.881026925485286e-06, "loss": 0.3333, "step": 14589, "task_loss": 0.358211874961853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46205854415893555, "epoch": 12.33, "learning_rate": 9.87789605510332e-06, "loss": 0.472, "step": 14590, "task_loss": 1.241190791130066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3688144087791443, "epoch": 12.33, "learning_rate": 9.874765184721353e-06, "loss": 0.452, "step": 14591, "task_loss": 0.5442970395088196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7647260427474976, "epoch": 12.33, "learning_rate": 9.871634314339387e-06, "loss": 0.459, "step": 14592, "task_loss": 0.497050017118454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8304710984230042, "epoch": 12.34, "learning_rate": 9.86850344395742e-06, "loss": 0.4615, "step": 14593, "task_loss": 0.9980489015579224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22805802524089813, "epoch": 12.34, "learning_rate": 9.865372573575454e-06, "loss": 0.322, "step": 14594, "task_loss": 0.4270934760570526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5816479921340942, "epoch": 12.34, "learning_rate": 9.862241703193487e-06, "loss": 0.5357, "step": 14595, "task_loss": 0.7955360412597656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4684339761734009, "epoch": 12.34, "learning_rate": 9.859110832811522e-06, "loss": 0.5133, "step": 14596, "task_loss": 0.4916442334651947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5337986350059509, "epoch": 12.34, "learning_rate": 9.855979962429556e-06, "loss": 0.4422, "step": 14597, "task_loss": 0.7394416928291321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6985774040222168, "epoch": 12.34, "learning_rate": 9.85284909204759e-06, "loss": 0.5568, "step": 14598, "task_loss": 0.7867944836616516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6276285648345947, "epoch": 12.34, "learning_rate": 9.849718221665625e-06, "loss": 0.4848, "step": 14599, "task_loss": 1.0783599615097046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5108577013015747, "epoch": 12.34, "learning_rate": 9.846587351283658e-06, "loss": 0.3603, "step": 14600, "task_loss": 0.4007719159126282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5835545063018799, "epoch": 12.34, "learning_rate": 9.843456480901692e-06, "loss": 0.5057, "step": 14601, "task_loss": 0.33065810799598694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5385329723358154, "epoch": 12.34, "learning_rate": 9.840325610519725e-06, "loss": 0.4444, "step": 14602, "task_loss": 0.8973146080970764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5698426961898804, "epoch": 12.34, "learning_rate": 9.837194740137758e-06, "loss": 0.4618, "step": 14603, "task_loss": 0.7909348607063293 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42060238122940063, "epoch": 12.34, "learning_rate": 9.834063869755792e-06, "loss": 0.4997, "step": 14604, "task_loss": 1.2903310060501099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4480874538421631, "epoch": 12.35, "learning_rate": 9.830932999373827e-06, "loss": 0.4441, "step": 14605, "task_loss": 0.7152438163757324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2684760093688965, "epoch": 12.35, "learning_rate": 9.82780212899186e-06, "loss": 0.4421, "step": 14606, "task_loss": 0.9044064879417419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41207993030548096, "epoch": 12.35, "learning_rate": 9.824671258609894e-06, "loss": 0.4526, "step": 14607, "task_loss": 0.6082610487937927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35425013303756714, "epoch": 12.35, "learning_rate": 9.821540388227928e-06, "loss": 0.4158, "step": 14608, "task_loss": 0.7137708067893982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4714858829975128, "epoch": 12.35, "learning_rate": 9.818409517845961e-06, "loss": 0.4742, "step": 14609, "task_loss": 0.5835442543029785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6468356847763062, "epoch": 12.35, "learning_rate": 9.815278647463994e-06, "loss": 0.5564, "step": 14610, "task_loss": 1.0354785919189453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5841670632362366, "epoch": 12.35, "learning_rate": 9.812147777082028e-06, "loss": 0.4216, "step": 14611, "task_loss": 0.20871593058109283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4072497487068176, "epoch": 12.35, "learning_rate": 9.809016906700063e-06, "loss": 0.4143, "step": 14612, "task_loss": 0.6521146297454834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3376009166240692, "epoch": 12.35, "learning_rate": 9.805886036318097e-06, "loss": 0.3141, "step": 14613, "task_loss": 0.027771824970841408 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8500915765762329, "epoch": 12.35, "learning_rate": 9.802755165936132e-06, "loss": 0.519, "step": 14614, "task_loss": 1.1849905252456665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5400584936141968, "epoch": 12.35, "learning_rate": 9.799624295554165e-06, "loss": 0.4575, "step": 14615, "task_loss": 0.5619030594825745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2888426184654236, "epoch": 12.35, "learning_rate": 9.796493425172199e-06, "loss": 0.4069, "step": 14616, "task_loss": 0.43865859508514404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.63481605052948, "epoch": 12.36, "learning_rate": 9.793362554790232e-06, "loss": 0.636, "step": 14617, "task_loss": 0.46654075384140015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5162043571472168, "epoch": 12.36, "learning_rate": 9.790231684408266e-06, "loss": 0.4059, "step": 14618, "task_loss": 0.23441144824028015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4185490608215332, "epoch": 12.36, "learning_rate": 9.7871008140263e-06, "loss": 0.4573, "step": 14619, "task_loss": 0.7340905070304871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3354703485965729, "epoch": 12.36, "learning_rate": 9.783969943644333e-06, "loss": 0.3415, "step": 14620, "task_loss": 0.5071995854377747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32242971658706665, "epoch": 12.36, "learning_rate": 9.780839073262368e-06, "loss": 0.3711, "step": 14621, "task_loss": 0.5025303363800049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46094629168510437, "epoch": 12.36, "learning_rate": 9.777708202880401e-06, "loss": 0.4942, "step": 14622, "task_loss": 0.9212746620178223 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31681114435195923, "epoch": 12.36, "learning_rate": 9.774577332498435e-06, "loss": 0.382, "step": 14623, "task_loss": 0.29950857162475586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44357356429100037, "epoch": 12.36, "learning_rate": 9.77144646211647e-06, "loss": 0.4232, "step": 14624, "task_loss": 1.4854581356048584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5406948328018188, "epoch": 12.36, "learning_rate": 9.768315591734503e-06, "loss": 0.539, "step": 14625, "task_loss": 0.9739474058151245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.630133867263794, "epoch": 12.36, "learning_rate": 9.765184721352537e-06, "loss": 0.5127, "step": 14626, "task_loss": 0.7882195115089417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5098047852516174, "epoch": 12.36, "learning_rate": 9.76205385097057e-06, "loss": 0.4591, "step": 14627, "task_loss": 0.6528857946395874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5079978704452515, "epoch": 12.36, "learning_rate": 9.758922980588604e-06, "loss": 0.448, "step": 14628, "task_loss": 0.6099722385406494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22728562355041504, "epoch": 12.37, "learning_rate": 9.755792110206637e-06, "loss": 0.3307, "step": 14629, "task_loss": 0.1330343782901764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4112245440483093, "epoch": 12.37, "learning_rate": 9.752661239824672e-06, "loss": 0.5032, "step": 14630, "task_loss": 0.11709866672754288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22041966021060944, "epoch": 12.37, "learning_rate": 9.749530369442706e-06, "loss": 0.3527, "step": 14631, "task_loss": 0.2695468068122864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31472906470298767, "epoch": 12.37, "learning_rate": 9.74639949906074e-06, "loss": 0.4248, "step": 14632, "task_loss": 0.711920440196991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5463173389434814, "epoch": 12.37, "learning_rate": 9.743268628678773e-06, "loss": 0.4901, "step": 14633, "task_loss": 0.45144322514533997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4406052529811859, "epoch": 12.37, "learning_rate": 9.740137758296806e-06, "loss": 0.6242, "step": 14634, "task_loss": 0.21091362833976746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42399442195892334, "epoch": 12.37, "learning_rate": 9.73700688791484e-06, "loss": 0.5736, "step": 14635, "task_loss": 0.8575429916381836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47509512305259705, "epoch": 12.37, "learning_rate": 9.733876017532875e-06, "loss": 0.5562, "step": 14636, "task_loss": 0.4169047772884369 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3030160963535309, "epoch": 12.37, "learning_rate": 9.730745147150908e-06, "loss": 0.3903, "step": 14637, "task_loss": 0.06099078431725502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6846891045570374, "epoch": 12.37, "learning_rate": 9.727614276768942e-06, "loss": 0.4957, "step": 14638, "task_loss": 0.1715846061706543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30436769127845764, "epoch": 12.37, "learning_rate": 9.724483406386977e-06, "loss": 0.3328, "step": 14639, "task_loss": 0.6216034889221191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5342002511024475, "epoch": 12.38, "learning_rate": 9.72135253600501e-06, "loss": 0.4352, "step": 14640, "task_loss": 1.5380922555923462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5149276256561279, "epoch": 12.38, "learning_rate": 9.718221665623044e-06, "loss": 0.7101, "step": 14641, "task_loss": 1.1941050291061401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.305865079164505, "epoch": 12.38, "learning_rate": 9.715090795241078e-06, "loss": 0.448, "step": 14642, "task_loss": 0.20444296300411224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5604305267333984, "epoch": 12.38, "learning_rate": 9.711959924859111e-06, "loss": 0.5486, "step": 14643, "task_loss": 1.3393365144729614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2830623686313629, "epoch": 12.38, "learning_rate": 9.708829054477145e-06, "loss": 0.3578, "step": 14644, "task_loss": 0.20703378319740295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5267239212989807, "epoch": 12.38, "learning_rate": 9.705698184095178e-06, "loss": 0.5139, "step": 14645, "task_loss": 1.1764203310012817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5067201256752014, "epoch": 12.38, "learning_rate": 9.702567313713211e-06, "loss": 0.4035, "step": 14646, "task_loss": 0.6470180153846741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25027093291282654, "epoch": 12.38, "learning_rate": 9.699436443331247e-06, "loss": 0.2157, "step": 14647, "task_loss": 0.47575843334198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6212410926818848, "epoch": 12.38, "learning_rate": 9.696305572949282e-06, "loss": 0.468, "step": 14648, "task_loss": 1.1924700736999512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41332608461380005, "epoch": 12.38, "learning_rate": 9.693174702567315e-06, "loss": 0.5759, "step": 14649, "task_loss": 0.39004701375961304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4155759811401367, "epoch": 12.38, "learning_rate": 9.690043832185349e-06, "loss": 0.3627, "step": 14650, "task_loss": 0.6793668866157532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20191055536270142, "epoch": 12.38, "learning_rate": 9.686912961803382e-06, "loss": 0.294, "step": 14651, "task_loss": 0.03252948075532913 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7466074824333191, "epoch": 12.39, "learning_rate": 9.683782091421416e-06, "loss": 0.5315, "step": 14652, "task_loss": 0.720479428768158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4304879903793335, "epoch": 12.39, "learning_rate": 9.68065122103945e-06, "loss": 0.5743, "step": 14653, "task_loss": 0.35586726665496826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6512069702148438, "epoch": 12.39, "learning_rate": 9.677520350657483e-06, "loss": 0.4158, "step": 14654, "task_loss": 1.739267110824585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6301002502441406, "epoch": 12.39, "learning_rate": 9.674389480275516e-06, "loss": 0.5386, "step": 14655, "task_loss": 0.9066369533538818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42994415760040283, "epoch": 12.39, "learning_rate": 9.671258609893551e-06, "loss": 0.4305, "step": 14656, "task_loss": 1.2509936094284058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6233416795730591, "epoch": 12.39, "learning_rate": 9.668127739511585e-06, "loss": 0.6312, "step": 14657, "task_loss": 0.9718807339668274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1944746971130371, "epoch": 12.39, "learning_rate": 9.664996869129618e-06, "loss": 0.2846, "step": 14658, "task_loss": 0.02547280117869377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34370681643486023, "epoch": 12.39, "learning_rate": 9.661865998747652e-06, "loss": 0.3827, "step": 14659, "task_loss": 0.4276195466518402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4661281704902649, "epoch": 12.39, "learning_rate": 9.658735128365685e-06, "loss": 0.4538, "step": 14660, "task_loss": 0.5934279561042786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42902278900146484, "epoch": 12.39, "learning_rate": 9.65560425798372e-06, "loss": 0.4271, "step": 14661, "task_loss": 0.25998833775520325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6767672300338745, "epoch": 12.39, "learning_rate": 9.652473387601754e-06, "loss": 0.505, "step": 14662, "task_loss": 0.37933090329170227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36435651779174805, "epoch": 12.39, "learning_rate": 9.649342517219787e-06, "loss": 0.4812, "step": 14663, "task_loss": 0.4493657946586609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24385309219360352, "epoch": 12.4, "learning_rate": 9.64621164683782e-06, "loss": 0.3649, "step": 14664, "task_loss": 0.09016034007072449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7486042976379395, "epoch": 12.4, "learning_rate": 9.643080776455856e-06, "loss": 0.4724, "step": 14665, "task_loss": 1.1427284479141235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2991369366645813, "epoch": 12.4, "learning_rate": 9.63994990607389e-06, "loss": 0.4384, "step": 14666, "task_loss": 0.1348358690738678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34151312708854675, "epoch": 12.4, "learning_rate": 9.636819035691923e-06, "loss": 0.3716, "step": 14667, "task_loss": 0.3809586763381958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5027100443840027, "epoch": 12.4, "learning_rate": 9.633688165309956e-06, "loss": 0.4536, "step": 14668, "task_loss": 0.6628871560096741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39671826362609863, "epoch": 12.4, "learning_rate": 9.63055729492799e-06, "loss": 0.3783, "step": 14669, "task_loss": 0.7762985825538635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5504164099693298, "epoch": 12.4, "learning_rate": 9.627426424546023e-06, "loss": 0.5058, "step": 14670, "task_loss": 0.5127440094947815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5288217067718506, "epoch": 12.4, "learning_rate": 9.624295554164057e-06, "loss": 0.5149, "step": 14671, "task_loss": 0.4127315580844879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.13443197309970856, "epoch": 12.4, "learning_rate": 9.621164683782092e-06, "loss": 0.3777, "step": 14672, "task_loss": 0.038654591888189316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31835269927978516, "epoch": 12.4, "learning_rate": 9.618033813400127e-06, "loss": 0.3828, "step": 14673, "task_loss": 0.48988431692123413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19893494248390198, "epoch": 12.4, "learning_rate": 9.61490294301816e-06, "loss": 0.4096, "step": 14674, "task_loss": 0.38575831055641174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5929683446884155, "epoch": 12.4, "learning_rate": 9.611772072636194e-06, "loss": 0.6687, "step": 14675, "task_loss": 1.1739789247512817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3107295632362366, "epoch": 12.41, "learning_rate": 9.608641202254228e-06, "loss": 0.3805, "step": 14676, "task_loss": 0.16289593279361725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20965439081192017, "epoch": 12.41, "learning_rate": 9.605510331872261e-06, "loss": 0.2337, "step": 14677, "task_loss": 0.20043714344501495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5195056796073914, "epoch": 12.41, "learning_rate": 9.602379461490295e-06, "loss": 0.5476, "step": 14678, "task_loss": 0.3178323805332184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5218809843063354, "epoch": 12.41, "learning_rate": 9.599248591108328e-06, "loss": 0.5445, "step": 14679, "task_loss": 0.9357633590698242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6424537897109985, "epoch": 12.41, "learning_rate": 9.596117720726361e-06, "loss": 0.4427, "step": 14680, "task_loss": 0.2480962574481964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1660028100013733, "epoch": 12.41, "learning_rate": 9.592986850344397e-06, "loss": 0.3356, "step": 14681, "task_loss": 0.08022197335958481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2944008409976959, "epoch": 12.41, "learning_rate": 9.58985597996243e-06, "loss": 0.3327, "step": 14682, "task_loss": 0.19394458830356598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4209335744380951, "epoch": 12.41, "learning_rate": 9.586725109580464e-06, "loss": 0.5712, "step": 14683, "task_loss": 0.23998570442199707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5634985566139221, "epoch": 12.41, "learning_rate": 9.583594239198497e-06, "loss": 0.4619, "step": 14684, "task_loss": 0.7807695269584656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15960168838500977, "epoch": 12.41, "learning_rate": 9.580463368816532e-06, "loss": 0.2972, "step": 14685, "task_loss": 0.03930402174592018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2224849909543991, "epoch": 12.41, "learning_rate": 9.577332498434566e-06, "loss": 0.4644, "step": 14686, "task_loss": 0.34739094972610474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40588581562042236, "epoch": 12.41, "learning_rate": 9.5742016280526e-06, "loss": 0.5299, "step": 14687, "task_loss": 0.2770046293735504 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40965020656585693, "epoch": 12.42, "learning_rate": 9.571070757670633e-06, "loss": 0.4455, "step": 14688, "task_loss": 0.6112362146377563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3977915048599243, "epoch": 12.42, "learning_rate": 9.567939887288666e-06, "loss": 0.4702, "step": 14689, "task_loss": 0.3116898238658905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3956308960914612, "epoch": 12.42, "learning_rate": 9.564809016906701e-06, "loss": 0.4802, "step": 14690, "task_loss": 0.5766125917434692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.280308336019516, "epoch": 12.42, "learning_rate": 9.561678146524735e-06, "loss": 0.5091, "step": 14691, "task_loss": 0.5423598885536194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2877221405506134, "epoch": 12.42, "learning_rate": 9.558547276142768e-06, "loss": 0.5014, "step": 14692, "task_loss": 0.3525705337524414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25418004393577576, "epoch": 12.42, "learning_rate": 9.555416405760802e-06, "loss": 0.3708, "step": 14693, "task_loss": 0.47025999426841736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3530370593070984, "epoch": 12.42, "learning_rate": 9.552285535378835e-06, "loss": 0.519, "step": 14694, "task_loss": 0.5205552577972412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40140771865844727, "epoch": 12.42, "learning_rate": 9.549154664996869e-06, "loss": 0.5383, "step": 14695, "task_loss": 0.8942793011665344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0197367668151855, "epoch": 12.42, "learning_rate": 9.546023794614902e-06, "loss": 0.6198, "step": 14696, "task_loss": 0.9088011980056763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34911832213401794, "epoch": 12.42, "learning_rate": 9.542892924232937e-06, "loss": 0.393, "step": 14697, "task_loss": 1.264119029045105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3217768967151642, "epoch": 12.42, "learning_rate": 9.53976205385097e-06, "loss": 0.5089, "step": 14698, "task_loss": 0.7098134756088257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49779653549194336, "epoch": 12.42, "learning_rate": 9.536631183469006e-06, "loss": 0.4197, "step": 14699, "task_loss": 0.2708057463169098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48438316583633423, "epoch": 12.43, "learning_rate": 9.53350031308704e-06, "loss": 0.3927, "step": 14700, "task_loss": 0.16387489438056946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7375249266624451, "epoch": 12.43, "learning_rate": 9.530369442705073e-06, "loss": 0.7019, "step": 14701, "task_loss": 0.7715138792991638 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4997698962688446, "epoch": 12.43, "learning_rate": 9.527238572323106e-06, "loss": 0.5343, "step": 14702, "task_loss": 0.8951702117919922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41002216935157776, "epoch": 12.43, "learning_rate": 9.52410770194114e-06, "loss": 0.5324, "step": 14703, "task_loss": 0.1820056289434433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47828397154808044, "epoch": 12.43, "learning_rate": 9.520976831559173e-06, "loss": 0.4476, "step": 14704, "task_loss": 0.5142514109611511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33517986536026, "epoch": 12.43, "learning_rate": 9.517845961177207e-06, "loss": 0.5345, "step": 14705, "task_loss": 0.30157166719436646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33635851740837097, "epoch": 12.43, "learning_rate": 9.514715090795242e-06, "loss": 0.4317, "step": 14706, "task_loss": 0.3909105360507965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.658149003982544, "epoch": 12.43, "learning_rate": 9.511584220413275e-06, "loss": 0.4768, "step": 14707, "task_loss": 0.49377384781837463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4715956449508667, "epoch": 12.43, "learning_rate": 9.508453350031309e-06, "loss": 0.4893, "step": 14708, "task_loss": 0.5213782787322998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6918294429779053, "epoch": 12.43, "learning_rate": 9.505322479649342e-06, "loss": 0.4636, "step": 14709, "task_loss": 1.0169475078582764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4416685700416565, "epoch": 12.43, "learning_rate": 9.502191609267378e-06, "loss": 0.4828, "step": 14710, "task_loss": 1.2408461570739746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6740844249725342, "epoch": 12.44, "learning_rate": 9.499060738885411e-06, "loss": 0.5616, "step": 14711, "task_loss": 0.4483846426010132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.555360734462738, "epoch": 12.44, "learning_rate": 9.495929868503445e-06, "loss": 0.4829, "step": 14712, "task_loss": 1.0016950368881226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3836659789085388, "epoch": 12.44, "learning_rate": 9.492798998121478e-06, "loss": 0.4405, "step": 14713, "task_loss": 0.1776493638753891 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3667300343513489, "epoch": 12.44, "learning_rate": 9.489668127739511e-06, "loss": 0.4547, "step": 14714, "task_loss": 1.446820855140686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34350311756134033, "epoch": 12.44, "learning_rate": 9.486537257357547e-06, "loss": 0.3629, "step": 14715, "task_loss": 0.9705151915550232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9336172938346863, "epoch": 12.44, "learning_rate": 9.48340638697558e-06, "loss": 0.5345, "step": 14716, "task_loss": 1.4975101947784424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3528476059436798, "epoch": 12.44, "learning_rate": 9.480275516593614e-06, "loss": 0.4568, "step": 14717, "task_loss": 0.6302543878555298 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5202656984329224, "epoch": 12.44, "learning_rate": 9.477144646211647e-06, "loss": 0.5818, "step": 14718, "task_loss": 0.303771436214447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3514782786369324, "epoch": 12.44, "learning_rate": 9.47401377582968e-06, "loss": 0.4715, "step": 14719, "task_loss": 0.1686304807662964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29416295886039734, "epoch": 12.44, "learning_rate": 9.470882905447714e-06, "loss": 0.5413, "step": 14720, "task_loss": 0.7457135319709778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29905009269714355, "epoch": 12.44, "learning_rate": 9.467752035065748e-06, "loss": 0.4541, "step": 14721, "task_loss": 0.40872201323509216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4388824999332428, "epoch": 12.44, "learning_rate": 9.464621164683783e-06, "loss": 0.4521, "step": 14722, "task_loss": 0.3378850221633911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37401193380355835, "epoch": 12.45, "learning_rate": 9.461490294301816e-06, "loss": 0.4141, "step": 14723, "task_loss": 0.5091358423233032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5174989700317383, "epoch": 12.45, "learning_rate": 9.458359423919851e-06, "loss": 0.445, "step": 14724, "task_loss": 0.7709438800811768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4937956929206848, "epoch": 12.45, "learning_rate": 9.455228553537885e-06, "loss": 0.3895, "step": 14725, "task_loss": 1.0100135803222656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1652224361896515, "epoch": 12.45, "learning_rate": 9.452097683155918e-06, "loss": 0.3328, "step": 14726, "task_loss": 0.11145079135894775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7828906178474426, "epoch": 12.45, "learning_rate": 9.448966812773952e-06, "loss": 0.5431, "step": 14727, "task_loss": 0.7964751124382019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3485608398914337, "epoch": 12.45, "learning_rate": 9.445835942391985e-06, "loss": 0.4946, "step": 14728, "task_loss": 0.7106661796569824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19091659784317017, "epoch": 12.45, "learning_rate": 9.442705072010019e-06, "loss": 0.3139, "step": 14729, "task_loss": 0.0290604867041111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49220603704452515, "epoch": 12.45, "learning_rate": 9.439574201628052e-06, "loss": 0.5273, "step": 14730, "task_loss": 0.9913598299026489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28971412777900696, "epoch": 12.45, "learning_rate": 9.436443331246086e-06, "loss": 0.3745, "step": 14731, "task_loss": 0.4848233163356781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4716796576976776, "epoch": 12.45, "learning_rate": 9.43331246086412e-06, "loss": 0.4201, "step": 14732, "task_loss": 2.1291587352752686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33806532621383667, "epoch": 12.45, "learning_rate": 9.430181590482154e-06, "loss": 0.4063, "step": 14733, "task_loss": 0.27196282148361206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42938530445098877, "epoch": 12.45, "learning_rate": 9.42705072010019e-06, "loss": 0.4614, "step": 14734, "task_loss": 0.6874755620956421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5388944745063782, "epoch": 12.46, "learning_rate": 9.423919849718223e-06, "loss": 0.4367, "step": 14735, "task_loss": 0.5549059510231018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2121753692626953, "epoch": 12.46, "learning_rate": 9.420788979336256e-06, "loss": 0.3498, "step": 14736, "task_loss": 0.3075242340564728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3643001317977905, "epoch": 12.46, "learning_rate": 9.41765810895429e-06, "loss": 0.4602, "step": 14737, "task_loss": 0.3658583462238312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4629366993904114, "epoch": 12.46, "learning_rate": 9.414527238572323e-06, "loss": 0.4913, "step": 14738, "task_loss": 0.41415727138519287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48514366149902344, "epoch": 12.46, "learning_rate": 9.411396368190357e-06, "loss": 0.5198, "step": 14739, "task_loss": 0.5795744061470032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29062122106552124, "epoch": 12.46, "learning_rate": 9.40826549780839e-06, "loss": 0.3481, "step": 14740, "task_loss": 0.662377655506134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5343993902206421, "epoch": 12.46, "learning_rate": 9.405134627426425e-06, "loss": 0.419, "step": 14741, "task_loss": 0.5455877780914307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28928765654563904, "epoch": 12.46, "learning_rate": 9.402003757044459e-06, "loss": 0.4293, "step": 14742, "task_loss": 0.43869298696517944 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4269055426120758, "epoch": 12.46, "learning_rate": 9.398872886662492e-06, "loss": 0.3969, "step": 14743, "task_loss": 0.8294216990470886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9751777648925781, "epoch": 12.46, "learning_rate": 9.395742016280526e-06, "loss": 0.4931, "step": 14744, "task_loss": 1.0532697439193726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6504216194152832, "epoch": 12.46, "learning_rate": 9.39261114589856e-06, "loss": 0.5575, "step": 14745, "task_loss": 0.8153703212738037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6969987154006958, "epoch": 12.46, "learning_rate": 9.389480275516593e-06, "loss": 0.3946, "step": 14746, "task_loss": 1.3281341791152954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7146795988082886, "epoch": 12.47, "learning_rate": 9.386349405134628e-06, "loss": 0.5695, "step": 14747, "task_loss": 1.8197622299194336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2572377622127533, "epoch": 12.47, "learning_rate": 9.383218534752662e-06, "loss": 0.3668, "step": 14748, "task_loss": 0.1913469135761261 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5378312468528748, "epoch": 12.47, "learning_rate": 9.380087664370697e-06, "loss": 0.4284, "step": 14749, "task_loss": 1.1747208833694458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.619360089302063, "epoch": 12.47, "learning_rate": 9.37695679398873e-06, "loss": 0.5563, "step": 14750, "task_loss": 0.41669991612434387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2337883710861206, "epoch": 12.47, "learning_rate": 9.373825923606764e-06, "loss": 0.4426, "step": 14751, "task_loss": 0.3589722216129303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3447960615158081, "epoch": 12.47, "learning_rate": 9.370695053224797e-06, "loss": 0.4329, "step": 14752, "task_loss": 0.28791600465774536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31859976053237915, "epoch": 12.47, "learning_rate": 9.36756418284283e-06, "loss": 0.3042, "step": 14753, "task_loss": 0.3572631776332855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5608527660369873, "epoch": 12.47, "learning_rate": 9.364433312460864e-06, "loss": 0.4605, "step": 14754, "task_loss": 0.37427449226379395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4765760600566864, "epoch": 12.47, "learning_rate": 9.361302442078898e-06, "loss": 0.4275, "step": 14755, "task_loss": 0.4654463529586792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5343742370605469, "epoch": 12.47, "learning_rate": 9.358171571696931e-06, "loss": 0.5094, "step": 14756, "task_loss": 1.0607335567474365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4404110312461853, "epoch": 12.47, "learning_rate": 9.355040701314966e-06, "loss": 0.4531, "step": 14757, "task_loss": 0.6221530437469482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3925117552280426, "epoch": 12.47, "learning_rate": 9.351909830933e-06, "loss": 0.4115, "step": 14758, "task_loss": 0.31641948223114014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.314395546913147, "epoch": 12.48, "learning_rate": 9.348778960551035e-06, "loss": 0.3065, "step": 14759, "task_loss": 0.13915862143039703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2686830759048462, "epoch": 12.48, "learning_rate": 9.345648090169068e-06, "loss": 0.4494, "step": 14760, "task_loss": 0.27264153957366943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4587165117263794, "epoch": 12.48, "learning_rate": 9.342517219787102e-06, "loss": 0.4427, "step": 14761, "task_loss": 1.1759498119354248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25527793169021606, "epoch": 12.48, "learning_rate": 9.339386349405135e-06, "loss": 0.5057, "step": 14762, "task_loss": 1.2347959280014038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5503698587417603, "epoch": 12.48, "learning_rate": 9.336255479023169e-06, "loss": 0.5601, "step": 14763, "task_loss": 1.3858389854431152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 1.0298269987106323, "epoch": 12.48, "learning_rate": 9.333124608641202e-06, "loss": 0.5806, "step": 14764, "task_loss": 0.20229163765907288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42258068919181824, "epoch": 12.48, "learning_rate": 9.329993738259236e-06, "loss": 0.4469, "step": 14765, "task_loss": 0.32776740193367004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46958276629447937, "epoch": 12.48, "learning_rate": 9.326862867877271e-06, "loss": 0.5523, "step": 14766, "task_loss": 0.969241738319397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6538046598434448, "epoch": 12.48, "learning_rate": 9.323731997495304e-06, "loss": 0.5552, "step": 14767, "task_loss": 0.8257696032524109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3902484178543091, "epoch": 12.48, "learning_rate": 9.320601127113338e-06, "loss": 0.4234, "step": 14768, "task_loss": 0.29135861992836 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5277549624443054, "epoch": 12.48, "learning_rate": 9.317470256731371e-06, "loss": 0.5901, "step": 14769, "task_loss": 0.5943080186843872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18280300498008728, "epoch": 12.48, "learning_rate": 9.314339386349405e-06, "loss": 0.389, "step": 14770, "task_loss": 0.4441087543964386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47903358936309814, "epoch": 12.49, "learning_rate": 9.31120851596744e-06, "loss": 0.3376, "step": 14771, "task_loss": 0.6071392297744751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27390629053115845, "epoch": 12.49, "learning_rate": 9.308077645585473e-06, "loss": 0.4219, "step": 14772, "task_loss": 0.8729051947593689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2938857674598694, "epoch": 12.49, "learning_rate": 9.304946775203507e-06, "loss": 0.3999, "step": 14773, "task_loss": 0.5005175471305847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27590954303741455, "epoch": 12.49, "learning_rate": 9.30181590482154e-06, "loss": 0.4467, "step": 14774, "task_loss": 0.3048876225948334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6292552351951599, "epoch": 12.49, "learning_rate": 9.298685034439576e-06, "loss": 0.5503, "step": 14775, "task_loss": 1.0168445110321045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3161303699016571, "epoch": 12.49, "learning_rate": 9.295554164057609e-06, "loss": 0.4623, "step": 14776, "task_loss": 0.1794702559709549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4448234438896179, "epoch": 12.49, "learning_rate": 9.292423293675642e-06, "loss": 0.5079, "step": 14777, "task_loss": 0.2850111722946167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38823816180229187, "epoch": 12.49, "learning_rate": 9.289292423293676e-06, "loss": 0.4297, "step": 14778, "task_loss": 0.7041850090026855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5654295682907104, "epoch": 12.49, "learning_rate": 9.28616155291171e-06, "loss": 0.5877, "step": 14779, "task_loss": 0.39959582686424255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3978337049484253, "epoch": 12.49, "learning_rate": 9.283030682529743e-06, "loss": 0.4826, "step": 14780, "task_loss": 0.28038275241851807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3381667137145996, "epoch": 12.49, "learning_rate": 9.279899812147776e-06, "loss": 0.4336, "step": 14781, "task_loss": 0.8681100606918335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20870530605316162, "epoch": 12.5, "learning_rate": 9.276768941765812e-06, "loss": 0.4858, "step": 14782, "task_loss": 0.731257438659668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4148085117340088, "epoch": 12.5, "learning_rate": 9.273638071383845e-06, "loss": 0.4073, "step": 14783, "task_loss": 1.3033674955368042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7664452791213989, "epoch": 12.5, "learning_rate": 9.27050720100188e-06, "loss": 0.5217, "step": 14784, "task_loss": 0.20467814803123474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35447439551353455, "epoch": 12.5, "learning_rate": 9.267376330619914e-06, "loss": 0.3945, "step": 14785, "task_loss": 0.4034945070743561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5855526924133301, "epoch": 12.5, "learning_rate": 9.264245460237947e-06, "loss": 0.6158, "step": 14786, "task_loss": 2.5622410774230957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5133451223373413, "epoch": 12.5, "learning_rate": 9.26111458985598e-06, "loss": 0.3883, "step": 14787, "task_loss": 1.1576846837997437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36307042837142944, "epoch": 12.5, "learning_rate": 9.257983719474014e-06, "loss": 0.4625, "step": 14788, "task_loss": 0.17004548013210297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26710015535354614, "epoch": 12.5, "learning_rate": 9.254852849092048e-06, "loss": 0.3527, "step": 14789, "task_loss": 0.3383556604385376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3150787353515625, "epoch": 12.5, "learning_rate": 9.251721978710081e-06, "loss": 0.3632, "step": 14790, "task_loss": 0.46556973457336426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4971507787704468, "epoch": 12.5, "learning_rate": 9.248591108328116e-06, "loss": 0.3812, "step": 14791, "task_loss": 0.4647805392742157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5970168113708496, "epoch": 12.5, "learning_rate": 9.24546023794615e-06, "loss": 0.5116, "step": 14792, "task_loss": 0.7078662514686584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31826290488243103, "epoch": 12.5, "learning_rate": 9.242329367564183e-06, "loss": 0.4577, "step": 14793, "task_loss": 0.43603435158729553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5124253630638123, "epoch": 12.51, "learning_rate": 9.239198497182217e-06, "loss": 0.4393, "step": 14794, "task_loss": 0.6310442090034485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49633485078811646, "epoch": 12.51, "learning_rate": 9.23606762680025e-06, "loss": 0.4703, "step": 14795, "task_loss": 0.7685193419456482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4141228199005127, "epoch": 12.51, "learning_rate": 9.232936756418285e-06, "loss": 0.4991, "step": 14796, "task_loss": 0.6475130319595337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5477544665336609, "epoch": 12.51, "learning_rate": 9.229805886036319e-06, "loss": 0.5954, "step": 14797, "task_loss": 1.1237543821334839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31644219160079956, "epoch": 12.51, "learning_rate": 9.226675015654352e-06, "loss": 0.3543, "step": 14798, "task_loss": 0.020511195063591003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5590564012527466, "epoch": 12.51, "learning_rate": 9.223544145272386e-06, "loss": 0.4516, "step": 14799, "task_loss": 0.26460370421409607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3406335115432739, "epoch": 12.51, "learning_rate": 9.220413274890421e-06, "loss": 0.5747, "step": 14800, "task_loss": 0.5486583709716797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19640100002288818, "epoch": 12.51, "learning_rate": 9.217282404508454e-06, "loss": 0.3582, "step": 14801, "task_loss": 0.18767358362674713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6727412343025208, "epoch": 12.51, "learning_rate": 9.214151534126488e-06, "loss": 0.502, "step": 14802, "task_loss": 1.3470044136047363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30053409934043884, "epoch": 12.51, "learning_rate": 9.211020663744521e-06, "loss": 0.3873, "step": 14803, "task_loss": 0.7502648830413818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38175174593925476, "epoch": 12.51, "learning_rate": 9.207889793362555e-06, "loss": 0.4816, "step": 14804, "task_loss": 1.0261878967285156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27959558367729187, "epoch": 12.51, "learning_rate": 9.204758922980588e-06, "loss": 0.4181, "step": 14805, "task_loss": 0.7289585471153259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39650458097457886, "epoch": 12.52, "learning_rate": 9.201628052598622e-06, "loss": 0.5371, "step": 14806, "task_loss": 0.7591821551322937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30708885192871094, "epoch": 12.52, "learning_rate": 9.198497182216655e-06, "loss": 0.3903, "step": 14807, "task_loss": 0.8028452396392822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30306535959243774, "epoch": 12.52, "learning_rate": 9.19536631183469e-06, "loss": 0.4364, "step": 14808, "task_loss": 0.20523175597190857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.524661660194397, "epoch": 12.52, "learning_rate": 9.192235441452726e-06, "loss": 0.4571, "step": 14809, "task_loss": 0.4419706463813782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6500345468521118, "epoch": 12.52, "learning_rate": 9.189104571070759e-06, "loss": 0.4295, "step": 14810, "task_loss": 0.2693732976913452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4778960347175598, "epoch": 12.52, "learning_rate": 9.185973700688792e-06, "loss": 0.4775, "step": 14811, "task_loss": 0.7072603106498718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5730737447738647, "epoch": 12.52, "learning_rate": 9.182842830306826e-06, "loss": 0.4864, "step": 14812, "task_loss": 0.9676092267036438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46137577295303345, "epoch": 12.52, "learning_rate": 9.17971195992486e-06, "loss": 0.4563, "step": 14813, "task_loss": 0.9787250757217407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7786381244659424, "epoch": 12.52, "learning_rate": 9.176581089542893e-06, "loss": 0.528, "step": 14814, "task_loss": 0.8841501474380493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7144179344177246, "epoch": 12.52, "learning_rate": 9.173450219160926e-06, "loss": 0.4977, "step": 14815, "task_loss": 0.6124595403671265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4053078293800354, "epoch": 12.52, "learning_rate": 9.17031934877896e-06, "loss": 0.3875, "step": 14816, "task_loss": 1.1427698135375977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34539926052093506, "epoch": 12.52, "learning_rate": 9.167188478396995e-06, "loss": 0.457, "step": 14817, "task_loss": 0.885179877281189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2574256360530853, "epoch": 12.53, "learning_rate": 9.164057608015029e-06, "loss": 0.4714, "step": 14818, "task_loss": 0.06290093064308167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4153746962547302, "epoch": 12.53, "learning_rate": 9.160926737633062e-06, "loss": 0.3892, "step": 14819, "task_loss": 0.6476618647575378 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6739339232444763, "epoch": 12.53, "learning_rate": 9.157795867251097e-06, "loss": 0.6893, "step": 14820, "task_loss": 0.6069790720939636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4405362010002136, "epoch": 12.53, "learning_rate": 9.15466499686913e-06, "loss": 0.4039, "step": 14821, "task_loss": 0.7054619789123535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42134663462638855, "epoch": 12.53, "learning_rate": 9.151534126487164e-06, "loss": 0.5107, "step": 14822, "task_loss": 1.0117172002792358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37993183732032776, "epoch": 12.53, "learning_rate": 9.148403256105198e-06, "loss": 0.4709, "step": 14823, "task_loss": 0.23894810676574707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6472810506820679, "epoch": 12.53, "learning_rate": 9.145272385723231e-06, "loss": 0.4472, "step": 14824, "task_loss": 0.3982540965080261 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4686659574508667, "epoch": 12.53, "learning_rate": 9.142141515341266e-06, "loss": 0.4767, "step": 14825, "task_loss": 0.06919242441654205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5461742281913757, "epoch": 12.53, "learning_rate": 9.1390106449593e-06, "loss": 0.5926, "step": 14826, "task_loss": 0.6300997734069824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4741583466529846, "epoch": 12.53, "learning_rate": 9.135879774577333e-06, "loss": 0.3444, "step": 14827, "task_loss": 0.25290846824645996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2842337489128113, "epoch": 12.53, "learning_rate": 9.132748904195367e-06, "loss": 0.4933, "step": 14828, "task_loss": 0.906880259513855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2514612078666687, "epoch": 12.53, "learning_rate": 9.1296180338134e-06, "loss": 0.3341, "step": 14829, "task_loss": 0.6692126989364624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4997767508029938, "epoch": 12.54, "learning_rate": 9.126487163431434e-06, "loss": 0.5567, "step": 14830, "task_loss": 1.0277732610702515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2992899417877197, "epoch": 12.54, "learning_rate": 9.123356293049467e-06, "loss": 0.3704, "step": 14831, "task_loss": 0.5155743956565857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4006918966770172, "epoch": 12.54, "learning_rate": 9.120225422667502e-06, "loss": 0.4708, "step": 14832, "task_loss": 0.5859904885292053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31008055806159973, "epoch": 12.54, "learning_rate": 9.117094552285536e-06, "loss": 0.4545, "step": 14833, "task_loss": 1.403303623199463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3254470229148865, "epoch": 12.54, "learning_rate": 9.113963681903571e-06, "loss": 0.4906, "step": 14834, "task_loss": 0.5497162938117981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3262040913105011, "epoch": 12.54, "learning_rate": 9.110832811521604e-06, "loss": 0.3102, "step": 14835, "task_loss": 0.5727691650390625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5231195092201233, "epoch": 12.54, "learning_rate": 9.107701941139638e-06, "loss": 0.4676, "step": 14836, "task_loss": 0.32439523935317993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.410333514213562, "epoch": 12.54, "learning_rate": 9.104571070757671e-06, "loss": 0.3882, "step": 14837, "task_loss": 0.921076238155365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15760651230812073, "epoch": 12.54, "learning_rate": 9.101440200375705e-06, "loss": 0.3978, "step": 14838, "task_loss": 0.32616233825683594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42282819747924805, "epoch": 12.54, "learning_rate": 9.098309329993738e-06, "loss": 0.4338, "step": 14839, "task_loss": 0.5304746031761169 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.210385262966156, "epoch": 12.54, "learning_rate": 9.095178459611772e-06, "loss": 0.4242, "step": 14840, "task_loss": 0.40476536750793457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.570370614528656, "epoch": 12.54, "learning_rate": 9.092047589229805e-06, "loss": 0.4641, "step": 14841, "task_loss": 0.9843732118606567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4609305262565613, "epoch": 12.55, "learning_rate": 9.08891671884784e-06, "loss": 0.4279, "step": 14842, "task_loss": 0.43809637427330017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3463022708892822, "epoch": 12.55, "learning_rate": 9.085785848465874e-06, "loss": 0.5245, "step": 14843, "task_loss": 1.1386363506317139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3915640115737915, "epoch": 12.55, "learning_rate": 9.082654978083907e-06, "loss": 0.3639, "step": 14844, "task_loss": 0.3761036694049835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2670002579689026, "epoch": 12.55, "learning_rate": 9.079524107701943e-06, "loss": 0.3845, "step": 14845, "task_loss": 0.4945778250694275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37853819131851196, "epoch": 12.55, "learning_rate": 9.076393237319976e-06, "loss": 0.6093, "step": 14846, "task_loss": 0.3354540169239044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5245522260665894, "epoch": 12.55, "learning_rate": 9.07326236693801e-06, "loss": 0.4019, "step": 14847, "task_loss": 0.7204300165176392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5145204067230225, "epoch": 12.55, "learning_rate": 9.070131496556043e-06, "loss": 0.5326, "step": 14848, "task_loss": 0.8969506025314331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23125410079956055, "epoch": 12.55, "learning_rate": 9.067000626174076e-06, "loss": 0.3388, "step": 14849, "task_loss": 0.18930307030677795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23507562279701233, "epoch": 12.55, "learning_rate": 9.06386975579211e-06, "loss": 0.4336, "step": 14850, "task_loss": 0.0729483887553215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29923492670059204, "epoch": 12.55, "learning_rate": 9.060738885410145e-06, "loss": 0.371, "step": 14851, "task_loss": 0.30405858159065247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4804536700248718, "epoch": 12.55, "learning_rate": 9.057608015028179e-06, "loss": 0.4415, "step": 14852, "task_loss": 0.6187456846237183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39918583631515503, "epoch": 12.56, "learning_rate": 9.054477144646212e-06, "loss": 0.4267, "step": 14853, "task_loss": 0.8656633496284485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30365490913391113, "epoch": 12.56, "learning_rate": 9.051346274264245e-06, "loss": 0.3913, "step": 14854, "task_loss": 0.3493107259273529 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2962064743041992, "epoch": 12.56, "learning_rate": 9.048215403882279e-06, "loss": 0.4499, "step": 14855, "task_loss": 0.6856915950775146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5198525190353394, "epoch": 12.56, "learning_rate": 9.045084533500312e-06, "loss": 0.4222, "step": 14856, "task_loss": 0.3112236261367798 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5841454267501831, "epoch": 12.56, "learning_rate": 9.041953663118348e-06, "loss": 0.5336, "step": 14857, "task_loss": 0.4352759122848511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36045849323272705, "epoch": 12.56, "learning_rate": 9.038822792736381e-06, "loss": 0.4059, "step": 14858, "task_loss": 0.7454867362976074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30854836106300354, "epoch": 12.56, "learning_rate": 9.035691922354415e-06, "loss": 0.3545, "step": 14859, "task_loss": 0.5559686422348022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4939218759536743, "epoch": 12.56, "learning_rate": 9.03256105197245e-06, "loss": 0.507, "step": 14860, "task_loss": 0.27223098278045654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42527633905410767, "epoch": 12.56, "learning_rate": 9.029430181590483e-06, "loss": 0.3733, "step": 14861, "task_loss": 0.23771452903747559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7278570532798767, "epoch": 12.56, "learning_rate": 9.026299311208517e-06, "loss": 0.5247, "step": 14862, "task_loss": 1.6112620830535889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6096972823143005, "epoch": 12.56, "learning_rate": 9.02316844082655e-06, "loss": 0.5183, "step": 14863, "task_loss": 0.8716512322425842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44267725944519043, "epoch": 12.56, "learning_rate": 9.020037570444584e-06, "loss": 0.5377, "step": 14864, "task_loss": 1.2528915405273438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5987818241119385, "epoch": 12.57, "learning_rate": 9.016906700062617e-06, "loss": 0.4573, "step": 14865, "task_loss": 1.374953269958496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21130748093128204, "epoch": 12.57, "learning_rate": 9.01377582968065e-06, "loss": 0.4418, "step": 14866, "task_loss": 0.8155895471572876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3173671364784241, "epoch": 12.57, "learning_rate": 9.010644959298686e-06, "loss": 0.4484, "step": 14867, "task_loss": 1.8693255186080933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3460475206375122, "epoch": 12.57, "learning_rate": 9.00751408891672e-06, "loss": 0.4166, "step": 14868, "task_loss": 0.7854889631271362 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47989988327026367, "epoch": 12.57, "learning_rate": 9.004383218534754e-06, "loss": 0.6576, "step": 14869, "task_loss": 0.681489109992981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20407040417194366, "epoch": 12.57, "learning_rate": 9.001252348152788e-06, "loss": 0.6076, "step": 14870, "task_loss": 0.5795115828514099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7606086730957031, "epoch": 12.57, "learning_rate": 8.998121477770821e-06, "loss": 0.5839, "step": 14871, "task_loss": 1.3302158117294312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46591681241989136, "epoch": 12.57, "learning_rate": 8.994990607388855e-06, "loss": 0.4886, "step": 14872, "task_loss": 0.9293614625930786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3414246439933777, "epoch": 12.57, "learning_rate": 8.991859737006888e-06, "loss": 0.3835, "step": 14873, "task_loss": 0.32577601075172424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32009774446487427, "epoch": 12.57, "learning_rate": 8.988728866624922e-06, "loss": 0.3894, "step": 14874, "task_loss": 0.29216548800468445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3869706988334656, "epoch": 12.57, "learning_rate": 8.985597996242955e-06, "loss": 0.3728, "step": 14875, "task_loss": 0.24032433331012726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3061337471008301, "epoch": 12.57, "learning_rate": 8.98246712586099e-06, "loss": 0.4532, "step": 14876, "task_loss": 0.7054287791252136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5846126079559326, "epoch": 12.58, "learning_rate": 8.979336255479024e-06, "loss": 0.4852, "step": 14877, "task_loss": 0.7050690650939941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6806417107582092, "epoch": 12.58, "learning_rate": 8.976205385097057e-06, "loss": 0.4388, "step": 14878, "task_loss": 0.4574021100997925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4445473551750183, "epoch": 12.58, "learning_rate": 8.97307451471509e-06, "loss": 0.4679, "step": 14879, "task_loss": 0.831595778465271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5354905128479004, "epoch": 12.58, "learning_rate": 8.969943644333124e-06, "loss": 0.3658, "step": 14880, "task_loss": 0.5835623145103455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4151933491230011, "epoch": 12.58, "learning_rate": 8.96681277395116e-06, "loss": 0.4481, "step": 14881, "task_loss": 0.8246848583221436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39621567726135254, "epoch": 12.58, "learning_rate": 8.963681903569193e-06, "loss": 0.5239, "step": 14882, "task_loss": 1.6599743366241455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5255345702171326, "epoch": 12.58, "learning_rate": 8.960551033187226e-06, "loss": 0.5195, "step": 14883, "task_loss": 0.6217733025550842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5051659345626831, "epoch": 12.58, "learning_rate": 8.95742016280526e-06, "loss": 0.4788, "step": 14884, "task_loss": 0.790012001991272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3009679615497589, "epoch": 12.58, "learning_rate": 8.954289292423295e-06, "loss": 0.3915, "step": 14885, "task_loss": 0.7301338911056519 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3211176097393036, "epoch": 12.58, "learning_rate": 8.951158422041329e-06, "loss": 0.3789, "step": 14886, "task_loss": 0.7887259125709534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3420470356941223, "epoch": 12.58, "learning_rate": 8.948027551659362e-06, "loss": 0.6004, "step": 14887, "task_loss": 0.8467986583709717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32181307673454285, "epoch": 12.58, "learning_rate": 8.944896681277395e-06, "loss": 0.4471, "step": 14888, "task_loss": 0.6851255893707275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6755098700523376, "epoch": 12.59, "learning_rate": 8.941765810895429e-06, "loss": 0.4483, "step": 14889, "task_loss": 0.7558509707450867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4806479811668396, "epoch": 12.59, "learning_rate": 8.938634940513462e-06, "loss": 0.4369, "step": 14890, "task_loss": 1.129317045211792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3778035640716553, "epoch": 12.59, "learning_rate": 8.935504070131496e-06, "loss": 0.3909, "step": 14891, "task_loss": 0.34499871730804443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47518977522850037, "epoch": 12.59, "learning_rate": 8.932373199749531e-06, "loss": 0.3996, "step": 14892, "task_loss": 0.3795377016067505 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4834267497062683, "epoch": 12.59, "learning_rate": 8.929242329367565e-06, "loss": 0.3894, "step": 14893, "task_loss": 0.5793136954307556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7098901271820068, "epoch": 12.59, "learning_rate": 8.9261114589856e-06, "loss": 0.4549, "step": 14894, "task_loss": 0.7222098112106323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5735839009284973, "epoch": 12.59, "learning_rate": 8.922980588603633e-06, "loss": 0.5961, "step": 14895, "task_loss": 0.46923232078552246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26213258504867554, "epoch": 12.59, "learning_rate": 8.919849718221667e-06, "loss": 0.3998, "step": 14896, "task_loss": 0.3808481991291046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3118307590484619, "epoch": 12.59, "learning_rate": 8.9167188478397e-06, "loss": 0.4835, "step": 14897, "task_loss": 0.45929980278015137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37938612699508667, "epoch": 12.59, "learning_rate": 8.913587977457734e-06, "loss": 0.4167, "step": 14898, "task_loss": 1.044651985168457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45125889778137207, "epoch": 12.59, "learning_rate": 8.910457107075767e-06, "loss": 0.4982, "step": 14899, "task_loss": 0.6281577944755554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45382773876190186, "epoch": 12.59, "learning_rate": 8.9073262366938e-06, "loss": 0.5269, "step": 14900, "task_loss": 0.6415404677391052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48325225710868835, "epoch": 12.6, "learning_rate": 8.904195366311836e-06, "loss": 0.4764, "step": 14901, "task_loss": 0.7526668310165405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24653780460357666, "epoch": 12.6, "learning_rate": 8.90106449592987e-06, "loss": 0.3515, "step": 14902, "task_loss": 0.15054550766944885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31439048051834106, "epoch": 12.6, "learning_rate": 8.897933625547903e-06, "loss": 0.467, "step": 14903, "task_loss": 0.23628124594688416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5740969181060791, "epoch": 12.6, "learning_rate": 8.894802755165936e-06, "loss": 0.4354, "step": 14904, "task_loss": 0.19161462783813477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34406524896621704, "epoch": 12.6, "learning_rate": 8.89167188478397e-06, "loss": 0.4207, "step": 14905, "task_loss": 0.820719838142395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7161105275154114, "epoch": 12.6, "learning_rate": 8.888541014402005e-06, "loss": 0.481, "step": 14906, "task_loss": 1.4422625303268433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3610774576663971, "epoch": 12.6, "learning_rate": 8.885410144020038e-06, "loss": 0.4467, "step": 14907, "task_loss": 0.16686297953128815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39500850439071655, "epoch": 12.6, "learning_rate": 8.882279273638072e-06, "loss": 0.4579, "step": 14908, "task_loss": 0.7884155511856079 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3458581566810608, "epoch": 12.6, "learning_rate": 8.879148403256105e-06, "loss": 0.5436, "step": 14909, "task_loss": 1.1006944179534912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31750744581222534, "epoch": 12.6, "learning_rate": 8.87601753287414e-06, "loss": 0.3039, "step": 14910, "task_loss": 0.4141862094402313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9385615587234497, "epoch": 12.6, "learning_rate": 8.872886662492174e-06, "loss": 0.5488, "step": 14911, "task_loss": 0.6118607521057129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6013718247413635, "epoch": 12.6, "learning_rate": 8.869755792110207e-06, "loss": 0.4875, "step": 14912, "task_loss": 0.6711826324462891 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6636090278625488, "epoch": 12.61, "learning_rate": 8.86662492172824e-06, "loss": 0.5151, "step": 14913, "task_loss": 0.47691214084625244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4290599822998047, "epoch": 12.61, "learning_rate": 8.863494051346274e-06, "loss": 0.52, "step": 14914, "task_loss": 0.24055971205234528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42754459381103516, "epoch": 12.61, "learning_rate": 8.860363180964308e-06, "loss": 0.4045, "step": 14915, "task_loss": 0.7849160432815552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7903364896774292, "epoch": 12.61, "learning_rate": 8.857232310582341e-06, "loss": 0.5599, "step": 14916, "task_loss": 1.6914043426513672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5573076009750366, "epoch": 12.61, "learning_rate": 8.854101440200375e-06, "loss": 0.4696, "step": 14917, "task_loss": 0.5687537789344788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29863184690475464, "epoch": 12.61, "learning_rate": 8.85097056981841e-06, "loss": 0.4541, "step": 14918, "task_loss": 0.6171025633811951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34824633598327637, "epoch": 12.61, "learning_rate": 8.847839699436445e-06, "loss": 0.3202, "step": 14919, "task_loss": 0.6095053553581238 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5643831491470337, "epoch": 12.61, "learning_rate": 8.844708829054479e-06, "loss": 0.5015, "step": 14920, "task_loss": 0.6175873279571533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3126521110534668, "epoch": 12.61, "learning_rate": 8.841577958672512e-06, "loss": 0.4436, "step": 14921, "task_loss": 0.10836179554462433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9112505912780762, "epoch": 12.61, "learning_rate": 8.838447088290546e-06, "loss": 0.5502, "step": 14922, "task_loss": 0.9713506102561951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3166190981864929, "epoch": 12.61, "learning_rate": 8.835316217908579e-06, "loss": 0.4, "step": 14923, "task_loss": 1.2923506498336792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49920135736465454, "epoch": 12.61, "learning_rate": 8.832185347526612e-06, "loss": 0.6045, "step": 14924, "task_loss": 1.0267150402069092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4112991690635681, "epoch": 12.62, "learning_rate": 8.829054477144646e-06, "loss": 0.5515, "step": 14925, "task_loss": 0.8441474437713623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4441414475440979, "epoch": 12.62, "learning_rate": 8.82592360676268e-06, "loss": 0.4459, "step": 14926, "task_loss": 0.6768850684165955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27627891302108765, "epoch": 12.62, "learning_rate": 8.822792736380715e-06, "loss": 0.4571, "step": 14927, "task_loss": 0.462390273809433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3476416766643524, "epoch": 12.62, "learning_rate": 8.819661865998748e-06, "loss": 0.4509, "step": 14928, "task_loss": 0.5561189651489258 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6250041127204895, "epoch": 12.62, "learning_rate": 8.816530995616782e-06, "loss": 0.6055, "step": 14929, "task_loss": 1.1451953649520874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3513980805873871, "epoch": 12.62, "learning_rate": 8.813400125234815e-06, "loss": 0.4151, "step": 14930, "task_loss": 0.6729755997657776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3823819160461426, "epoch": 12.62, "learning_rate": 8.81026925485285e-06, "loss": 0.4276, "step": 14931, "task_loss": 0.15493148565292358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5778601169586182, "epoch": 12.62, "learning_rate": 8.807138384470884e-06, "loss": 0.6624, "step": 14932, "task_loss": 0.6659271717071533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41028738021850586, "epoch": 12.62, "learning_rate": 8.804007514088917e-06, "loss": 0.536, "step": 14933, "task_loss": 1.0005998611450195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3477420508861542, "epoch": 12.62, "learning_rate": 8.80087664370695e-06, "loss": 0.4557, "step": 14934, "task_loss": 0.6053062677383423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28235718607902527, "epoch": 12.62, "learning_rate": 8.797745773324984e-06, "loss": 0.3911, "step": 14935, "task_loss": 0.41435712575912476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4828604459762573, "epoch": 12.63, "learning_rate": 8.79461490294302e-06, "loss": 0.4839, "step": 14936, "task_loss": 0.4614100754261017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7266366481781006, "epoch": 12.63, "learning_rate": 8.791484032561053e-06, "loss": 0.495, "step": 14937, "task_loss": 0.8454301357269287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24613283574581146, "epoch": 12.63, "learning_rate": 8.788353162179086e-06, "loss": 0.3982, "step": 14938, "task_loss": 0.40139010548591614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3832569122314453, "epoch": 12.63, "learning_rate": 8.78522229179712e-06, "loss": 0.451, "step": 14939, "task_loss": 0.7386264801025391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4367101788520813, "epoch": 12.63, "learning_rate": 8.782091421415153e-06, "loss": 0.6277, "step": 14940, "task_loss": 0.7650018930435181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3139549791812897, "epoch": 12.63, "learning_rate": 8.778960551033187e-06, "loss": 0.4034, "step": 14941, "task_loss": 0.3505925238132477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3821285367012024, "epoch": 12.63, "learning_rate": 8.77582968065122e-06, "loss": 0.5148, "step": 14942, "task_loss": 0.6908835768699646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2906680703163147, "epoch": 12.63, "learning_rate": 8.772698810269255e-06, "loss": 0.381, "step": 14943, "task_loss": 0.22858025133609772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.263770192861557, "epoch": 12.63, "learning_rate": 8.769567939887289e-06, "loss": 0.5299, "step": 14944, "task_loss": 0.25909313559532166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6255160570144653, "epoch": 12.63, "learning_rate": 8.766437069505324e-06, "loss": 0.5225, "step": 14945, "task_loss": 0.7047940492630005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39091822504997253, "epoch": 12.63, "learning_rate": 8.763306199123357e-06, "loss": 0.406, "step": 14946, "task_loss": 0.5998892188072205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5574036836624146, "epoch": 12.63, "learning_rate": 8.760175328741391e-06, "loss": 0.4638, "step": 14947, "task_loss": 0.6567528247833252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4889408349990845, "epoch": 12.64, "learning_rate": 8.757044458359424e-06, "loss": 0.3718, "step": 14948, "task_loss": 1.0475822687149048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42036545276641846, "epoch": 12.64, "learning_rate": 8.753913587977458e-06, "loss": 0.5009, "step": 14949, "task_loss": 0.6813563704490662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5711225867271423, "epoch": 12.64, "learning_rate": 8.750782717595491e-06, "loss": 0.5603, "step": 14950, "task_loss": 0.5468470454216003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42140138149261475, "epoch": 12.64, "learning_rate": 8.747651847213525e-06, "loss": 0.4752, "step": 14951, "task_loss": 0.8426974415779114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47556638717651367, "epoch": 12.64, "learning_rate": 8.74452097683156e-06, "loss": 0.5032, "step": 14952, "task_loss": 0.4180910289287567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5288169980049133, "epoch": 12.64, "learning_rate": 8.741390106449593e-06, "loss": 0.7492, "step": 14953, "task_loss": 0.6630289554595947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3038691580295563, "epoch": 12.64, "learning_rate": 8.738259236067627e-06, "loss": 0.5786, "step": 14954, "task_loss": 0.8012465238571167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27326661348342896, "epoch": 12.64, "learning_rate": 8.735128365685662e-06, "loss": 0.5251, "step": 14955, "task_loss": 0.6093032956123352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32037919759750366, "epoch": 12.64, "learning_rate": 8.731997495303696e-06, "loss": 0.3509, "step": 14956, "task_loss": 0.8187767863273621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5741370320320129, "epoch": 12.64, "learning_rate": 8.728866624921729e-06, "loss": 0.4343, "step": 14957, "task_loss": 0.9801803827285767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4221974313259125, "epoch": 12.64, "learning_rate": 8.725735754539762e-06, "loss": 0.5776, "step": 14958, "task_loss": 0.7329771518707275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37081044912338257, "epoch": 12.64, "learning_rate": 8.722604884157796e-06, "loss": 0.3724, "step": 14959, "task_loss": 0.4264255464076996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22438254952430725, "epoch": 12.65, "learning_rate": 8.71947401377583e-06, "loss": 0.4334, "step": 14960, "task_loss": 0.4466226398944855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37071657180786133, "epoch": 12.65, "learning_rate": 8.716343143393865e-06, "loss": 0.4792, "step": 14961, "task_loss": 0.41348642110824585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29575419425964355, "epoch": 12.65, "learning_rate": 8.713212273011898e-06, "loss": 0.3174, "step": 14962, "task_loss": 0.3414704501628876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45498210191726685, "epoch": 12.65, "learning_rate": 8.710081402629932e-06, "loss": 0.5149, "step": 14963, "task_loss": 0.5863996148109436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5624514818191528, "epoch": 12.65, "learning_rate": 8.706950532247965e-06, "loss": 0.5069, "step": 14964, "task_loss": 0.9843158721923828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5029947757720947, "epoch": 12.65, "learning_rate": 8.703819661865998e-06, "loss": 0.3941, "step": 14965, "task_loss": 0.8024125099182129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5880112648010254, "epoch": 12.65, "learning_rate": 8.700688791484032e-06, "loss": 0.3724, "step": 14966, "task_loss": 0.8361860513687134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.679474949836731, "epoch": 12.65, "learning_rate": 8.697557921102067e-06, "loss": 0.4487, "step": 14967, "task_loss": 0.971507728099823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3482791483402252, "epoch": 12.65, "learning_rate": 8.6944270507201e-06, "loss": 0.452, "step": 14968, "task_loss": 0.7911660075187683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4422319531440735, "epoch": 12.65, "learning_rate": 8.691296180338134e-06, "loss": 0.4749, "step": 14969, "task_loss": 0.5685135126113892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36695343255996704, "epoch": 12.65, "learning_rate": 8.68816530995617e-06, "loss": 0.3949, "step": 14970, "task_loss": 1.1965309381484985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4971216320991516, "epoch": 12.65, "learning_rate": 8.685034439574203e-06, "loss": 0.4771, "step": 14971, "task_loss": 0.9376870393753052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36114317178726196, "epoch": 12.66, "learning_rate": 8.681903569192236e-06, "loss": 0.4439, "step": 14972, "task_loss": 1.0273821353912354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6220627427101135, "epoch": 12.66, "learning_rate": 8.67877269881027e-06, "loss": 0.3869, "step": 14973, "task_loss": 0.6485153436660767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21099475026130676, "epoch": 12.66, "learning_rate": 8.675641828428303e-06, "loss": 0.3933, "step": 14974, "task_loss": 0.5774315595626831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41004353761672974, "epoch": 12.66, "learning_rate": 8.672510958046337e-06, "loss": 0.3929, "step": 14975, "task_loss": 0.6806992292404175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.326510488986969, "epoch": 12.66, "learning_rate": 8.66938008766437e-06, "loss": 0.51, "step": 14976, "task_loss": 0.27831995487213135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23943543434143066, "epoch": 12.66, "learning_rate": 8.666249217282405e-06, "loss": 0.4407, "step": 14977, "task_loss": 0.9112436175346375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.807584285736084, "epoch": 12.66, "learning_rate": 8.663118346900439e-06, "loss": 0.5566, "step": 14978, "task_loss": 1.2227028608322144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4104318916797638, "epoch": 12.66, "learning_rate": 8.659987476518472e-06, "loss": 0.4564, "step": 14979, "task_loss": 0.4504373371601105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4987604022026062, "epoch": 12.66, "learning_rate": 8.656856606136507e-06, "loss": 0.4816, "step": 14980, "task_loss": 0.30625608563423157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38612037897109985, "epoch": 12.66, "learning_rate": 8.653725735754541e-06, "loss": 0.3954, "step": 14981, "task_loss": 0.13562066853046417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6069244146347046, "epoch": 12.66, "learning_rate": 8.650594865372574e-06, "loss": 0.4238, "step": 14982, "task_loss": 0.6440811157226562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.498496413230896, "epoch": 12.66, "learning_rate": 8.647463994990608e-06, "loss": 0.4746, "step": 14983, "task_loss": 0.3945971429347992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.494968444108963, "epoch": 12.67, "learning_rate": 8.644333124608641e-06, "loss": 0.4522, "step": 14984, "task_loss": 0.5184865593910217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5119041204452515, "epoch": 12.67, "learning_rate": 8.641202254226675e-06, "loss": 0.5074, "step": 14985, "task_loss": 0.9071097373962402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43620777130126953, "epoch": 12.67, "learning_rate": 8.63807138384471e-06, "loss": 0.4099, "step": 14986, "task_loss": 0.340157151222229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29797831177711487, "epoch": 12.67, "learning_rate": 8.634940513462743e-06, "loss": 0.3369, "step": 14987, "task_loss": 0.6407220959663391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25314584374427795, "epoch": 12.67, "learning_rate": 8.631809643080777e-06, "loss": 0.3413, "step": 14988, "task_loss": 0.43689221143722534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3709731698036194, "epoch": 12.67, "learning_rate": 8.62867877269881e-06, "loss": 0.4353, "step": 14989, "task_loss": 0.841255784034729 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4508817195892334, "epoch": 12.67, "learning_rate": 8.625547902316844e-06, "loss": 0.4275, "step": 14990, "task_loss": 0.19127006828784943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39889663457870483, "epoch": 12.67, "learning_rate": 8.622417031934877e-06, "loss": 0.2793, "step": 14991, "task_loss": 0.2130555957555771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3514246344566345, "epoch": 12.67, "learning_rate": 8.619286161552912e-06, "loss": 0.5293, "step": 14992, "task_loss": 0.31811821460723877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42009270191192627, "epoch": 12.67, "learning_rate": 8.616155291170946e-06, "loss": 0.3934, "step": 14993, "task_loss": 0.13368967175483704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44955652952194214, "epoch": 12.67, "learning_rate": 8.61302442078898e-06, "loss": 0.3961, "step": 14994, "task_loss": 0.09357498586177826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5810348391532898, "epoch": 12.67, "learning_rate": 8.609893550407015e-06, "loss": 0.4516, "step": 14995, "task_loss": 0.28797656297683716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2958005666732788, "epoch": 12.68, "learning_rate": 8.606762680025048e-06, "loss": 0.3648, "step": 14996, "task_loss": 1.1024057865142822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21284520626068115, "epoch": 12.68, "learning_rate": 8.603631809643082e-06, "loss": 0.4757, "step": 14997, "task_loss": 0.6410262584686279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.588051438331604, "epoch": 12.68, "learning_rate": 8.600500939261115e-06, "loss": 0.5022, "step": 14998, "task_loss": 1.3106516599655151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5341254472732544, "epoch": 12.68, "learning_rate": 8.597370068879149e-06, "loss": 0.5114, "step": 14999, "task_loss": 0.2657007575035095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4458681344985962, "epoch": 12.68, "learning_rate": 8.594239198497182e-06, "loss": 0.4763, "step": 15000, "task_loss": 0.7662859559059143 }, { "epoch": 12.68, "eval_accuracy": 0.9108514851485149, "eval_loss": 0.31584760546684265, "eval_runtime": 207.0915, "eval_samples_per_second": 121.927, "eval_steps_per_second": 0.956, "step": 15000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2992781400680542, "epoch": 12.68, "learning_rate": 8.591108328115215e-06, "loss": 0.4955, "step": 15001, "task_loss": 0.6928538084030151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38270285725593567, "epoch": 12.68, "learning_rate": 8.587977457733249e-06, "loss": 0.3923, "step": 15002, "task_loss": 0.6056111454963684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3749743700027466, "epoch": 12.68, "learning_rate": 8.584846587351284e-06, "loss": 0.3126, "step": 15003, "task_loss": 1.0059964656829834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4838295578956604, "epoch": 12.68, "learning_rate": 8.58171571696932e-06, "loss": 0.4377, "step": 15004, "task_loss": 0.6992678046226501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4968317747116089, "epoch": 12.68, "learning_rate": 8.578584846587353e-06, "loss": 0.5219, "step": 15005, "task_loss": 0.588087797164917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4912208616733551, "epoch": 12.68, "learning_rate": 8.575453976205386e-06, "loss": 0.4412, "step": 15006, "task_loss": 0.6697782874107361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5966281890869141, "epoch": 12.69, "learning_rate": 8.57232310582342e-06, "loss": 0.4837, "step": 15007, "task_loss": 0.594540536403656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3577840030193329, "epoch": 12.69, "learning_rate": 8.569192235441453e-06, "loss": 0.4615, "step": 15008, "task_loss": 1.0611470937728882 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6294047236442566, "epoch": 12.69, "learning_rate": 8.566061365059487e-06, "loss": 0.4126, "step": 15009, "task_loss": 1.1358230113983154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5233056545257568, "epoch": 12.69, "learning_rate": 8.56293049467752e-06, "loss": 0.3816, "step": 15010, "task_loss": 1.2729169130325317 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.304970383644104, "epoch": 12.69, "learning_rate": 8.559799624295554e-06, "loss": 0.3933, "step": 15011, "task_loss": 0.3947198987007141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3857519328594208, "epoch": 12.69, "learning_rate": 8.556668753913589e-06, "loss": 0.4002, "step": 15012, "task_loss": 0.7845688462257385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4348602294921875, "epoch": 12.69, "learning_rate": 8.553537883531622e-06, "loss": 0.4338, "step": 15013, "task_loss": 0.8083392381668091 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4202806353569031, "epoch": 12.69, "learning_rate": 8.550407013149656e-06, "loss": 0.3718, "step": 15014, "task_loss": 0.11979993432760239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5966145396232605, "epoch": 12.69, "learning_rate": 8.54727614276769e-06, "loss": 0.4411, "step": 15015, "task_loss": 0.49939852952957153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41635221242904663, "epoch": 12.69, "learning_rate": 8.544145272385724e-06, "loss": 0.4113, "step": 15016, "task_loss": 1.2755002975463867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3873010575771332, "epoch": 12.69, "learning_rate": 8.541014402003758e-06, "loss": 0.6276, "step": 15017, "task_loss": 0.33996373414993286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28069382905960083, "epoch": 12.69, "learning_rate": 8.537883531621791e-06, "loss": 0.4465, "step": 15018, "task_loss": 0.3020857870578766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16080352663993835, "epoch": 12.7, "learning_rate": 8.534752661239825e-06, "loss": 0.3865, "step": 15019, "task_loss": 0.046947963535785675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.14076992869377136, "epoch": 12.7, "learning_rate": 8.531621790857858e-06, "loss": 0.3075, "step": 15020, "task_loss": 0.012121761217713356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30893757939338684, "epoch": 12.7, "learning_rate": 8.528490920475893e-06, "loss": 0.3619, "step": 15021, "task_loss": 0.5024336576461792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28613483905792236, "epoch": 12.7, "learning_rate": 8.525360050093927e-06, "loss": 0.4245, "step": 15022, "task_loss": 0.1648285686969757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.590583860874176, "epoch": 12.7, "learning_rate": 8.52222917971196e-06, "loss": 0.5147, "step": 15023, "task_loss": 0.5786076188087463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5034299492835999, "epoch": 12.7, "learning_rate": 8.519098309329994e-06, "loss": 0.4489, "step": 15024, "task_loss": 0.6493264436721802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45695415139198303, "epoch": 12.7, "learning_rate": 8.515967438948027e-06, "loss": 0.5493, "step": 15025, "task_loss": 1.2891731262207031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42161285877227783, "epoch": 12.7, "learning_rate": 8.51283656856606e-06, "loss": 0.4461, "step": 15026, "task_loss": 0.8062981367111206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46077123284339905, "epoch": 12.7, "learning_rate": 8.509705698184094e-06, "loss": 0.3794, "step": 15027, "task_loss": 0.13389770686626434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6203373670578003, "epoch": 12.7, "learning_rate": 8.50657482780213e-06, "loss": 0.6002, "step": 15028, "task_loss": 0.47173383831977844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40667325258255005, "epoch": 12.7, "learning_rate": 8.503443957420165e-06, "loss": 0.4816, "step": 15029, "task_loss": 1.5712803602218628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5335782766342163, "epoch": 12.7, "learning_rate": 8.500313087038198e-06, "loss": 0.5265, "step": 15030, "task_loss": 0.48202869296073914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5498442649841309, "epoch": 12.71, "learning_rate": 8.497182216656232e-06, "loss": 0.4976, "step": 15031, "task_loss": 0.6282045245170593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2927628755569458, "epoch": 12.71, "learning_rate": 8.494051346274265e-06, "loss": 0.5468, "step": 15032, "task_loss": 0.47194430232048035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8482906818389893, "epoch": 12.71, "learning_rate": 8.490920475892299e-06, "loss": 0.5457, "step": 15033, "task_loss": 1.2491860389709473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4372267723083496, "epoch": 12.71, "learning_rate": 8.487789605510332e-06, "loss": 0.4818, "step": 15034, "task_loss": 0.33062660694122314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6873219013214111, "epoch": 12.71, "learning_rate": 8.484658735128365e-06, "loss": 0.477, "step": 15035, "task_loss": 1.012682318687439 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7454384565353394, "epoch": 12.71, "learning_rate": 8.481527864746399e-06, "loss": 0.5039, "step": 15036, "task_loss": 0.609855592250824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16595464944839478, "epoch": 12.71, "learning_rate": 8.478396994364434e-06, "loss": 0.3763, "step": 15037, "task_loss": 0.04033783823251724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4268585741519928, "epoch": 12.71, "learning_rate": 8.475266123982468e-06, "loss": 0.4309, "step": 15038, "task_loss": 0.5433082580566406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5056393146514893, "epoch": 12.71, "learning_rate": 8.472135253600501e-06, "loss": 0.4719, "step": 15039, "task_loss": 0.8161755204200745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.734204888343811, "epoch": 12.71, "learning_rate": 8.469004383218535e-06, "loss": 0.5386, "step": 15040, "task_loss": 1.6317133903503418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3977241516113281, "epoch": 12.71, "learning_rate": 8.46587351283657e-06, "loss": 0.4351, "step": 15041, "task_loss": 0.9748689532279968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3906921148300171, "epoch": 12.71, "learning_rate": 8.462742642454603e-06, "loss": 0.4876, "step": 15042, "task_loss": 0.39076122641563416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.641484260559082, "epoch": 12.72, "learning_rate": 8.459611772072637e-06, "loss": 0.5019, "step": 15043, "task_loss": 1.1150962114334106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7718719840049744, "epoch": 12.72, "learning_rate": 8.45648090169067e-06, "loss": 0.5943, "step": 15044, "task_loss": 1.0566012859344482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4358503222465515, "epoch": 12.72, "learning_rate": 8.453350031308704e-06, "loss": 0.4276, "step": 15045, "task_loss": 0.4601661264896393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26912760734558105, "epoch": 12.72, "learning_rate": 8.450219160926739e-06, "loss": 0.4132, "step": 15046, "task_loss": 1.4157288074493408 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4863024055957794, "epoch": 12.72, "learning_rate": 8.447088290544772e-06, "loss": 0.4274, "step": 15047, "task_loss": 0.49509820342063904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1290842890739441, "epoch": 12.72, "learning_rate": 8.443957420162806e-06, "loss": 0.3337, "step": 15048, "task_loss": 0.06690125912427902 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7423638105392456, "epoch": 12.72, "learning_rate": 8.44082654978084e-06, "loss": 0.4884, "step": 15049, "task_loss": 1.613747000694275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5061672925949097, "epoch": 12.72, "learning_rate": 8.437695679398873e-06, "loss": 0.5902, "step": 15050, "task_loss": 0.6571119427680969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4697355031967163, "epoch": 12.72, "learning_rate": 8.434564809016906e-06, "loss": 0.594, "step": 15051, "task_loss": 0.3488698899745941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.390642911195755, "epoch": 12.72, "learning_rate": 8.43143393863494e-06, "loss": 0.453, "step": 15052, "task_loss": 0.3068805932998657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5271430611610413, "epoch": 12.72, "learning_rate": 8.428303068252975e-06, "loss": 0.3938, "step": 15053, "task_loss": 2.2373833656311035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3082345724105835, "epoch": 12.72, "learning_rate": 8.425172197871008e-06, "loss": 0.4888, "step": 15054, "task_loss": 0.5996771454811096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5031503438949585, "epoch": 12.73, "learning_rate": 8.422041327489043e-06, "loss": 0.4548, "step": 15055, "task_loss": 0.955868124961853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5646097660064697, "epoch": 12.73, "learning_rate": 8.418910457107077e-06, "loss": 0.4809, "step": 15056, "task_loss": 0.447252482175827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5505818128585815, "epoch": 12.73, "learning_rate": 8.41577958672511e-06, "loss": 0.5313, "step": 15057, "task_loss": 1.0125007629394531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3821646571159363, "epoch": 12.73, "learning_rate": 8.412648716343144e-06, "loss": 0.4221, "step": 15058, "task_loss": 0.8555232286453247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2605704367160797, "epoch": 12.73, "learning_rate": 8.409517845961177e-06, "loss": 0.386, "step": 15059, "task_loss": 0.14093275368213654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5707604289054871, "epoch": 12.73, "learning_rate": 8.40638697557921e-06, "loss": 0.4147, "step": 15060, "task_loss": 0.16486231982707977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30758675932884216, "epoch": 12.73, "learning_rate": 8.403256105197244e-06, "loss": 0.4744, "step": 15061, "task_loss": 1.019897222518921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4703580141067505, "epoch": 12.73, "learning_rate": 8.40012523481528e-06, "loss": 0.5051, "step": 15062, "task_loss": 0.8240983486175537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2263239473104477, "epoch": 12.73, "learning_rate": 8.396994364433313e-06, "loss": 0.3454, "step": 15063, "task_loss": 0.5318767428398132 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2875639796257019, "epoch": 12.73, "learning_rate": 8.393863494051346e-06, "loss": 0.4644, "step": 15064, "task_loss": 0.5164564847946167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6147679090499878, "epoch": 12.73, "learning_rate": 8.390732623669382e-06, "loss": 0.4802, "step": 15065, "task_loss": 0.7188316583633423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4352780282497406, "epoch": 12.73, "learning_rate": 8.387601753287415e-06, "loss": 0.4443, "step": 15066, "task_loss": 0.9864170551300049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3694538474082947, "epoch": 12.74, "learning_rate": 8.384470882905449e-06, "loss": 0.4853, "step": 15067, "task_loss": 0.4424612820148468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3271103501319885, "epoch": 12.74, "learning_rate": 8.381340012523482e-06, "loss": 0.4768, "step": 15068, "task_loss": 0.29143452644348145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44124555587768555, "epoch": 12.74, "learning_rate": 8.378209142141516e-06, "loss": 0.577, "step": 15069, "task_loss": 0.8069020509719849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4634045958518982, "epoch": 12.74, "learning_rate": 8.375078271759549e-06, "loss": 0.4479, "step": 15070, "task_loss": 0.4777064919471741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6982874870300293, "epoch": 12.74, "learning_rate": 8.371947401377584e-06, "loss": 0.5415, "step": 15071, "task_loss": 0.7495650053024292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24343471229076385, "epoch": 12.74, "learning_rate": 8.368816530995618e-06, "loss": 0.4408, "step": 15072, "task_loss": 0.023372666910290718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47557100653648376, "epoch": 12.74, "learning_rate": 8.365685660613651e-06, "loss": 0.5045, "step": 15073, "task_loss": 1.2324203252792358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49406564235687256, "epoch": 12.74, "learning_rate": 8.362554790231685e-06, "loss": 0.4944, "step": 15074, "task_loss": 0.5806005001068115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39305564761161804, "epoch": 12.74, "learning_rate": 8.359423919849718e-06, "loss": 0.5138, "step": 15075, "task_loss": 0.4907389283180237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43902501463890076, "epoch": 12.74, "learning_rate": 8.356293049467752e-06, "loss": 0.4051, "step": 15076, "task_loss": 0.18382695317268372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3826756477355957, "epoch": 12.74, "learning_rate": 8.353162179085785e-06, "loss": 0.5751, "step": 15077, "task_loss": 0.2872302234172821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5986155271530151, "epoch": 12.75, "learning_rate": 8.35003130870382e-06, "loss": 0.4269, "step": 15078, "task_loss": 1.0268625020980835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7232465744018555, "epoch": 12.75, "learning_rate": 8.346900438321854e-06, "loss": 0.484, "step": 15079, "task_loss": 0.24175921082496643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4814614951610565, "epoch": 12.75, "learning_rate": 8.343769567939889e-06, "loss": 0.4896, "step": 15080, "task_loss": 0.8093700408935547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1947760283946991, "epoch": 12.75, "learning_rate": 8.340638697557922e-06, "loss": 0.3286, "step": 15081, "task_loss": 0.2038325071334839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3054892122745514, "epoch": 12.75, "learning_rate": 8.337507827175956e-06, "loss": 0.4254, "step": 15082, "task_loss": 1.113093614578247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7153237462043762, "epoch": 12.75, "learning_rate": 8.33437695679399e-06, "loss": 0.5236, "step": 15083, "task_loss": 1.1103063821792603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23369050025939941, "epoch": 12.75, "learning_rate": 8.331246086412023e-06, "loss": 0.4021, "step": 15084, "task_loss": 0.8393123745918274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46605896949768066, "epoch": 12.75, "learning_rate": 8.328115216030056e-06, "loss": 0.5551, "step": 15085, "task_loss": 0.7559438943862915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5458028316497803, "epoch": 12.75, "learning_rate": 8.32498434564809e-06, "loss": 0.5161, "step": 15086, "task_loss": 1.122238278388977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47603312134742737, "epoch": 12.75, "learning_rate": 8.321853475266123e-06, "loss": 0.4368, "step": 15087, "task_loss": 0.22567547857761383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38248762488365173, "epoch": 12.75, "learning_rate": 8.318722604884158e-06, "loss": 0.4277, "step": 15088, "task_loss": 0.4340163767337799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1996513307094574, "epoch": 12.75, "learning_rate": 8.315591734502192e-06, "loss": 0.4068, "step": 15089, "task_loss": 0.31379055976867676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37884488701820374, "epoch": 12.76, "learning_rate": 8.312460864120227e-06, "loss": 0.4144, "step": 15090, "task_loss": 0.7757403254508972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39560070633888245, "epoch": 12.76, "learning_rate": 8.30932999373826e-06, "loss": 0.3792, "step": 15091, "task_loss": 0.45480918884277344 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5417288541793823, "epoch": 12.76, "learning_rate": 8.306199123356294e-06, "loss": 0.5084, "step": 15092, "task_loss": 0.6813675165176392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27661001682281494, "epoch": 12.76, "learning_rate": 8.303068252974327e-06, "loss": 0.3833, "step": 15093, "task_loss": 0.5405244827270508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6737629175186157, "epoch": 12.76, "learning_rate": 8.299937382592361e-06, "loss": 0.4761, "step": 15094, "task_loss": 1.156436800956726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3934747576713562, "epoch": 12.76, "learning_rate": 8.296806512210394e-06, "loss": 0.4088, "step": 15095, "task_loss": 0.2633993625640869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.424890398979187, "epoch": 12.76, "learning_rate": 8.29367564182843e-06, "loss": 0.4158, "step": 15096, "task_loss": 0.6755290031433105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24059560894966125, "epoch": 12.76, "learning_rate": 8.290544771446463e-06, "loss": 0.337, "step": 15097, "task_loss": 0.21757838129997253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6875965595245361, "epoch": 12.76, "learning_rate": 8.287413901064496e-06, "loss": 0.56, "step": 15098, "task_loss": 0.9525506496429443 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6170996427536011, "epoch": 12.76, "learning_rate": 8.28428303068253e-06, "loss": 0.4781, "step": 15099, "task_loss": 0.09396238625049591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5607519149780273, "epoch": 12.76, "learning_rate": 8.281152160300563e-06, "loss": 0.5249, "step": 15100, "task_loss": 0.8779231309890747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.515758752822876, "epoch": 12.76, "learning_rate": 8.278021289918597e-06, "loss": 0.4727, "step": 15101, "task_loss": 0.39772865176200867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33578214049339294, "epoch": 12.77, "learning_rate": 8.274890419536632e-06, "loss": 0.3865, "step": 15102, "task_loss": 0.4746631383895874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5520968437194824, "epoch": 12.77, "learning_rate": 8.271759549154666e-06, "loss": 0.4515, "step": 15103, "task_loss": 0.3915548324584961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1831102967262268, "epoch": 12.77, "learning_rate": 8.268628678772699e-06, "loss": 0.3685, "step": 15104, "task_loss": 0.26154056191444397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32874640822410583, "epoch": 12.77, "learning_rate": 8.265497808390734e-06, "loss": 0.4037, "step": 15105, "task_loss": 0.38359662890434265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4311980903148651, "epoch": 12.77, "learning_rate": 8.262366938008768e-06, "loss": 0.3395, "step": 15106, "task_loss": 0.3332987427711487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27064216136932373, "epoch": 12.77, "learning_rate": 8.259236067626801e-06, "loss": 0.3097, "step": 15107, "task_loss": 0.3173295557498932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4113309979438782, "epoch": 12.77, "learning_rate": 8.256105197244835e-06, "loss": 0.5423, "step": 15108, "task_loss": 0.5051556825637817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3381524085998535, "epoch": 12.77, "learning_rate": 8.252974326862868e-06, "loss": 0.4481, "step": 15109, "task_loss": 0.793628454208374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5149365067481995, "epoch": 12.77, "learning_rate": 8.249843456480902e-06, "loss": 0.4406, "step": 15110, "task_loss": 0.636237382888794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.518126368522644, "epoch": 12.77, "learning_rate": 8.246712586098935e-06, "loss": 0.5662, "step": 15111, "task_loss": 0.6599653959274292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6060736179351807, "epoch": 12.77, "learning_rate": 8.243581715716968e-06, "loss": 0.5743, "step": 15112, "task_loss": 1.0623202323913574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3276849091053009, "epoch": 12.77, "learning_rate": 8.240450845335004e-06, "loss": 0.432, "step": 15113, "task_loss": 0.39427804946899414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40531426668167114, "epoch": 12.78, "learning_rate": 8.237319974953039e-06, "loss": 0.383, "step": 15114, "task_loss": 1.0292454957962036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3040403723716736, "epoch": 12.78, "learning_rate": 8.234189104571072e-06, "loss": 0.4426, "step": 15115, "task_loss": 0.2899932861328125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45941615104675293, "epoch": 12.78, "learning_rate": 8.231058234189106e-06, "loss": 0.4505, "step": 15116, "task_loss": 0.8208111524581909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48519062995910645, "epoch": 12.78, "learning_rate": 8.22792736380714e-06, "loss": 0.4647, "step": 15117, "task_loss": 0.41498520970344543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6094616055488586, "epoch": 12.78, "learning_rate": 8.224796493425173e-06, "loss": 0.4717, "step": 15118, "task_loss": 0.29686683416366577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6549577713012695, "epoch": 12.78, "learning_rate": 8.221665623043206e-06, "loss": 0.526, "step": 15119, "task_loss": 0.6501080393791199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49366435408592224, "epoch": 12.78, "learning_rate": 8.21853475266124e-06, "loss": 0.5402, "step": 15120, "task_loss": 1.0373622179031372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3885655999183655, "epoch": 12.78, "learning_rate": 8.215403882279273e-06, "loss": 0.4478, "step": 15121, "task_loss": 0.43641456961631775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5823874473571777, "epoch": 12.78, "learning_rate": 8.212273011897308e-06, "loss": 0.4149, "step": 15122, "task_loss": 0.2769356369972229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.461578369140625, "epoch": 12.78, "learning_rate": 8.209142141515342e-06, "loss": 0.4146, "step": 15123, "task_loss": 1.0345757007598877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28108373284339905, "epoch": 12.78, "learning_rate": 8.206011271133375e-06, "loss": 0.4598, "step": 15124, "task_loss": 0.7791920900344849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20075403153896332, "epoch": 12.78, "learning_rate": 8.202880400751409e-06, "loss": 0.3356, "step": 15125, "task_loss": 0.047646913677453995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.616647481918335, "epoch": 12.79, "learning_rate": 8.199749530369442e-06, "loss": 0.5844, "step": 15126, "task_loss": 1.0795977115631104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5026472210884094, "epoch": 12.79, "learning_rate": 8.196618659987477e-06, "loss": 0.473, "step": 15127, "task_loss": 0.6997684240341187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23186761140823364, "epoch": 12.79, "learning_rate": 8.193487789605511e-06, "loss": 0.4128, "step": 15128, "task_loss": 0.528324544429779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42913562059402466, "epoch": 12.79, "learning_rate": 8.190356919223544e-06, "loss": 0.4653, "step": 15129, "task_loss": 0.6621823906898499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47786372900009155, "epoch": 12.79, "learning_rate": 8.187226048841578e-06, "loss": 0.3448, "step": 15130, "task_loss": 0.28554415702819824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7372227907180786, "epoch": 12.79, "learning_rate": 8.184095178459613e-06, "loss": 0.5516, "step": 15131, "task_loss": 1.7815985679626465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4972548484802246, "epoch": 12.79, "learning_rate": 8.180964308077646e-06, "loss": 0.4451, "step": 15132, "task_loss": 1.0583202838897705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38560301065444946, "epoch": 12.79, "learning_rate": 8.17783343769568e-06, "loss": 0.5625, "step": 15133, "task_loss": 0.580557107925415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5026946067810059, "epoch": 12.79, "learning_rate": 8.174702567313713e-06, "loss": 0.39, "step": 15134, "task_loss": 0.3972512483596802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40198445320129395, "epoch": 12.79, "learning_rate": 8.171571696931747e-06, "loss": 0.332, "step": 15135, "task_loss": 0.12112169712781906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5274694561958313, "epoch": 12.79, "learning_rate": 8.16844082654978e-06, "loss": 0.4055, "step": 15136, "task_loss": 1.4157201051712036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21591629087924957, "epoch": 12.79, "learning_rate": 8.165309956167814e-06, "loss": 0.3752, "step": 15137, "task_loss": 0.12735731899738312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3008508086204529, "epoch": 12.8, "learning_rate": 8.162179085785849e-06, "loss": 0.445, "step": 15138, "task_loss": 0.1483256071805954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2397840917110443, "epoch": 12.8, "learning_rate": 8.159048215403882e-06, "loss": 0.3237, "step": 15139, "task_loss": 0.1954803764820099 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22094877064228058, "epoch": 12.8, "learning_rate": 8.155917345021918e-06, "loss": 0.3596, "step": 15140, "task_loss": 0.4163200855255127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3794596195220947, "epoch": 12.8, "learning_rate": 8.152786474639951e-06, "loss": 0.3924, "step": 15141, "task_loss": 0.7091025710105896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46666640043258667, "epoch": 12.8, "learning_rate": 8.149655604257985e-06, "loss": 0.4143, "step": 15142, "task_loss": 0.8894482851028442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26045793294906616, "epoch": 12.8, "learning_rate": 8.146524733876018e-06, "loss": 0.4092, "step": 15143, "task_loss": 0.2824817895889282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4233574867248535, "epoch": 12.8, "learning_rate": 8.143393863494052e-06, "loss": 0.4733, "step": 15144, "task_loss": 0.1280556619167328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6126713156700134, "epoch": 12.8, "learning_rate": 8.140262993112085e-06, "loss": 0.4199, "step": 15145, "task_loss": 0.37902507185935974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40584132075309753, "epoch": 12.8, "learning_rate": 8.137132122730119e-06, "loss": 0.4636, "step": 15146, "task_loss": 0.9270690679550171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6905832290649414, "epoch": 12.8, "learning_rate": 8.134001252348154e-06, "loss": 0.4238, "step": 15147, "task_loss": 1.7999076843261719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2989085912704468, "epoch": 12.8, "learning_rate": 8.130870381966187e-06, "loss": 0.4193, "step": 15148, "task_loss": 0.7128101587295532 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4741278886795044, "epoch": 12.81, "learning_rate": 8.12773951158422e-06, "loss": 0.383, "step": 15149, "task_loss": 0.704267680644989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6060717701911926, "epoch": 12.81, "learning_rate": 8.124608641202254e-06, "loss": 0.42, "step": 15150, "task_loss": 1.131994366645813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4544646739959717, "epoch": 12.81, "learning_rate": 8.12147777082029e-06, "loss": 0.4304, "step": 15151, "task_loss": 1.322706699371338 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30471694469451904, "epoch": 12.81, "learning_rate": 8.118346900438323e-06, "loss": 0.2955, "step": 15152, "task_loss": 0.43322035670280457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.575636625289917, "epoch": 12.81, "learning_rate": 8.115216030056356e-06, "loss": 0.6032, "step": 15153, "task_loss": 0.7624963521957397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31838279962539673, "epoch": 12.81, "learning_rate": 8.11208515967439e-06, "loss": 0.4201, "step": 15154, "task_loss": 0.3550806939601898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3913980722427368, "epoch": 12.81, "learning_rate": 8.108954289292423e-06, "loss": 0.4532, "step": 15155, "task_loss": 0.059252988547086716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.12242439389228821, "epoch": 12.81, "learning_rate": 8.105823418910458e-06, "loss": 0.4472, "step": 15156, "task_loss": 0.021395418792963028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3696136474609375, "epoch": 12.81, "learning_rate": 8.102692548528492e-06, "loss": 0.39, "step": 15157, "task_loss": 0.31781628727912903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4102433919906616, "epoch": 12.81, "learning_rate": 8.099561678146525e-06, "loss": 0.3616, "step": 15158, "task_loss": 0.34029918909072876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40863290429115295, "epoch": 12.81, "learning_rate": 8.096430807764559e-06, "loss": 0.5075, "step": 15159, "task_loss": 0.37557753920555115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4345672130584717, "epoch": 12.81, "learning_rate": 8.093299937382592e-06, "loss": 0.3638, "step": 15160, "task_loss": 0.23027805984020233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4722925126552582, "epoch": 12.82, "learning_rate": 8.090169067000626e-06, "loss": 0.5095, "step": 15161, "task_loss": 0.6217884421348572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.526729941368103, "epoch": 12.82, "learning_rate": 8.08703819661866e-06, "loss": 0.5348, "step": 15162, "task_loss": 0.7768153548240662 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36488547921180725, "epoch": 12.82, "learning_rate": 8.083907326236693e-06, "loss": 0.2728, "step": 15163, "task_loss": 0.6121854186058044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31790822744369507, "epoch": 12.82, "learning_rate": 8.080776455854728e-06, "loss": 0.3657, "step": 15164, "task_loss": 0.4479720890522003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2967868149280548, "epoch": 12.82, "learning_rate": 8.077645585472763e-06, "loss": 0.5848, "step": 15165, "task_loss": 0.42774778604507446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26563897728919983, "epoch": 12.82, "learning_rate": 8.074514715090796e-06, "loss": 0.3802, "step": 15166, "task_loss": 0.2908059358596802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5762462019920349, "epoch": 12.82, "learning_rate": 8.07138384470883e-06, "loss": 0.5221, "step": 15167, "task_loss": 1.0857816934585571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36670345067977905, "epoch": 12.82, "learning_rate": 8.068252974326863e-06, "loss": 0.4712, "step": 15168, "task_loss": 0.9931355714797974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3589041829109192, "epoch": 12.82, "learning_rate": 8.065122103944897e-06, "loss": 0.3605, "step": 15169, "task_loss": 0.2488766759634018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3501005172729492, "epoch": 12.82, "learning_rate": 8.06199123356293e-06, "loss": 0.3635, "step": 15170, "task_loss": 0.36642810702323914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3068643808364868, "epoch": 12.82, "learning_rate": 8.058860363180964e-06, "loss": 0.3888, "step": 15171, "task_loss": 0.27529123425483704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4399908781051636, "epoch": 12.82, "learning_rate": 8.055729492798999e-06, "loss": 0.4446, "step": 15172, "task_loss": 0.9912252426147461 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4224013090133667, "epoch": 12.83, "learning_rate": 8.052598622417033e-06, "loss": 0.3992, "step": 15173, "task_loss": 0.2532173991203308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2874601483345032, "epoch": 12.83, "learning_rate": 8.049467752035066e-06, "loss": 0.4078, "step": 15174, "task_loss": 0.5967553853988647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5350997447967529, "epoch": 12.83, "learning_rate": 8.0463368816531e-06, "loss": 0.4497, "step": 15175, "task_loss": 1.5191155672073364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44628775119781494, "epoch": 12.83, "learning_rate": 8.043206011271135e-06, "loss": 0.3957, "step": 15176, "task_loss": 0.3648539185523987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17913776636123657, "epoch": 12.83, "learning_rate": 8.040075140889168e-06, "loss": 0.3824, "step": 15177, "task_loss": 0.35453081130981445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41314348578453064, "epoch": 12.83, "learning_rate": 8.036944270507202e-06, "loss": 0.4417, "step": 15178, "task_loss": 0.04466142877936363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30963993072509766, "epoch": 12.83, "learning_rate": 8.033813400125235e-06, "loss": 0.4834, "step": 15179, "task_loss": 0.4478282928466797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23368825018405914, "epoch": 12.83, "learning_rate": 8.030682529743269e-06, "loss": 0.2778, "step": 15180, "task_loss": 0.41587281227111816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6241939663887024, "epoch": 12.83, "learning_rate": 8.027551659361304e-06, "loss": 0.4262, "step": 15181, "task_loss": 0.4135003089904785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3891737163066864, "epoch": 12.83, "learning_rate": 8.024420788979337e-06, "loss": 0.5779, "step": 15182, "task_loss": 1.0710744857788086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.320060133934021, "epoch": 12.83, "learning_rate": 8.02128991859737e-06, "loss": 0.4247, "step": 15183, "task_loss": 0.4308677017688751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24384525418281555, "epoch": 12.83, "learning_rate": 8.018159048215404e-06, "loss": 0.3946, "step": 15184, "task_loss": 1.0173413753509521 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39347389340400696, "epoch": 12.84, "learning_rate": 8.015028177833438e-06, "loss": 0.4491, "step": 15185, "task_loss": 1.0560331344604492 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5356789231300354, "epoch": 12.84, "learning_rate": 8.011897307451471e-06, "loss": 0.4058, "step": 15186, "task_loss": 0.7919492721557617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6641342043876648, "epoch": 12.84, "learning_rate": 8.008766437069505e-06, "loss": 0.4014, "step": 15187, "task_loss": 0.31703341007232666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6331648826599121, "epoch": 12.84, "learning_rate": 8.00563556668754e-06, "loss": 0.508, "step": 15188, "task_loss": 0.8470935225486755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4910878539085388, "epoch": 12.84, "learning_rate": 8.002504696305573e-06, "loss": 0.4993, "step": 15189, "task_loss": 0.18939638137817383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.494658887386322, "epoch": 12.84, "learning_rate": 7.999373825923608e-06, "loss": 0.5324, "step": 15190, "task_loss": 0.30864813923835754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3688880503177643, "epoch": 12.84, "learning_rate": 7.996242955541642e-06, "loss": 0.4139, "step": 15191, "task_loss": 0.5797186493873596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3158721923828125, "epoch": 12.84, "learning_rate": 7.993112085159675e-06, "loss": 0.4245, "step": 15192, "task_loss": 0.6146361827850342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.487257719039917, "epoch": 12.84, "learning_rate": 7.989981214777709e-06, "loss": 0.6994, "step": 15193, "task_loss": 0.6171837449073792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7400456070899963, "epoch": 12.84, "learning_rate": 7.986850344395742e-06, "loss": 0.5472, "step": 15194, "task_loss": 0.31951698660850525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3933796286582947, "epoch": 12.84, "learning_rate": 7.983719474013776e-06, "loss": 0.4742, "step": 15195, "task_loss": 0.1987144947052002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5569505095481873, "epoch": 12.84, "learning_rate": 7.98058860363181e-06, "loss": 0.4856, "step": 15196, "task_loss": 0.3642317056655884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5024429559707642, "epoch": 12.85, "learning_rate": 7.977457733249843e-06, "loss": 0.4593, "step": 15197, "task_loss": 0.8055276870727539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35167714953422546, "epoch": 12.85, "learning_rate": 7.974326862867878e-06, "loss": 0.3425, "step": 15198, "task_loss": 0.09174197167158127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3416876792907715, "epoch": 12.85, "learning_rate": 7.971195992485911e-06, "loss": 0.3794, "step": 15199, "task_loss": 0.5483516454696655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5245484709739685, "epoch": 12.85, "learning_rate": 7.968065122103947e-06, "loss": 0.5144, "step": 15200, "task_loss": 0.39217954874038696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4142347574234009, "epoch": 12.85, "learning_rate": 7.96493425172198e-06, "loss": 0.5101, "step": 15201, "task_loss": 1.0188331604003906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46882742643356323, "epoch": 12.85, "learning_rate": 7.961803381340013e-06, "loss": 0.4721, "step": 15202, "task_loss": 0.09661860018968582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3688562512397766, "epoch": 12.85, "learning_rate": 7.958672510958047e-06, "loss": 0.4287, "step": 15203, "task_loss": 1.338944673538208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33169177174568176, "epoch": 12.85, "learning_rate": 7.95554164057608e-06, "loss": 0.3804, "step": 15204, "task_loss": 0.8495321869850159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4540468454360962, "epoch": 12.85, "learning_rate": 7.952410770194114e-06, "loss": 0.4888, "step": 15205, "task_loss": 0.23345641791820526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3395997881889343, "epoch": 12.85, "learning_rate": 7.949279899812147e-06, "loss": 0.4309, "step": 15206, "task_loss": 0.25209423899650574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7066575884819031, "epoch": 12.85, "learning_rate": 7.946149029430183e-06, "loss": 0.5005, "step": 15207, "task_loss": 0.8855158090591431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3238891661167145, "epoch": 12.85, "learning_rate": 7.943018159048216e-06, "loss": 0.5097, "step": 15208, "task_loss": 0.9093174934387207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44236817955970764, "epoch": 12.86, "learning_rate": 7.93988728866625e-06, "loss": 0.5763, "step": 15209, "task_loss": 1.2390238046646118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33615148067474365, "epoch": 12.86, "learning_rate": 7.936756418284283e-06, "loss": 0.5912, "step": 15210, "task_loss": 0.3758845627307892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5341407060623169, "epoch": 12.86, "learning_rate": 7.933625547902316e-06, "loss": 0.6009, "step": 15211, "task_loss": 0.9127920866012573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3309248983860016, "epoch": 12.86, "learning_rate": 7.93049467752035e-06, "loss": 0.5695, "step": 15212, "task_loss": 0.5305825471878052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40733739733695984, "epoch": 12.86, "learning_rate": 7.927363807138385e-06, "loss": 0.4253, "step": 15213, "task_loss": 0.7410995960235596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41072380542755127, "epoch": 12.86, "learning_rate": 7.924232936756419e-06, "loss": 0.4794, "step": 15214, "task_loss": 0.44091862440109253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.329230397939682, "epoch": 12.86, "learning_rate": 7.921102066374452e-06, "loss": 0.4278, "step": 15215, "task_loss": 0.10100438445806503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5159057378768921, "epoch": 12.86, "learning_rate": 7.917971195992487e-06, "loss": 0.5437, "step": 15216, "task_loss": 0.6670057773590088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6916142106056213, "epoch": 12.86, "learning_rate": 7.91484032561052e-06, "loss": 0.5287, "step": 15217, "task_loss": 0.4844367504119873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19014376401901245, "epoch": 12.86, "learning_rate": 7.911709455228554e-06, "loss": 0.4216, "step": 15218, "task_loss": 0.18856310844421387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3435025215148926, "epoch": 12.86, "learning_rate": 7.908578584846588e-06, "loss": 0.5397, "step": 15219, "task_loss": 0.8422865867614746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.267857164144516, "epoch": 12.87, "learning_rate": 7.905447714464621e-06, "loss": 0.4231, "step": 15220, "task_loss": 0.8602243065834045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5416452288627625, "epoch": 12.87, "learning_rate": 7.902316844082655e-06, "loss": 0.4758, "step": 15221, "task_loss": 0.6842418909072876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38208067417144775, "epoch": 12.87, "learning_rate": 7.899185973700688e-06, "loss": 0.3331, "step": 15222, "task_loss": 0.5484828948974609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4916917681694031, "epoch": 12.87, "learning_rate": 7.896055103318723e-06, "loss": 0.4583, "step": 15223, "task_loss": 0.8663843870162964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35101377964019775, "epoch": 12.87, "learning_rate": 7.892924232936757e-06, "loss": 0.4186, "step": 15224, "task_loss": 0.4190223813056946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5463391542434692, "epoch": 12.87, "learning_rate": 7.889793362554792e-06, "loss": 0.4317, "step": 15225, "task_loss": 0.538010835647583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5483894348144531, "epoch": 12.87, "learning_rate": 7.886662492172825e-06, "loss": 0.4288, "step": 15226, "task_loss": 0.3964768350124359 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37194859981536865, "epoch": 12.87, "learning_rate": 7.883531621790859e-06, "loss": 0.5341, "step": 15227, "task_loss": 0.642602801322937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4046512246131897, "epoch": 12.87, "learning_rate": 7.880400751408892e-06, "loss": 0.5401, "step": 15228, "task_loss": 0.8445967435836792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.426999568939209, "epoch": 12.87, "learning_rate": 7.877269881026926e-06, "loss": 0.5493, "step": 15229, "task_loss": 0.7257580161094666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4662858247756958, "epoch": 12.87, "learning_rate": 7.87413901064496e-06, "loss": 0.4378, "step": 15230, "task_loss": 0.5252858400344849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2679852545261383, "epoch": 12.87, "learning_rate": 7.871008140262993e-06, "loss": 0.47, "step": 15231, "task_loss": 1.5578557252883911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4135597050189972, "epoch": 12.88, "learning_rate": 7.867877269881028e-06, "loss": 0.3945, "step": 15232, "task_loss": 1.0302000045776367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5569745302200317, "epoch": 12.88, "learning_rate": 7.864746399499061e-06, "loss": 0.4356, "step": 15233, "task_loss": 0.7342290282249451 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25607043504714966, "epoch": 12.88, "learning_rate": 7.861615529117095e-06, "loss": 0.3896, "step": 15234, "task_loss": 0.7711193561553955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45784619450569153, "epoch": 12.88, "learning_rate": 7.858484658735128e-06, "loss": 0.5141, "step": 15235, "task_loss": 1.2063376903533936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.371005117893219, "epoch": 12.88, "learning_rate": 7.855353788353162e-06, "loss": 0.4102, "step": 15236, "task_loss": 0.3092201054096222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19859810173511505, "epoch": 12.88, "learning_rate": 7.852222917971197e-06, "loss": 0.3669, "step": 15237, "task_loss": 0.49385368824005127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24509066343307495, "epoch": 12.88, "learning_rate": 7.84909204758923e-06, "loss": 0.4255, "step": 15238, "task_loss": 0.2668977379798889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2680928707122803, "epoch": 12.88, "learning_rate": 7.845961177207264e-06, "loss": 0.4098, "step": 15239, "task_loss": 0.16901789605617523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34642505645751953, "epoch": 12.88, "learning_rate": 7.842830306825297e-06, "loss": 0.4267, "step": 15240, "task_loss": 0.9716415405273438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31507277488708496, "epoch": 12.88, "learning_rate": 7.839699436443333e-06, "loss": 0.3277, "step": 15241, "task_loss": 0.6092053055763245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46610328555107117, "epoch": 12.88, "learning_rate": 7.836568566061366e-06, "loss": 0.4165, "step": 15242, "task_loss": 0.944611132144928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4540313482284546, "epoch": 12.88, "learning_rate": 7.8334376956794e-06, "loss": 0.4591, "step": 15243, "task_loss": 0.4704122841358185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23531608283519745, "epoch": 12.89, "learning_rate": 7.830306825297433e-06, "loss": 0.3304, "step": 15244, "task_loss": 0.10988514125347137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5617852210998535, "epoch": 12.89, "learning_rate": 7.827175954915466e-06, "loss": 0.4114, "step": 15245, "task_loss": 0.4992009103298187 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5601819753646851, "epoch": 12.89, "learning_rate": 7.8240450845335e-06, "loss": 0.3576, "step": 15246, "task_loss": 0.9130175113677979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4167180061340332, "epoch": 12.89, "learning_rate": 7.820914214151533e-06, "loss": 0.5038, "step": 15247, "task_loss": 0.7111597061157227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5416098833084106, "epoch": 12.89, "learning_rate": 7.817783343769569e-06, "loss": 0.4807, "step": 15248, "task_loss": 0.9027553200721741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5279645919799805, "epoch": 12.89, "learning_rate": 7.814652473387602e-06, "loss": 0.5202, "step": 15249, "task_loss": 0.40995079278945923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3399283289909363, "epoch": 12.89, "learning_rate": 7.811521603005637e-06, "loss": 0.2855, "step": 15250, "task_loss": 0.7613968849182129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5155447125434875, "epoch": 12.89, "learning_rate": 7.80839073262367e-06, "loss": 0.4282, "step": 15251, "task_loss": 0.17406253516674042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40457624197006226, "epoch": 12.89, "learning_rate": 7.805259862241704e-06, "loss": 0.4313, "step": 15252, "task_loss": 0.8109288215637207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6773537397384644, "epoch": 12.89, "learning_rate": 7.802128991859738e-06, "loss": 0.4758, "step": 15253, "task_loss": 0.910824179649353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47935110330581665, "epoch": 12.89, "learning_rate": 7.798998121477771e-06, "loss": 0.5391, "step": 15254, "task_loss": 1.1123260259628296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5191774964332581, "epoch": 12.89, "learning_rate": 7.795867251095805e-06, "loss": 0.5657, "step": 15255, "task_loss": 0.4282732307910919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5225976705551147, "epoch": 12.9, "learning_rate": 7.792736380713838e-06, "loss": 0.4671, "step": 15256, "task_loss": 1.6111301183700562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4163248538970947, "epoch": 12.9, "learning_rate": 7.789605510331873e-06, "loss": 0.3331, "step": 15257, "task_loss": 0.21828517317771912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3991513252258301, "epoch": 12.9, "learning_rate": 7.786474639949907e-06, "loss": 0.4062, "step": 15258, "task_loss": 0.6669830083847046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4054928421974182, "epoch": 12.9, "learning_rate": 7.78334376956794e-06, "loss": 0.4311, "step": 15259, "task_loss": 1.0585383176803589 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3025341033935547, "epoch": 12.9, "learning_rate": 7.780212899185974e-06, "loss": 0.5465, "step": 15260, "task_loss": 0.1759706288576126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49994951486587524, "epoch": 12.9, "learning_rate": 7.777082028804007e-06, "loss": 0.4411, "step": 15261, "task_loss": 0.9649634957313538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24849379062652588, "epoch": 12.9, "learning_rate": 7.773951158422042e-06, "loss": 0.3343, "step": 15262, "task_loss": 0.35998034477233887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38301873207092285, "epoch": 12.9, "learning_rate": 7.770820288040076e-06, "loss": 0.4518, "step": 15263, "task_loss": 0.4017236530780792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46912863850593567, "epoch": 12.9, "learning_rate": 7.76768941765811e-06, "loss": 0.3951, "step": 15264, "task_loss": 0.4002848267555237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45375382900238037, "epoch": 12.9, "learning_rate": 7.764558547276143e-06, "loss": 0.5585, "step": 15265, "task_loss": 0.18672826886177063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24296584725379944, "epoch": 12.9, "learning_rate": 7.761427676894178e-06, "loss": 0.4316, "step": 15266, "task_loss": 0.24157153069972992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3325657844543457, "epoch": 12.9, "learning_rate": 7.758296806512211e-06, "loss": 0.4409, "step": 15267, "task_loss": 0.4601380527019501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4123963415622711, "epoch": 12.91, "learning_rate": 7.755165936130245e-06, "loss": 0.4696, "step": 15268, "task_loss": 0.2915157377719879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49533140659332275, "epoch": 12.91, "learning_rate": 7.752035065748278e-06, "loss": 0.4202, "step": 15269, "task_loss": 0.5265724658966064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5072195529937744, "epoch": 12.91, "learning_rate": 7.748904195366312e-06, "loss": 0.4945, "step": 15270, "task_loss": 0.33158162236213684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17207181453704834, "epoch": 12.91, "learning_rate": 7.745773324984345e-06, "loss": 0.4582, "step": 15271, "task_loss": 1.0218472480773926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5777761936187744, "epoch": 12.91, "learning_rate": 7.742642454602379e-06, "loss": 0.4994, "step": 15272, "task_loss": 0.723659336566925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49003905057907104, "epoch": 12.91, "learning_rate": 7.739511584220412e-06, "loss": 0.3935, "step": 15273, "task_loss": 0.6310247778892517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42810410261154175, "epoch": 12.91, "learning_rate": 7.736380713838447e-06, "loss": 0.5485, "step": 15274, "task_loss": 0.3607807755470276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6322226524353027, "epoch": 12.91, "learning_rate": 7.733249843456483e-06, "loss": 0.5878, "step": 15275, "task_loss": 0.5664782524108887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44558584690093994, "epoch": 12.91, "learning_rate": 7.730118973074516e-06, "loss": 0.4732, "step": 15276, "task_loss": 0.7604743838310242 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5896517038345337, "epoch": 12.91, "learning_rate": 7.72698810269255e-06, "loss": 0.5078, "step": 15277, "task_loss": 0.4782739579677582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.303438663482666, "epoch": 12.91, "learning_rate": 7.723857232310583e-06, "loss": 0.465, "step": 15278, "task_loss": 0.22025980055332184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3217439651489258, "epoch": 12.91, "learning_rate": 7.720726361928616e-06, "loss": 0.4649, "step": 15279, "task_loss": 0.3325444161891937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22881753742694855, "epoch": 12.92, "learning_rate": 7.71759549154665e-06, "loss": 0.4272, "step": 15280, "task_loss": 0.7499513626098633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35345539450645447, "epoch": 12.92, "learning_rate": 7.714464621164683e-06, "loss": 0.3646, "step": 15281, "task_loss": 0.6055485606193542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25629615783691406, "epoch": 12.92, "learning_rate": 7.711333750782717e-06, "loss": 0.3839, "step": 15282, "task_loss": 1.5923879146575928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.374103307723999, "epoch": 12.92, "learning_rate": 7.708202880400752e-06, "loss": 0.4577, "step": 15283, "task_loss": 0.6170212626457214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33950555324554443, "epoch": 12.92, "learning_rate": 7.705072010018786e-06, "loss": 0.3065, "step": 15284, "task_loss": 0.39084869623184204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.262216717004776, "epoch": 12.92, "learning_rate": 7.701941139636819e-06, "loss": 0.566, "step": 15285, "task_loss": 0.22768767178058624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.718660295009613, "epoch": 12.92, "learning_rate": 7.698810269254854e-06, "loss": 0.3568, "step": 15286, "task_loss": 0.7661762833595276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6054259538650513, "epoch": 12.92, "learning_rate": 7.695679398872888e-06, "loss": 0.4418, "step": 15287, "task_loss": 0.4746646583080292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3929705023765564, "epoch": 12.92, "learning_rate": 7.692548528490921e-06, "loss": 0.4218, "step": 15288, "task_loss": 0.2747204303741455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33548861742019653, "epoch": 12.92, "learning_rate": 7.689417658108955e-06, "loss": 0.4425, "step": 15289, "task_loss": 1.104590892791748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32473766803741455, "epoch": 12.92, "learning_rate": 7.686286787726988e-06, "loss": 0.297, "step": 15290, "task_loss": 0.6890411376953125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8009471893310547, "epoch": 12.93, "learning_rate": 7.683155917345022e-06, "loss": 0.5009, "step": 15291, "task_loss": 0.6574572920799255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5028818249702454, "epoch": 12.93, "learning_rate": 7.680025046963057e-06, "loss": 0.4784, "step": 15292, "task_loss": 0.30788296461105347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7119000554084778, "epoch": 12.93, "learning_rate": 7.67689417658109e-06, "loss": 0.5721, "step": 15293, "task_loss": 0.32893791794776917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24635136127471924, "epoch": 12.93, "learning_rate": 7.673763306199124e-06, "loss": 0.4376, "step": 15294, "task_loss": 0.17312178015708923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.781951904296875, "epoch": 12.93, "learning_rate": 7.670632435817157e-06, "loss": 0.5946, "step": 15295, "task_loss": 1.1767971515655518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5624918341636658, "epoch": 12.93, "learning_rate": 7.66750156543519e-06, "loss": 0.5643, "step": 15296, "task_loss": 1.746488332748413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30762070417404175, "epoch": 12.93, "learning_rate": 7.664370695053224e-06, "loss": 0.3586, "step": 15297, "task_loss": 0.15987111628055573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2429676502943039, "epoch": 12.93, "learning_rate": 7.66123982467126e-06, "loss": 0.5029, "step": 15298, "task_loss": 0.37764132022857666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27134382724761963, "epoch": 12.93, "learning_rate": 7.658108954289293e-06, "loss": 0.4105, "step": 15299, "task_loss": 0.13080085813999176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3412669897079468, "epoch": 12.93, "learning_rate": 7.654978083907326e-06, "loss": 0.4799, "step": 15300, "task_loss": 0.4067259728908539 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28756630420684814, "epoch": 12.93, "learning_rate": 7.651847213525361e-06, "loss": 0.4376, "step": 15301, "task_loss": 1.0748213529586792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.399681955575943, "epoch": 12.93, "learning_rate": 7.648716343143395e-06, "loss": 0.4083, "step": 15302, "task_loss": 0.8346064686775208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3580392003059387, "epoch": 12.94, "learning_rate": 7.645585472761428e-06, "loss": 0.4069, "step": 15303, "task_loss": 0.33361729979515076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41818463802337646, "epoch": 12.94, "learning_rate": 7.642454602379462e-06, "loss": 0.4638, "step": 15304, "task_loss": 0.6627475023269653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4708825349807739, "epoch": 12.94, "learning_rate": 7.639323731997495e-06, "loss": 0.4701, "step": 15305, "task_loss": 0.7985402941703796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3810613453388214, "epoch": 12.94, "learning_rate": 7.636192861615529e-06, "loss": 0.4613, "step": 15306, "task_loss": 0.2653803527355194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7278398275375366, "epoch": 12.94, "learning_rate": 7.633061991233562e-06, "loss": 0.5722, "step": 15307, "task_loss": 0.6845921277999878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38235291838645935, "epoch": 12.94, "learning_rate": 7.629931120851597e-06, "loss": 0.3731, "step": 15308, "task_loss": 0.7196389436721802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28251203894615173, "epoch": 12.94, "learning_rate": 7.62680025046963e-06, "loss": 0.3977, "step": 15309, "task_loss": 0.40877634286880493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4327950179576874, "epoch": 12.94, "learning_rate": 7.623669380087664e-06, "loss": 0.4707, "step": 15310, "task_loss": 0.18791384994983673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47162336111068726, "epoch": 12.94, "learning_rate": 7.6205385097056995e-06, "loss": 0.5101, "step": 15311, "task_loss": 0.6834566593170166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6825882196426392, "epoch": 12.94, "learning_rate": 7.617407639323733e-06, "loss": 0.4344, "step": 15312, "task_loss": 1.1438628435134888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5120607018470764, "epoch": 12.94, "learning_rate": 7.6142767689417665e-06, "loss": 0.3959, "step": 15313, "task_loss": 0.9145627021789551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6952458024024963, "epoch": 12.94, "learning_rate": 7.6111458985598e-06, "loss": 0.5665, "step": 15314, "task_loss": 0.7577742338180542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44008904695510864, "epoch": 12.95, "learning_rate": 7.608015028177834e-06, "loss": 0.4898, "step": 15315, "task_loss": 0.48448631167411804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29470378160476685, "epoch": 12.95, "learning_rate": 7.604884157795868e-06, "loss": 0.3405, "step": 15316, "task_loss": 0.23403948545455933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47543591260910034, "epoch": 12.95, "learning_rate": 7.601753287413901e-06, "loss": 0.4568, "step": 15317, "task_loss": 0.6803188323974609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5055257081985474, "epoch": 12.95, "learning_rate": 7.598622417031935e-06, "loss": 0.5578, "step": 15318, "task_loss": 0.7982637286186218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31345149874687195, "epoch": 12.95, "learning_rate": 7.595491546649969e-06, "loss": 0.3351, "step": 15319, "task_loss": 0.48321229219436646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.327326238155365, "epoch": 12.95, "learning_rate": 7.5923606762680025e-06, "loss": 0.4597, "step": 15320, "task_loss": 0.5853113532066345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3606812059879303, "epoch": 12.95, "learning_rate": 7.589229805886036e-06, "loss": 0.3376, "step": 15321, "task_loss": 0.7681702375411987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7143148183822632, "epoch": 12.95, "learning_rate": 7.5860989355040695e-06, "loss": 0.5461, "step": 15322, "task_loss": 0.34911614656448364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3993212580680847, "epoch": 12.95, "learning_rate": 7.582968065122105e-06, "loss": 0.4403, "step": 15323, "task_loss": 0.5227065086364746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18168854713439941, "epoch": 12.95, "learning_rate": 7.579837194740139e-06, "loss": 0.469, "step": 15324, "task_loss": 0.05253950506448746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32148808240890503, "epoch": 12.95, "learning_rate": 7.576706324358172e-06, "loss": 0.4021, "step": 15325, "task_loss": 0.3869611620903015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6484072804450989, "epoch": 12.95, "learning_rate": 7.573575453976206e-06, "loss": 0.5715, "step": 15326, "task_loss": 1.1163721084594727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38472700119018555, "epoch": 12.96, "learning_rate": 7.570444583594239e-06, "loss": 0.4884, "step": 15327, "task_loss": 0.8253822922706604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3755068778991699, "epoch": 12.96, "learning_rate": 7.567313713212274e-06, "loss": 0.3596, "step": 15328, "task_loss": 1.0673632621765137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3576873540878296, "epoch": 12.96, "learning_rate": 7.564182842830307e-06, "loss": 0.4909, "step": 15329, "task_loss": 0.5310108065605164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5460309982299805, "epoch": 12.96, "learning_rate": 7.561051972448341e-06, "loss": 0.4855, "step": 15330, "task_loss": 0.22078770399093628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3876587152481079, "epoch": 12.96, "learning_rate": 7.557921102066375e-06, "loss": 0.4403, "step": 15331, "task_loss": 0.658905565738678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4022054076194763, "epoch": 12.96, "learning_rate": 7.5547902316844084e-06, "loss": 0.5432, "step": 15332, "task_loss": 0.6873363852500916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3286052346229553, "epoch": 12.96, "learning_rate": 7.551659361302442e-06, "loss": 0.474, "step": 15333, "task_loss": 1.0945624113082886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45007017254829407, "epoch": 12.96, "learning_rate": 7.548528490920475e-06, "loss": 0.3448, "step": 15334, "task_loss": 0.6118227243423462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5970985293388367, "epoch": 12.96, "learning_rate": 7.5453976205385106e-06, "loss": 0.5094, "step": 15335, "task_loss": 0.9049884676933289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4584995210170746, "epoch": 12.96, "learning_rate": 7.542266750156544e-06, "loss": 0.4613, "step": 15336, "task_loss": 0.6082422733306885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5734057426452637, "epoch": 12.96, "learning_rate": 7.539135879774578e-06, "loss": 0.521, "step": 15337, "task_loss": 0.5921640396118164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3252413868904114, "epoch": 12.96, "learning_rate": 7.536005009392612e-06, "loss": 0.4559, "step": 15338, "task_loss": 0.38807785511016846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5344895124435425, "epoch": 12.97, "learning_rate": 7.532874139010645e-06, "loss": 0.6144, "step": 15339, "task_loss": 1.4255121946334839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5660786628723145, "epoch": 12.97, "learning_rate": 7.52974326862868e-06, "loss": 0.5282, "step": 15340, "task_loss": 0.550841748714447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5434730648994446, "epoch": 12.97, "learning_rate": 7.526612398246713e-06, "loss": 0.3414, "step": 15341, "task_loss": 0.7330165505409241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30182719230651855, "epoch": 12.97, "learning_rate": 7.523481527864747e-06, "loss": 0.3685, "step": 15342, "task_loss": 0.46499672532081604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3826676607131958, "epoch": 12.97, "learning_rate": 7.52035065748278e-06, "loss": 0.5255, "step": 15343, "task_loss": 0.3070363998413086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6877956390380859, "epoch": 12.97, "learning_rate": 7.517219787100814e-06, "loss": 0.5365, "step": 15344, "task_loss": 0.7236526608467102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3663501441478729, "epoch": 12.97, "learning_rate": 7.514088916718848e-06, "loss": 0.4052, "step": 15345, "task_loss": 0.7101629972457886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2748071253299713, "epoch": 12.97, "learning_rate": 7.510958046336881e-06, "loss": 0.3434, "step": 15346, "task_loss": 1.4028936624526978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6319712400436401, "epoch": 12.97, "learning_rate": 7.507827175954915e-06, "loss": 0.5747, "step": 15347, "task_loss": 1.2346066236495972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41050148010253906, "epoch": 12.97, "learning_rate": 7.50469630557295e-06, "loss": 0.4967, "step": 15348, "task_loss": 1.2430706024169922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5288330912590027, "epoch": 12.97, "learning_rate": 7.501565435190984e-06, "loss": 0.5081, "step": 15349, "task_loss": 0.3331802487373352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5100999474525452, "epoch": 12.97, "learning_rate": 7.498434564809018e-06, "loss": 0.4747, "step": 15350, "task_loss": 0.5292544364929199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5038935542106628, "epoch": 12.98, "learning_rate": 7.495303694427051e-06, "loss": 0.4626, "step": 15351, "task_loss": 0.48801976442337036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5286833047866821, "epoch": 12.98, "learning_rate": 7.492172824045085e-06, "loss": 0.428, "step": 15352, "task_loss": 0.9028392434120178 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4898107647895813, "epoch": 12.98, "learning_rate": 7.489041953663119e-06, "loss": 0.5084, "step": 15353, "task_loss": 0.7478304505348206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2902902364730835, "epoch": 12.98, "learning_rate": 7.4859110832811525e-06, "loss": 0.4133, "step": 15354, "task_loss": 0.4177205264568329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6715673804283142, "epoch": 12.98, "learning_rate": 7.482780212899186e-06, "loss": 0.5017, "step": 15355, "task_loss": 0.6669435501098633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4699553847312927, "epoch": 12.98, "learning_rate": 7.4796493425172195e-06, "loss": 0.504, "step": 15356, "task_loss": 0.9070311784744263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7150346040725708, "epoch": 12.98, "learning_rate": 7.476518472135254e-06, "loss": 0.4819, "step": 15357, "task_loss": 0.5739755034446716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40844160318374634, "epoch": 12.98, "learning_rate": 7.473387601753287e-06, "loss": 0.4833, "step": 15358, "task_loss": 0.6216624975204468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.365091472864151, "epoch": 12.98, "learning_rate": 7.470256731371321e-06, "loss": 0.6336, "step": 15359, "task_loss": 0.5667142868041992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2714356780052185, "epoch": 12.98, "learning_rate": 7.467125860989356e-06, "loss": 0.455, "step": 15360, "task_loss": 0.7003498077392578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5175207853317261, "epoch": 12.98, "learning_rate": 7.463994990607389e-06, "loss": 0.3489, "step": 15361, "task_loss": 0.8288613557815552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38513413071632385, "epoch": 12.99, "learning_rate": 7.460864120225424e-06, "loss": 0.5313, "step": 15362, "task_loss": 1.0728403329849243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.342849999666214, "epoch": 12.99, "learning_rate": 7.457733249843457e-06, "loss": 0.452, "step": 15363, "task_loss": 0.9992603063583374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41736721992492676, "epoch": 12.99, "learning_rate": 7.454602379461491e-06, "loss": 0.5785, "step": 15364, "task_loss": 0.7730987071990967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5396575331687927, "epoch": 12.99, "learning_rate": 7.451471509079524e-06, "loss": 0.4833, "step": 15365, "task_loss": 0.490310400724411 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23182907700538635, "epoch": 12.99, "learning_rate": 7.4483406386975585e-06, "loss": 0.468, "step": 15366, "task_loss": 0.2747322916984558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5332299470901489, "epoch": 12.99, "learning_rate": 7.445209768315592e-06, "loss": 0.3689, "step": 15367, "task_loss": 0.3321979343891144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5534368753433228, "epoch": 12.99, "learning_rate": 7.442078897933625e-06, "loss": 0.5218, "step": 15368, "task_loss": 0.7460137009620667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37796664237976074, "epoch": 12.99, "learning_rate": 7.43894802755166e-06, "loss": 0.6047, "step": 15369, "task_loss": 0.1892431378364563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.667976438999176, "epoch": 12.99, "learning_rate": 7.435817157169693e-06, "loss": 0.4791, "step": 15370, "task_loss": 0.20628605782985687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5583152770996094, "epoch": 12.99, "learning_rate": 7.432686286787727e-06, "loss": 0.447, "step": 15371, "task_loss": 0.9003366231918335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3601500988006592, "epoch": 12.99, "learning_rate": 7.429555416405762e-06, "loss": 0.3773, "step": 15372, "task_loss": 0.3166872262954712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1878199726343155, "epoch": 12.99, "learning_rate": 7.426424546023795e-06, "loss": 0.326, "step": 15373, "task_loss": 0.4057905375957489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28328895568847656, "epoch": 13.0, "learning_rate": 7.423293675641829e-06, "loss": 0.2842, "step": 15374, "task_loss": 0.43711838126182556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45870229601860046, "epoch": 13.0, "learning_rate": 7.420162805259863e-06, "loss": 0.4214, "step": 15375, "task_loss": 1.4152915477752686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5225862860679626, "epoch": 13.0, "learning_rate": 7.417031934877897e-06, "loss": 0.3428, "step": 15376, "task_loss": 0.2687109112739563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3653562068939209, "epoch": 13.0, "learning_rate": 7.41390106449593e-06, "loss": 0.325, "step": 15377, "task_loss": 0.2463110238313675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17662674188613892, "epoch": 13.0, "learning_rate": 7.410770194113964e-06, "loss": 0.3663, "step": 15378, "task_loss": 0.9427986145019531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3081802725791931, "epoch": 13.0, "learning_rate": 7.407639323731998e-06, "loss": 0.3917, "step": 15379, "task_loss": 0.8130849003791809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27258771657943726, "epoch": 13.0, "learning_rate": 7.404508453350031e-06, "loss": 0.7301, "step": 15380, "task_loss": 0.678367555141449 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25799044966697693, "epoch": 13.0, "learning_rate": 7.401377582968065e-06, "loss": 0.6031, "step": 15381, "task_loss": 0.4641134440898895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47943204641342163, "epoch": 13.0, "learning_rate": 7.398246712586099e-06, "loss": 0.5403, "step": 15382, "task_loss": 0.2238268405199051 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32236215472221375, "epoch": 13.0, "learning_rate": 7.395115842204133e-06, "loss": 0.5588, "step": 15383, "task_loss": 0.02191711589694023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49950340390205383, "epoch": 13.0, "learning_rate": 7.391984971822168e-06, "loss": 0.434, "step": 15384, "task_loss": 1.1389753818511963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5366899967193604, "epoch": 13.01, "learning_rate": 7.388854101440201e-06, "loss": 0.4137, "step": 15385, "task_loss": 0.3923911452293396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5020517706871033, "epoch": 13.01, "learning_rate": 7.385723231058235e-06, "loss": 0.4667, "step": 15386, "task_loss": 0.44370725750923157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30842140316963196, "epoch": 13.01, "learning_rate": 7.382592360676269e-06, "loss": 0.4728, "step": 15387, "task_loss": 0.6410271525382996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.273977130651474, "epoch": 13.01, "learning_rate": 7.3794614902943025e-06, "loss": 0.2995, "step": 15388, "task_loss": 0.09168991446495056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3294694125652313, "epoch": 13.01, "learning_rate": 7.376330619912336e-06, "loss": 0.4517, "step": 15389, "task_loss": 0.5184553861618042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4567108154296875, "epoch": 13.01, "learning_rate": 7.3731997495303695e-06, "loss": 0.4858, "step": 15390, "task_loss": 1.633603811264038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2776956856250763, "epoch": 13.01, "learning_rate": 7.370068879148404e-06, "loss": 0.3621, "step": 15391, "task_loss": 0.716662585735321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5726830959320068, "epoch": 13.01, "learning_rate": 7.366938008766437e-06, "loss": 0.4088, "step": 15392, "task_loss": 0.6336005926132202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2244763821363449, "epoch": 13.01, "learning_rate": 7.363807138384471e-06, "loss": 0.5382, "step": 15393, "task_loss": 0.042773351073265076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4866150915622711, "epoch": 13.01, "learning_rate": 7.360676268002504e-06, "loss": 0.4493, "step": 15394, "task_loss": 0.46690407395362854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19327935576438904, "epoch": 13.01, "learning_rate": 7.3575453976205386e-06, "loss": 0.386, "step": 15395, "task_loss": 0.25247326493263245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2710760831832886, "epoch": 13.01, "learning_rate": 7.354414527238572e-06, "loss": 0.3894, "step": 15396, "task_loss": 0.3773190379142761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2996262311935425, "epoch": 13.02, "learning_rate": 7.351283656856607e-06, "loss": 0.2792, "step": 15397, "task_loss": 0.5466613173484802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.679813027381897, "epoch": 13.02, "learning_rate": 7.348152786474641e-06, "loss": 0.6522, "step": 15398, "task_loss": 0.4399926960468292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44117271900177, "epoch": 13.02, "learning_rate": 7.345021916092674e-06, "loss": 0.3879, "step": 15399, "task_loss": 0.7977515459060669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24426762759685516, "epoch": 13.02, "learning_rate": 7.3418910457107085e-06, "loss": 0.3865, "step": 15400, "task_loss": 0.39651158452033997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4287922978401184, "epoch": 13.02, "learning_rate": 7.338760175328742e-06, "loss": 0.593, "step": 15401, "task_loss": 1.318279504776001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3147316575050354, "epoch": 13.02, "learning_rate": 7.3356293049467754e-06, "loss": 0.3597, "step": 15402, "task_loss": 0.43380308151245117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4830401539802551, "epoch": 13.02, "learning_rate": 7.332498434564809e-06, "loss": 0.4515, "step": 15403, "task_loss": 0.2616911232471466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4194740951061249, "epoch": 13.02, "learning_rate": 7.329367564182843e-06, "loss": 0.4174, "step": 15404, "task_loss": 0.6082265973091125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16417133808135986, "epoch": 13.02, "learning_rate": 7.326236693800877e-06, "loss": 0.2409, "step": 15405, "task_loss": 0.5197821855545044 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3983297348022461, "epoch": 13.02, "learning_rate": 7.32310582341891e-06, "loss": 0.3956, "step": 15406, "task_loss": 0.9691120386123657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42162472009658813, "epoch": 13.02, "learning_rate": 7.3199749530369445e-06, "loss": 0.5771, "step": 15407, "task_loss": 0.43963342905044556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.303422749042511, "epoch": 13.02, "learning_rate": 7.316844082654978e-06, "loss": 0.5934, "step": 15408, "task_loss": 0.27565649151802063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21294379234313965, "epoch": 13.03, "learning_rate": 7.313713212273013e-06, "loss": 0.3323, "step": 15409, "task_loss": 0.11839675903320312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.282875657081604, "epoch": 13.03, "learning_rate": 7.310582341891047e-06, "loss": 0.4749, "step": 15410, "task_loss": 0.0811990275979042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32419270277023315, "epoch": 13.03, "learning_rate": 7.30745147150908e-06, "loss": 0.4536, "step": 15411, "task_loss": 1.016769528388977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7044914960861206, "epoch": 13.03, "learning_rate": 7.3043206011271136e-06, "loss": 0.4361, "step": 15412, "task_loss": 1.2637015581130981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18543587625026703, "epoch": 13.03, "learning_rate": 7.301189730745148e-06, "loss": 0.353, "step": 15413, "task_loss": 0.40418171882629395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38841694593429565, "epoch": 13.03, "learning_rate": 7.298058860363181e-06, "loss": 0.4416, "step": 15414, "task_loss": 0.5349991321563721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4493221342563629, "epoch": 13.03, "learning_rate": 7.294927989981215e-06, "loss": 0.426, "step": 15415, "task_loss": 0.7516605854034424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3056989312171936, "epoch": 13.03, "learning_rate": 7.291797119599249e-06, "loss": 0.2741, "step": 15416, "task_loss": 0.38950589299201965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49188435077667236, "epoch": 13.03, "learning_rate": 7.288666249217283e-06, "loss": 0.4902, "step": 15417, "task_loss": 1.1986382007598877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3174676299095154, "epoch": 13.03, "learning_rate": 7.285535378835316e-06, "loss": 0.3748, "step": 15418, "task_loss": 0.46183905005455017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2152678370475769, "epoch": 13.03, "learning_rate": 7.28240450845335e-06, "loss": 0.2741, "step": 15419, "task_loss": 0.45740270614624023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2254955768585205, "epoch": 13.03, "learning_rate": 7.279273638071384e-06, "loss": 0.4546, "step": 15420, "task_loss": 0.6414120197296143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3521943688392639, "epoch": 13.04, "learning_rate": 7.276142767689418e-06, "loss": 0.457, "step": 15421, "task_loss": 0.8290771842002869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44183436036109924, "epoch": 13.04, "learning_rate": 7.2730118973074526e-06, "loss": 0.4339, "step": 15422, "task_loss": 0.3301568627357483 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6506335735321045, "epoch": 13.04, "learning_rate": 7.269881026925486e-06, "loss": 0.4698, "step": 15423, "task_loss": 0.9605718851089478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2918151617050171, "epoch": 13.04, "learning_rate": 7.2667501565435195e-06, "loss": 0.3724, "step": 15424, "task_loss": 0.964990496635437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4942999482154846, "epoch": 13.04, "learning_rate": 7.263619286161554e-06, "loss": 0.4299, "step": 15425, "task_loss": 1.023828387260437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3814018964767456, "epoch": 13.04, "learning_rate": 7.260488415779587e-06, "loss": 0.3858, "step": 15426, "task_loss": 0.5225023627281189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3237658441066742, "epoch": 13.04, "learning_rate": 7.257357545397621e-06, "loss": 0.3285, "step": 15427, "task_loss": 0.3306533694267273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5143022537231445, "epoch": 13.04, "learning_rate": 7.254226675015654e-06, "loss": 0.5041, "step": 15428, "task_loss": 0.5311129689216614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4853132367134094, "epoch": 13.04, "learning_rate": 7.251095804633689e-06, "loss": 0.4654, "step": 15429, "task_loss": 0.5726667642593384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.380922794342041, "epoch": 13.04, "learning_rate": 7.247964934251722e-06, "loss": 0.4798, "step": 15430, "task_loss": 0.19770634174346924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25563111901283264, "epoch": 13.04, "learning_rate": 7.2448340638697555e-06, "loss": 0.5077, "step": 15431, "task_loss": 0.1806224286556244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45023709535598755, "epoch": 13.04, "learning_rate": 7.241703193487789e-06, "loss": 0.5302, "step": 15432, "task_loss": 0.5900902152061462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4827898442745209, "epoch": 13.05, "learning_rate": 7.238572323105824e-06, "loss": 0.5323, "step": 15433, "task_loss": 0.1831880360841751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31507810950279236, "epoch": 13.05, "learning_rate": 7.2354414527238585e-06, "loss": 0.3836, "step": 15434, "task_loss": 0.2950150966644287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6125561594963074, "epoch": 13.05, "learning_rate": 7.232310582341892e-06, "loss": 0.4123, "step": 15435, "task_loss": 0.7238138318061829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.372895210981369, "epoch": 13.05, "learning_rate": 7.2291797119599254e-06, "loss": 0.4428, "step": 15436, "task_loss": 0.27551907300949097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48281624913215637, "epoch": 13.05, "learning_rate": 7.226048841577959e-06, "loss": 0.4025, "step": 15437, "task_loss": 0.7428438067436218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5538859963417053, "epoch": 13.05, "learning_rate": 7.222917971195993e-06, "loss": 0.4962, "step": 15438, "task_loss": 0.8222858905792236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31720125675201416, "epoch": 13.05, "learning_rate": 7.219787100814027e-06, "loss": 0.4049, "step": 15439, "task_loss": 0.7433227300643921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2720515727996826, "epoch": 13.05, "learning_rate": 7.21665623043206e-06, "loss": 0.453, "step": 15440, "task_loss": 0.35061541199684143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2612040638923645, "epoch": 13.05, "learning_rate": 7.213525360050094e-06, "loss": 0.4553, "step": 15441, "task_loss": 0.3898374140262604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5389792919158936, "epoch": 13.05, "learning_rate": 7.210394489668128e-06, "loss": 0.4164, "step": 15442, "task_loss": 0.4804874062538147 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4161056876182556, "epoch": 13.05, "learning_rate": 7.2072636192861615e-06, "loss": 0.428, "step": 15443, "task_loss": 0.4979325532913208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5309693217277527, "epoch": 13.05, "learning_rate": 7.204132748904195e-06, "loss": 0.6313, "step": 15444, "task_loss": 0.7215347290039062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5719726085662842, "epoch": 13.06, "learning_rate": 7.201001878522229e-06, "loss": 0.3937, "step": 15445, "task_loss": 0.5455603003501892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3402819037437439, "epoch": 13.06, "learning_rate": 7.197871008140264e-06, "loss": 0.3534, "step": 15446, "task_loss": 0.12233906239271164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4278128743171692, "epoch": 13.06, "learning_rate": 7.194740137758298e-06, "loss": 0.4216, "step": 15447, "task_loss": 0.9273564219474792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6576439142227173, "epoch": 13.06, "learning_rate": 7.191609267376331e-06, "loss": 0.5531, "step": 15448, "task_loss": 0.8660739660263062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6217464804649353, "epoch": 13.06, "learning_rate": 7.188478396994365e-06, "loss": 0.5064, "step": 15449, "task_loss": 0.7387430667877197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4566691517829895, "epoch": 13.06, "learning_rate": 7.185347526612398e-06, "loss": 0.4668, "step": 15450, "task_loss": 0.3609434962272644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40029066801071167, "epoch": 13.06, "learning_rate": 7.182216656230433e-06, "loss": 0.5133, "step": 15451, "task_loss": 0.7454752922058105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3082987368106842, "epoch": 13.06, "learning_rate": 7.179085785848466e-06, "loss": 0.3632, "step": 15452, "task_loss": 0.2474270612001419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3012118935585022, "epoch": 13.06, "learning_rate": 7.1759549154665e-06, "loss": 0.3901, "step": 15453, "task_loss": 0.8273102641105652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8164876103401184, "epoch": 13.06, "learning_rate": 7.172824045084534e-06, "loss": 0.4985, "step": 15454, "task_loss": 1.7312880754470825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43346792459487915, "epoch": 13.06, "learning_rate": 7.169693174702567e-06, "loss": 0.4198, "step": 15455, "task_loss": 1.5442531108856201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44800660014152527, "epoch": 13.07, "learning_rate": 7.166562304320601e-06, "loss": 0.4182, "step": 15456, "task_loss": 0.32091233134269714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41767460107803345, "epoch": 13.07, "learning_rate": 7.163431433938634e-06, "loss": 0.3245, "step": 15457, "task_loss": 0.40960463881492615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34090501070022583, "epoch": 13.07, "learning_rate": 7.1603005635566695e-06, "loss": 0.4365, "step": 15458, "task_loss": 0.08256089687347412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4543449878692627, "epoch": 13.07, "learning_rate": 7.157169693174704e-06, "loss": 0.3534, "step": 15459, "task_loss": 1.2083475589752197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.279386430978775, "epoch": 13.07, "learning_rate": 7.154038822792737e-06, "loss": 0.3019, "step": 15460, "task_loss": 0.6656579971313477 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4244426488876343, "epoch": 13.07, "learning_rate": 7.150907952410771e-06, "loss": 0.4878, "step": 15461, "task_loss": 0.508213222026825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.515209436416626, "epoch": 13.07, "learning_rate": 7.147777082028804e-06, "loss": 0.5124, "step": 15462, "task_loss": 0.6005597114562988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3502766191959381, "epoch": 13.07, "learning_rate": 7.144646211646839e-06, "loss": 0.4527, "step": 15463, "task_loss": 1.1285113096237183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48076075315475464, "epoch": 13.07, "learning_rate": 7.141515341264872e-06, "loss": 0.4456, "step": 15464, "task_loss": 0.9046142101287842 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6511574983596802, "epoch": 13.07, "learning_rate": 7.1383844708829055e-06, "loss": 0.6085, "step": 15465, "task_loss": 0.7923080921173096 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4349067807197571, "epoch": 13.07, "learning_rate": 7.135253600500939e-06, "loss": 0.3752, "step": 15466, "task_loss": 0.9256941080093384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5489201545715332, "epoch": 13.07, "learning_rate": 7.132122730118973e-06, "loss": 0.4888, "step": 15467, "task_loss": 2.1060242652893066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4360518455505371, "epoch": 13.08, "learning_rate": 7.128991859737007e-06, "loss": 0.426, "step": 15468, "task_loss": 0.7704366445541382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37049081921577454, "epoch": 13.08, "learning_rate": 7.12586098935504e-06, "loss": 0.4988, "step": 15469, "task_loss": 0.1803593635559082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49410998821258545, "epoch": 13.08, "learning_rate": 7.1227301189730755e-06, "loss": 0.4347, "step": 15470, "task_loss": 1.0791268348693848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3068748712539673, "epoch": 13.08, "learning_rate": 7.119599248591109e-06, "loss": 0.5696, "step": 15471, "task_loss": 1.2606174945831299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4152895212173462, "epoch": 13.08, "learning_rate": 7.116468378209143e-06, "loss": 0.3397, "step": 15472, "task_loss": 0.5899979472160339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5557246208190918, "epoch": 13.08, "learning_rate": 7.113337507827177e-06, "loss": 0.4097, "step": 15473, "task_loss": 1.8532220125198364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3814261853694916, "epoch": 13.08, "learning_rate": 7.11020663744521e-06, "loss": 0.39, "step": 15474, "task_loss": 0.8736680150032043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5009466409683228, "epoch": 13.08, "learning_rate": 7.107075767063244e-06, "loss": 0.4676, "step": 15475, "task_loss": 0.27681007981300354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36113059520721436, "epoch": 13.08, "learning_rate": 7.103944896681278e-06, "loss": 0.4624, "step": 15476, "task_loss": 1.4055176973342896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2888143062591553, "epoch": 13.08, "learning_rate": 7.1008140262993115e-06, "loss": 0.3188, "step": 15477, "task_loss": 0.9014564752578735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26614344120025635, "epoch": 13.08, "learning_rate": 7.097683155917345e-06, "loss": 0.3541, "step": 15478, "task_loss": 0.33821165561676025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3786030113697052, "epoch": 13.08, "learning_rate": 7.0945522855353784e-06, "loss": 0.4159, "step": 15479, "task_loss": 0.6585286259651184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25429558753967285, "epoch": 13.09, "learning_rate": 7.091421415153413e-06, "loss": 0.411, "step": 15480, "task_loss": 0.5277988910675049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23837848007678986, "epoch": 13.09, "learning_rate": 7.088290544771446e-06, "loss": 0.3283, "step": 15481, "task_loss": 0.9252274632453918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24393779039382935, "epoch": 13.09, "learning_rate": 7.085159674389481e-06, "loss": 0.3758, "step": 15482, "task_loss": 0.39150184392929077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5807412266731262, "epoch": 13.09, "learning_rate": 7.082028804007515e-06, "loss": 0.4605, "step": 15483, "task_loss": 0.7286006212234497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37362098693847656, "epoch": 13.09, "learning_rate": 7.078897933625548e-06, "loss": 0.4187, "step": 15484, "task_loss": 0.6863133907318115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5030677318572998, "epoch": 13.09, "learning_rate": 7.075767063243583e-06, "loss": 0.5743, "step": 15485, "task_loss": 1.0103379487991333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2617281377315521, "epoch": 13.09, "learning_rate": 7.072636192861616e-06, "loss": 0.4304, "step": 15486, "task_loss": 0.13938350975513458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5439741611480713, "epoch": 13.09, "learning_rate": 7.06950532247965e-06, "loss": 0.5192, "step": 15487, "task_loss": 0.7421950101852417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21184472739696503, "epoch": 13.09, "learning_rate": 7.066374452097683e-06, "loss": 0.4218, "step": 15488, "task_loss": 0.2762277126312256 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31970471143722534, "epoch": 13.09, "learning_rate": 7.0632435817157174e-06, "loss": 0.4458, "step": 15489, "task_loss": 1.1469500064849854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41411900520324707, "epoch": 13.09, "learning_rate": 7.060112711333751e-06, "loss": 0.2996, "step": 15490, "task_loss": 0.6333039402961731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.393084853887558, "epoch": 13.09, "learning_rate": 7.056981840951784e-06, "loss": 0.4205, "step": 15491, "task_loss": 0.2219483107328415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43219053745269775, "epoch": 13.1, "learning_rate": 7.053850970569819e-06, "loss": 0.344, "step": 15492, "task_loss": 0.9724631309509277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4531767964363098, "epoch": 13.1, "learning_rate": 7.050720100187852e-06, "loss": 0.5025, "step": 15493, "task_loss": 0.5065772533416748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5843995213508606, "epoch": 13.1, "learning_rate": 7.047589229805886e-06, "loss": 0.4474, "step": 15494, "task_loss": 1.1474802494049072 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23714804649353027, "epoch": 13.1, "learning_rate": 7.044458359423921e-06, "loss": 0.4026, "step": 15495, "task_loss": 0.574388325214386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5762361288070679, "epoch": 13.1, "learning_rate": 7.041327489041954e-06, "loss": 0.4474, "step": 15496, "task_loss": 1.3689069747924805 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3123135566711426, "epoch": 13.1, "learning_rate": 7.038196618659989e-06, "loss": 0.3509, "step": 15497, "task_loss": 0.042127590626478195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4890047609806061, "epoch": 13.1, "learning_rate": 7.035065748278022e-06, "loss": 0.4268, "step": 15498, "task_loss": 0.5350996851921082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35865145921707153, "epoch": 13.1, "learning_rate": 7.0319348778960556e-06, "loss": 0.494, "step": 15499, "task_loss": 0.5785248279571533 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2767668664455414, "epoch": 13.1, "learning_rate": 7.028804007514089e-06, "loss": 0.3209, "step": 15500, "task_loss": 0.08218551427125931 }, { "epoch": 13.1, "eval_accuracy": 0.9123168316831683, "eval_loss": 0.31200969219207764, "eval_runtime": 206.1644, "eval_samples_per_second": 122.475, "eval_steps_per_second": 0.96, "step": 15500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3964611887931824, "epoch": 13.1, "learning_rate": 7.025673137132123e-06, "loss": 0.4051, "step": 15501, "task_loss": 1.5194894075393677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44219228625297546, "epoch": 13.1, "learning_rate": 7.022542266750157e-06, "loss": 0.4172, "step": 15502, "task_loss": 0.7031528353691101 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19514837861061096, "epoch": 13.1, "learning_rate": 7.01941139636819e-06, "loss": 0.3881, "step": 15503, "task_loss": 0.2722989320755005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4002978801727295, "epoch": 13.11, "learning_rate": 7.016280525986224e-06, "loss": 0.3137, "step": 15504, "task_loss": 0.19652700424194336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6123827695846558, "epoch": 13.11, "learning_rate": 7.013149655604258e-06, "loss": 0.4779, "step": 15505, "task_loss": 0.7349413633346558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1900118887424469, "epoch": 13.11, "learning_rate": 7.010018785222292e-06, "loss": 0.3151, "step": 15506, "task_loss": 0.34209030866622925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33799314498901367, "epoch": 13.11, "learning_rate": 7.006887914840327e-06, "loss": 0.4377, "step": 15507, "task_loss": 0.5069386959075928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42248812317848206, "epoch": 13.11, "learning_rate": 7.00375704445836e-06, "loss": 0.4215, "step": 15508, "task_loss": 1.2911802530288696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4389194846153259, "epoch": 13.11, "learning_rate": 7.000626174076394e-06, "loss": 0.4115, "step": 15509, "task_loss": 0.46491295099258423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42878928780555725, "epoch": 13.11, "learning_rate": 6.997495303694428e-06, "loss": 0.3625, "step": 15510, "task_loss": 0.4148316979408264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5441962480545044, "epoch": 13.11, "learning_rate": 6.9943644333124615e-06, "loss": 0.4434, "step": 15511, "task_loss": 0.28407829999923706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5693289041519165, "epoch": 13.11, "learning_rate": 6.991233562930495e-06, "loss": 0.522, "step": 15512, "task_loss": 0.40296733379364014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7205849289894104, "epoch": 13.11, "learning_rate": 6.9881026925485285e-06, "loss": 0.5101, "step": 15513, "task_loss": 0.46947941184043884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3553944528102875, "epoch": 13.11, "learning_rate": 6.984971822166563e-06, "loss": 0.5056, "step": 15514, "task_loss": 0.7115702629089355 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5333380699157715, "epoch": 13.11, "learning_rate": 6.981840951784596e-06, "loss": 0.4428, "step": 15515, "task_loss": 1.0916105508804321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2740231156349182, "epoch": 13.12, "learning_rate": 6.97871008140263e-06, "loss": 0.4645, "step": 15516, "task_loss": 0.47985178232192993 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1747560054063797, "epoch": 13.12, "learning_rate": 6.975579211020663e-06, "loss": 0.366, "step": 15517, "task_loss": 0.44715791940689087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45900759100914, "epoch": 13.12, "learning_rate": 6.9724483406386975e-06, "loss": 0.3753, "step": 15518, "task_loss": 0.18613751232624054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49262112379074097, "epoch": 13.12, "learning_rate": 6.969317470256733e-06, "loss": 0.3881, "step": 15519, "task_loss": 0.7501885294914246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16602128744125366, "epoch": 13.12, "learning_rate": 6.966186599874766e-06, "loss": 0.4902, "step": 15520, "task_loss": 0.6830518245697021 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5080870389938354, "epoch": 13.12, "learning_rate": 6.9630557294928e-06, "loss": 0.5636, "step": 15521, "task_loss": 0.7340821623802185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5023406744003296, "epoch": 13.12, "learning_rate": 6.959924859110833e-06, "loss": 0.4848, "step": 15522, "task_loss": 0.7579521536827087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3604196310043335, "epoch": 13.12, "learning_rate": 6.9567939887288674e-06, "loss": 0.3324, "step": 15523, "task_loss": 1.151607632637024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4461309313774109, "epoch": 13.12, "learning_rate": 6.953663118346901e-06, "loss": 0.3879, "step": 15524, "task_loss": 1.0798609256744385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5391840934753418, "epoch": 13.12, "learning_rate": 6.950532247964934e-06, "loss": 0.4194, "step": 15525, "task_loss": 1.134599208831787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5382905006408691, "epoch": 13.12, "learning_rate": 6.947401377582968e-06, "loss": 0.4466, "step": 15526, "task_loss": 0.6935966610908508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4673299789428711, "epoch": 13.13, "learning_rate": 6.944270507201002e-06, "loss": 0.6139, "step": 15527, "task_loss": 0.7176768183708191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6125930547714233, "epoch": 13.13, "learning_rate": 6.941139636819036e-06, "loss": 0.3884, "step": 15528, "task_loss": 0.9874147176742554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2613111734390259, "epoch": 13.13, "learning_rate": 6.938008766437069e-06, "loss": 0.388, "step": 15529, "task_loss": 0.02552110329270363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5230299830436707, "epoch": 13.13, "learning_rate": 6.9348778960551035e-06, "loss": 0.3583, "step": 15530, "task_loss": 0.533748984336853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.399405837059021, "epoch": 13.13, "learning_rate": 6.931747025673137e-06, "loss": 0.391, "step": 15531, "task_loss": 0.462937593460083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33781781792640686, "epoch": 13.13, "learning_rate": 6.928616155291172e-06, "loss": 0.327, "step": 15532, "task_loss": 0.44713130593299866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3694213628768921, "epoch": 13.13, "learning_rate": 6.925485284909206e-06, "loss": 0.3743, "step": 15533, "task_loss": 0.18917317688465118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44483280181884766, "epoch": 13.13, "learning_rate": 6.922354414527239e-06, "loss": 0.3837, "step": 15534, "task_loss": 0.6093291640281677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5345198512077332, "epoch": 13.13, "learning_rate": 6.919223544145273e-06, "loss": 0.4179, "step": 15535, "task_loss": 0.6213546991348267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5190179944038391, "epoch": 13.13, "learning_rate": 6.916092673763307e-06, "loss": 0.448, "step": 15536, "task_loss": 1.2681903839111328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3503086566925049, "epoch": 13.13, "learning_rate": 6.91296180338134e-06, "loss": 0.5432, "step": 15537, "task_loss": 0.969073474407196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4083949625492096, "epoch": 13.13, "learning_rate": 6.909830932999374e-06, "loss": 0.3993, "step": 15538, "task_loss": 1.028610110282898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4908480644226074, "epoch": 13.14, "learning_rate": 6.906700062617408e-06, "loss": 0.4379, "step": 15539, "task_loss": 1.451982855796814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36608242988586426, "epoch": 13.14, "learning_rate": 6.903569192235442e-06, "loss": 0.4189, "step": 15540, "task_loss": 0.6516835689544678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7072679400444031, "epoch": 13.14, "learning_rate": 6.900438321853475e-06, "loss": 0.4986, "step": 15541, "task_loss": 1.125620722770691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22943168878555298, "epoch": 13.14, "learning_rate": 6.8973074514715086e-06, "loss": 0.4098, "step": 15542, "task_loss": 0.7990049123764038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3815666437149048, "epoch": 13.14, "learning_rate": 6.894176581089543e-06, "loss": 0.4759, "step": 15543, "task_loss": 0.48730364441871643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43371254205703735, "epoch": 13.14, "learning_rate": 6.891045710707578e-06, "loss": 0.3267, "step": 15544, "task_loss": 0.3214567005634308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1592109203338623, "epoch": 13.14, "learning_rate": 6.8879148403256115e-06, "loss": 0.3649, "step": 15545, "task_loss": 0.10820893198251724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.324241578578949, "epoch": 13.14, "learning_rate": 6.884783969943645e-06, "loss": 0.4525, "step": 15546, "task_loss": 0.3651493787765503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38491636514663696, "epoch": 13.14, "learning_rate": 6.8816530995616785e-06, "loss": 0.3993, "step": 15547, "task_loss": 0.7282981276512146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3379245698451996, "epoch": 13.14, "learning_rate": 6.878522229179713e-06, "loss": 0.5479, "step": 15548, "task_loss": 0.47810015082359314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26740020513534546, "epoch": 13.14, "learning_rate": 6.875391358797746e-06, "loss": 0.3542, "step": 15549, "task_loss": 0.686267614364624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3084922432899475, "epoch": 13.14, "learning_rate": 6.87226048841578e-06, "loss": 0.4969, "step": 15550, "task_loss": 1.352903962135315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19211845099925995, "epoch": 13.15, "learning_rate": 6.869129618033813e-06, "loss": 0.5602, "step": 15551, "task_loss": 0.08173803985118866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3830808401107788, "epoch": 13.15, "learning_rate": 6.8659987476518475e-06, "loss": 0.5303, "step": 15552, "task_loss": 0.25520074367523193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5366668701171875, "epoch": 13.15, "learning_rate": 6.862867877269881e-06, "loss": 0.5287, "step": 15553, "task_loss": 0.6350709795951843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.757763147354126, "epoch": 13.15, "learning_rate": 6.8597370068879145e-06, "loss": 0.6021, "step": 15554, "task_loss": 2.1276345252990723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4233860969543457, "epoch": 13.15, "learning_rate": 6.856606136505948e-06, "loss": 0.529, "step": 15555, "task_loss": 0.6057830452919006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42090681195259094, "epoch": 13.15, "learning_rate": 6.853475266123983e-06, "loss": 0.3546, "step": 15556, "task_loss": 0.38565492630004883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.542489767074585, "epoch": 13.15, "learning_rate": 6.8503443957420175e-06, "loss": 0.4424, "step": 15557, "task_loss": 0.7858928442001343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2647233009338379, "epoch": 13.15, "learning_rate": 6.847213525360051e-06, "loss": 0.3597, "step": 15558, "task_loss": 0.22353370487689972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3825925886631012, "epoch": 13.15, "learning_rate": 6.844082654978084e-06, "loss": 0.3895, "step": 15559, "task_loss": 0.3470032811164856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47318676114082336, "epoch": 13.15, "learning_rate": 6.840951784596118e-06, "loss": 0.4641, "step": 15560, "task_loss": 0.37605002522468567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46309465169906616, "epoch": 13.15, "learning_rate": 6.837820914214152e-06, "loss": 0.4512, "step": 15561, "task_loss": 0.6153436899185181 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4497988224029541, "epoch": 13.15, "learning_rate": 6.834690043832186e-06, "loss": 0.4558, "step": 15562, "task_loss": 0.9570232629776001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3817433714866638, "epoch": 13.16, "learning_rate": 6.831559173450219e-06, "loss": 0.4155, "step": 15563, "task_loss": 0.7573544979095459 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20807000994682312, "epoch": 13.16, "learning_rate": 6.828428303068253e-06, "loss": 0.4494, "step": 15564, "task_loss": 0.38539016246795654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46638667583465576, "epoch": 13.16, "learning_rate": 6.825297432686287e-06, "loss": 0.4842, "step": 15565, "task_loss": 0.7535818219184875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3274652361869812, "epoch": 13.16, "learning_rate": 6.8221665623043204e-06, "loss": 0.4938, "step": 15566, "task_loss": 0.5900393128395081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.443392813205719, "epoch": 13.16, "learning_rate": 6.819035691922354e-06, "loss": 0.392, "step": 15567, "task_loss": 0.5349918603897095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4870808720588684, "epoch": 13.16, "learning_rate": 6.815904821540389e-06, "loss": 0.5367, "step": 15568, "task_loss": 0.8998364806175232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44572561979293823, "epoch": 13.16, "learning_rate": 6.8127739511584226e-06, "loss": 0.4276, "step": 15569, "task_loss": 0.729059636592865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4168373942375183, "epoch": 13.16, "learning_rate": 6.809643080776457e-06, "loss": 0.2746, "step": 15570, "task_loss": 0.32395076751708984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44819769263267517, "epoch": 13.16, "learning_rate": 6.80651221039449e-06, "loss": 0.4048, "step": 15571, "task_loss": 0.3261510729789734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5936762094497681, "epoch": 13.16, "learning_rate": 6.803381340012524e-06, "loss": 0.3248, "step": 15572, "task_loss": 0.616587221622467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48232877254486084, "epoch": 13.16, "learning_rate": 6.800250469630558e-06, "loss": 0.45, "step": 15573, "task_loss": 0.6764828562736511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4338821768760681, "epoch": 13.16, "learning_rate": 6.797119599248592e-06, "loss": 0.4607, "step": 15574, "task_loss": 0.5608813166618347 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5385029315948486, "epoch": 13.17, "learning_rate": 6.793988728866625e-06, "loss": 0.5618, "step": 15575, "task_loss": 1.122431993484497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5077014565467834, "epoch": 13.17, "learning_rate": 6.7908578584846586e-06, "loss": 0.4594, "step": 15576, "task_loss": 0.8680570721626282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6359014511108398, "epoch": 13.17, "learning_rate": 6.787726988102693e-06, "loss": 0.5507, "step": 15577, "task_loss": 1.1615824699401855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.625200629234314, "epoch": 13.17, "learning_rate": 6.784596117720726e-06, "loss": 0.5443, "step": 15578, "task_loss": 0.5311485528945923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3465232849121094, "epoch": 13.17, "learning_rate": 6.78146524733876e-06, "loss": 0.5194, "step": 15579, "task_loss": 1.0053082704544067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32505619525909424, "epoch": 13.17, "learning_rate": 6.778334376956793e-06, "loss": 0.395, "step": 15580, "task_loss": 0.8590647578239441 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49222972989082336, "epoch": 13.17, "learning_rate": 6.7752035065748285e-06, "loss": 0.4212, "step": 15581, "task_loss": 1.0679726600646973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2741157114505768, "epoch": 13.17, "learning_rate": 6.772072636192863e-06, "loss": 0.3301, "step": 15582, "task_loss": 0.3010544180870056 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38355395197868347, "epoch": 13.17, "learning_rate": 6.768941765810896e-06, "loss": 0.5611, "step": 15583, "task_loss": 0.19714397192001343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4713965058326721, "epoch": 13.17, "learning_rate": 6.76581089542893e-06, "loss": 0.4752, "step": 15584, "task_loss": 0.3447553217411041 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3603557348251343, "epoch": 13.17, "learning_rate": 6.762680025046963e-06, "loss": 0.3991, "step": 15585, "task_loss": 0.36226123571395874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3586062788963318, "epoch": 13.17, "learning_rate": 6.7595491546649976e-06, "loss": 0.4398, "step": 15586, "task_loss": 0.3363111615180969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3620554506778717, "epoch": 13.18, "learning_rate": 6.756418284283031e-06, "loss": 0.4578, "step": 15587, "task_loss": 0.7583397030830383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5003781914710999, "epoch": 13.18, "learning_rate": 6.7532874139010645e-06, "loss": 0.3694, "step": 15588, "task_loss": 0.846971333026886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5203860998153687, "epoch": 13.18, "learning_rate": 6.750156543519098e-06, "loss": 0.3659, "step": 15589, "task_loss": 0.8154274821281433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7132325172424316, "epoch": 13.18, "learning_rate": 6.747025673137132e-06, "loss": 0.4101, "step": 15590, "task_loss": 0.598547101020813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6433451771736145, "epoch": 13.18, "learning_rate": 6.743894802755166e-06, "loss": 0.5981, "step": 15591, "task_loss": 0.5125296711921692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4669308066368103, "epoch": 13.18, "learning_rate": 6.740763932373199e-06, "loss": 0.4732, "step": 15592, "task_loss": 0.5509132742881775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6546396017074585, "epoch": 13.18, "learning_rate": 6.7376330619912344e-06, "loss": 0.4569, "step": 15593, "task_loss": 0.6501541137695312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37119215726852417, "epoch": 13.18, "learning_rate": 6.734502191609268e-06, "loss": 0.4016, "step": 15594, "task_loss": 0.8799663782119751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39069610834121704, "epoch": 13.18, "learning_rate": 6.731371321227302e-06, "loss": 0.4209, "step": 15595, "task_loss": 0.4295230507850647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4048643112182617, "epoch": 13.18, "learning_rate": 6.728240450845336e-06, "loss": 0.4636, "step": 15596, "task_loss": 0.8896130323410034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41463473439216614, "epoch": 13.18, "learning_rate": 6.725109580463369e-06, "loss": 0.5051, "step": 15597, "task_loss": 0.38382554054260254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5896145105361938, "epoch": 13.19, "learning_rate": 6.721978710081403e-06, "loss": 0.4076, "step": 15598, "task_loss": 0.5780568718910217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38646984100341797, "epoch": 13.19, "learning_rate": 6.718847839699437e-06, "loss": 0.4167, "step": 15599, "task_loss": 0.4968283772468567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46282958984375, "epoch": 13.19, "learning_rate": 6.7157169693174705e-06, "loss": 0.4811, "step": 15600, "task_loss": 0.763498067855835 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3467065691947937, "epoch": 13.19, "learning_rate": 6.712586098935504e-06, "loss": 0.3422, "step": 15601, "task_loss": 0.4870515763759613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.392801433801651, "epoch": 13.19, "learning_rate": 6.709455228553537e-06, "loss": 0.4268, "step": 15602, "task_loss": 0.2001393437385559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30298060178756714, "epoch": 13.19, "learning_rate": 6.706324358171572e-06, "loss": 0.3528, "step": 15603, "task_loss": 0.09451206028461456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5681496858596802, "epoch": 13.19, "learning_rate": 6.703193487789605e-06, "loss": 0.4234, "step": 15604, "task_loss": 0.8339650630950928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30062246322631836, "epoch": 13.19, "learning_rate": 6.70006261740764e-06, "loss": 0.5671, "step": 15605, "task_loss": 0.4624735713005066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3816183805465698, "epoch": 13.19, "learning_rate": 6.696931747025674e-06, "loss": 0.5041, "step": 15606, "task_loss": 0.0796792134642601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5703842639923096, "epoch": 13.19, "learning_rate": 6.693800876643707e-06, "loss": 0.3696, "step": 15607, "task_loss": 0.4984630346298218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4663242995738983, "epoch": 13.19, "learning_rate": 6.690670006261742e-06, "loss": 0.5686, "step": 15608, "task_loss": 0.2963440716266632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3534380793571472, "epoch": 13.19, "learning_rate": 6.687539135879775e-06, "loss": 0.3165, "step": 15609, "task_loss": 0.0159663874655962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42660507559776306, "epoch": 13.2, "learning_rate": 6.684408265497809e-06, "loss": 0.4571, "step": 15610, "task_loss": 0.6219249367713928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.519824206829071, "epoch": 13.2, "learning_rate": 6.681277395115843e-06, "loss": 0.5152, "step": 15611, "task_loss": 0.6746348738670349 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3889700770378113, "epoch": 13.2, "learning_rate": 6.678146524733876e-06, "loss": 0.4898, "step": 15612, "task_loss": 0.5667321085929871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6739393472671509, "epoch": 13.2, "learning_rate": 6.67501565435191e-06, "loss": 0.4286, "step": 15613, "task_loss": 0.8986292481422424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2744065821170807, "epoch": 13.2, "learning_rate": 6.671884783969943e-06, "loss": 0.3901, "step": 15614, "task_loss": 0.39452415704727173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20522627234458923, "epoch": 13.2, "learning_rate": 6.668753913587978e-06, "loss": 0.2752, "step": 15615, "task_loss": 0.2401617020368576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.537897527217865, "epoch": 13.2, "learning_rate": 6.665623043206011e-06, "loss": 0.5741, "step": 15616, "task_loss": 0.6179659366607666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48612773418426514, "epoch": 13.2, "learning_rate": 6.662492172824046e-06, "loss": 0.4629, "step": 15617, "task_loss": 0.1290404498577118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7389791011810303, "epoch": 13.2, "learning_rate": 6.65936130244208e-06, "loss": 0.525, "step": 15618, "task_loss": 1.162497639656067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30664581060409546, "epoch": 13.2, "learning_rate": 6.656230432060113e-06, "loss": 0.5108, "step": 15619, "task_loss": 0.8750694394111633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7766778469085693, "epoch": 13.2, "learning_rate": 6.653099561678148e-06, "loss": 0.4822, "step": 15620, "task_loss": 0.9120073914527893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19965854287147522, "epoch": 13.2, "learning_rate": 6.649968691296181e-06, "loss": 0.2845, "step": 15621, "task_loss": 0.7652977705001831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5141395926475525, "epoch": 13.21, "learning_rate": 6.6468378209142145e-06, "loss": 0.5006, "step": 15622, "task_loss": 1.4484589099884033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3660731613636017, "epoch": 13.21, "learning_rate": 6.643706950532248e-06, "loss": 0.429, "step": 15623, "task_loss": 0.5182780623435974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3842109441757202, "epoch": 13.21, "learning_rate": 6.640576080150282e-06, "loss": 0.4152, "step": 15624, "task_loss": 0.7022645473480225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5092992782592773, "epoch": 13.21, "learning_rate": 6.637445209768316e-06, "loss": 0.5853, "step": 15625, "task_loss": 0.9267306923866272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3675076961517334, "epoch": 13.21, "learning_rate": 6.634314339386349e-06, "loss": 0.4264, "step": 15626, "task_loss": 0.9334899187088013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47779154777526855, "epoch": 13.21, "learning_rate": 6.631183469004383e-06, "loss": 0.3531, "step": 15627, "task_loss": 0.28824886679649353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6095781326293945, "epoch": 13.21, "learning_rate": 6.628052598622417e-06, "loss": 0.4509, "step": 15628, "task_loss": 0.19939595460891724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4197208881378174, "epoch": 13.21, "learning_rate": 6.6249217282404506e-06, "loss": 0.3493, "step": 15629, "task_loss": 0.9499028921127319 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.296069860458374, "epoch": 13.21, "learning_rate": 6.621790857858486e-06, "loss": 0.3936, "step": 15630, "task_loss": 0.35543206334114075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17775243520736694, "epoch": 13.21, "learning_rate": 6.618659987476519e-06, "loss": 0.3711, "step": 15631, "task_loss": 0.2983749210834503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5455096960067749, "epoch": 13.21, "learning_rate": 6.615529117094553e-06, "loss": 0.4148, "step": 15632, "task_loss": 0.5293278098106384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42120903730392456, "epoch": 13.21, "learning_rate": 6.612398246712587e-06, "loss": 0.4719, "step": 15633, "task_loss": 0.34073173999786377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.282362699508667, "epoch": 13.22, "learning_rate": 6.6092673763306205e-06, "loss": 0.3448, "step": 15634, "task_loss": 0.03445171192288399 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3952227234840393, "epoch": 13.22, "learning_rate": 6.606136505948654e-06, "loss": 0.5773, "step": 15635, "task_loss": 1.599684238433838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32834392786026, "epoch": 13.22, "learning_rate": 6.603005635566687e-06, "loss": 0.4803, "step": 15636, "task_loss": 0.38915932178497314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.55783611536026, "epoch": 13.22, "learning_rate": 6.599874765184722e-06, "loss": 0.52, "step": 15637, "task_loss": 0.24286334216594696 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35838884115219116, "epoch": 13.22, "learning_rate": 6.596743894802755e-06, "loss": 0.4695, "step": 15638, "task_loss": 0.633837878704071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32969361543655396, "epoch": 13.22, "learning_rate": 6.593613024420789e-06, "loss": 0.5564, "step": 15639, "task_loss": 0.37689968943595886 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5449967384338379, "epoch": 13.22, "learning_rate": 6.590482154038822e-06, "loss": 0.5912, "step": 15640, "task_loss": 1.185606837272644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6617679595947266, "epoch": 13.22, "learning_rate": 6.5873512836568565e-06, "loss": 0.6591, "step": 15641, "task_loss": 0.7502658367156982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5840612649917603, "epoch": 13.22, "learning_rate": 6.584220413274892e-06, "loss": 0.4395, "step": 15642, "task_loss": 0.3913317322731018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28851306438446045, "epoch": 13.22, "learning_rate": 6.581089542892925e-06, "loss": 0.346, "step": 15643, "task_loss": 0.7804163694381714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33835268020629883, "epoch": 13.22, "learning_rate": 6.577958672510959e-06, "loss": 0.3235, "step": 15644, "task_loss": 0.8991928100585938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5987402200698853, "epoch": 13.22, "learning_rate": 6.574827802128992e-06, "loss": 0.4027, "step": 15645, "task_loss": 0.8769959807395935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16755124926567078, "epoch": 13.23, "learning_rate": 6.571696931747026e-06, "loss": 0.5607, "step": 15646, "task_loss": 0.026051783934235573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4464420676231384, "epoch": 13.23, "learning_rate": 6.56856606136506e-06, "loss": 0.4088, "step": 15647, "task_loss": 0.4719759523868561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6850999593734741, "epoch": 13.23, "learning_rate": 6.565435190983093e-06, "loss": 0.5608, "step": 15648, "task_loss": 0.5511074066162109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5019482374191284, "epoch": 13.23, "learning_rate": 6.562304320601128e-06, "loss": 0.4421, "step": 15649, "task_loss": 0.49948275089263916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5827039480209351, "epoch": 13.23, "learning_rate": 6.559173450219161e-06, "loss": 0.5506, "step": 15650, "task_loss": 0.678712010383606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44340622425079346, "epoch": 13.23, "learning_rate": 6.556042579837195e-06, "loss": 0.4183, "step": 15651, "task_loss": 0.18433092534542084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40435129404067993, "epoch": 13.23, "learning_rate": 6.552911709455228e-06, "loss": 0.4265, "step": 15652, "task_loss": 0.5987377762794495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18238167464733124, "epoch": 13.23, "learning_rate": 6.5497808390732624e-06, "loss": 0.4166, "step": 15653, "task_loss": 0.02566603012382984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46591252088546753, "epoch": 13.23, "learning_rate": 6.546649968691297e-06, "loss": 0.3745, "step": 15654, "task_loss": 0.824142575263977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6744322180747986, "epoch": 13.23, "learning_rate": 6.543519098309331e-06, "loss": 0.541, "step": 15655, "task_loss": 0.7301437258720398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6006245613098145, "epoch": 13.23, "learning_rate": 6.5403882279273646e-06, "loss": 0.4686, "step": 15656, "task_loss": 0.4768677353858948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33800676465034485, "epoch": 13.23, "learning_rate": 6.537257357545398e-06, "loss": 0.4827, "step": 15657, "task_loss": 0.4164397120475769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3068740963935852, "epoch": 13.24, "learning_rate": 6.534126487163432e-06, "loss": 0.6262, "step": 15658, "task_loss": 0.4875120222568512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.382870614528656, "epoch": 13.24, "learning_rate": 6.530995616781466e-06, "loss": 0.3732, "step": 15659, "task_loss": 1.4594700336456299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4283331036567688, "epoch": 13.24, "learning_rate": 6.527864746399499e-06, "loss": 0.4993, "step": 15660, "task_loss": 0.3145095109939575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40635356307029724, "epoch": 13.24, "learning_rate": 6.524733876017533e-06, "loss": 0.5868, "step": 15661, "task_loss": 0.6886839866638184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21530523896217346, "epoch": 13.24, "learning_rate": 6.521603005635567e-06, "loss": 0.3612, "step": 15662, "task_loss": 0.48962804675102234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3535100221633911, "epoch": 13.24, "learning_rate": 6.5184721352536006e-06, "loss": 0.5039, "step": 15663, "task_loss": 0.4348219931125641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41981691122055054, "epoch": 13.24, "learning_rate": 6.515341264871634e-06, "loss": 0.3819, "step": 15664, "task_loss": 0.2980637550354004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40258634090423584, "epoch": 13.24, "learning_rate": 6.5122103944896675e-06, "loss": 0.4333, "step": 15665, "task_loss": 0.8352608680725098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4116130769252777, "epoch": 13.24, "learning_rate": 6.509079524107703e-06, "loss": 0.496, "step": 15666, "task_loss": 0.16036485135555267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47409823536872864, "epoch": 13.24, "learning_rate": 6.505948653725737e-06, "loss": 0.5424, "step": 15667, "task_loss": 1.042904019355774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3296762704849243, "epoch": 13.24, "learning_rate": 6.5028177833437705e-06, "loss": 0.4293, "step": 15668, "task_loss": 0.6177574992179871 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24027083814144135, "epoch": 13.24, "learning_rate": 6.499686912961804e-06, "loss": 0.4512, "step": 15669, "task_loss": 0.65507572889328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32908475399017334, "epoch": 13.25, "learning_rate": 6.4965560425798374e-06, "loss": 0.5665, "step": 15670, "task_loss": 0.7496634721755981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49319037795066833, "epoch": 13.25, "learning_rate": 6.493425172197872e-06, "loss": 0.5001, "step": 15671, "task_loss": 1.8468937873840332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4291069507598877, "epoch": 13.25, "learning_rate": 6.490294301815905e-06, "loss": 0.4503, "step": 15672, "task_loss": 0.6857532262802124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26001784205436707, "epoch": 13.25, "learning_rate": 6.487163431433939e-06, "loss": 0.3904, "step": 15673, "task_loss": 0.24949157238006592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.695191502571106, "epoch": 13.25, "learning_rate": 6.484032561051972e-06, "loss": 0.4553, "step": 15674, "task_loss": 1.3371074199676514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6361416578292847, "epoch": 13.25, "learning_rate": 6.4809016906700065e-06, "loss": 0.4746, "step": 15675, "task_loss": 0.6391172409057617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40681296586990356, "epoch": 13.25, "learning_rate": 6.47777082028804e-06, "loss": 0.4172, "step": 15676, "task_loss": 1.172755479812622 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5611404180526733, "epoch": 13.25, "learning_rate": 6.4746399499060735e-06, "loss": 0.4849, "step": 15677, "task_loss": 0.6437933444976807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5255622863769531, "epoch": 13.25, "learning_rate": 6.471509079524107e-06, "loss": 0.4789, "step": 15678, "task_loss": 1.821777582168579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5430673360824585, "epoch": 13.25, "learning_rate": 6.468378209142142e-06, "loss": 0.3722, "step": 15679, "task_loss": 0.8512192964553833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6027058362960815, "epoch": 13.25, "learning_rate": 6.4652473387601764e-06, "loss": 0.5333, "step": 15680, "task_loss": 0.38274693489074707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3052879273891449, "epoch": 13.26, "learning_rate": 6.46211646837821e-06, "loss": 0.4894, "step": 15681, "task_loss": 0.1533524990081787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3933626413345337, "epoch": 13.26, "learning_rate": 6.458985597996243e-06, "loss": 0.446, "step": 15682, "task_loss": 1.2516058683395386 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3914867043495178, "epoch": 13.26, "learning_rate": 6.455854727614277e-06, "loss": 0.4307, "step": 15683, "task_loss": 0.5937177538871765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5131840705871582, "epoch": 13.26, "learning_rate": 6.452723857232311e-06, "loss": 0.3918, "step": 15684, "task_loss": 0.4944615662097931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7823176383972168, "epoch": 13.26, "learning_rate": 6.449592986850345e-06, "loss": 0.4948, "step": 15685, "task_loss": 0.36362338066101074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.567520022392273, "epoch": 13.26, "learning_rate": 6.446462116468378e-06, "loss": 0.5, "step": 15686, "task_loss": 1.1968003511428833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5780608057975769, "epoch": 13.26, "learning_rate": 6.4433312460864124e-06, "loss": 0.3694, "step": 15687, "task_loss": 0.5421651005744934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3990764617919922, "epoch": 13.26, "learning_rate": 6.440200375704446e-06, "loss": 0.3317, "step": 15688, "task_loss": 0.14318346977233887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5042208433151245, "epoch": 13.26, "learning_rate": 6.437069505322479e-06, "loss": 0.3336, "step": 15689, "task_loss": 0.576438844203949 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40353524684906006, "epoch": 13.26, "learning_rate": 6.433938634940513e-06, "loss": 0.4833, "step": 15690, "task_loss": 0.561560869216919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3403359055519104, "epoch": 13.26, "learning_rate": 6.430807764558548e-06, "loss": 0.4102, "step": 15691, "task_loss": 0.5524148941040039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4272528290748596, "epoch": 13.26, "learning_rate": 6.4276768941765815e-06, "loss": 0.4459, "step": 15692, "task_loss": 0.8486605882644653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6075717210769653, "epoch": 13.27, "learning_rate": 6.424546023794616e-06, "loss": 0.4668, "step": 15693, "task_loss": 1.0654900074005127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41943442821502686, "epoch": 13.27, "learning_rate": 6.421415153412649e-06, "loss": 0.4123, "step": 15694, "task_loss": 0.9504008293151855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5945556163787842, "epoch": 13.27, "learning_rate": 6.418284283030683e-06, "loss": 0.4915, "step": 15695, "task_loss": 0.8987438082695007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45597732067108154, "epoch": 13.27, "learning_rate": 6.415153412648717e-06, "loss": 0.5638, "step": 15696, "task_loss": 0.4967585802078247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3890226483345032, "epoch": 13.27, "learning_rate": 6.412022542266751e-06, "loss": 0.4196, "step": 15697, "task_loss": 0.48184433579444885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28280341625213623, "epoch": 13.27, "learning_rate": 6.408891671884784e-06, "loss": 0.3358, "step": 15698, "task_loss": 0.4088994860649109 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7356595993041992, "epoch": 13.27, "learning_rate": 6.4057608015028175e-06, "loss": 0.5715, "step": 15699, "task_loss": 1.5892834663391113 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4774532914161682, "epoch": 13.27, "learning_rate": 6.402629931120852e-06, "loss": 0.4324, "step": 15700, "task_loss": 0.17891764640808105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6540156602859497, "epoch": 13.27, "learning_rate": 6.399499060738885e-06, "loss": 0.5652, "step": 15701, "task_loss": 0.8521887063980103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3348219692707062, "epoch": 13.27, "learning_rate": 6.396368190356919e-06, "loss": 0.4598, "step": 15702, "task_loss": 0.3301662802696228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4383712708950043, "epoch": 13.27, "learning_rate": 6.393237319974954e-06, "loss": 0.4161, "step": 15703, "task_loss": 1.0438227653503418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4651700258255005, "epoch": 13.27, "learning_rate": 6.3901064495929875e-06, "loss": 0.3982, "step": 15704, "task_loss": 0.5374500155448914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6765433549880981, "epoch": 13.28, "learning_rate": 6.386975579211022e-06, "loss": 0.4629, "step": 15705, "task_loss": 0.47514083981513977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3591102957725525, "epoch": 13.28, "learning_rate": 6.383844708829055e-06, "loss": 0.4189, "step": 15706, "task_loss": 0.66307532787323 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4468070864677429, "epoch": 13.28, "learning_rate": 6.380713838447089e-06, "loss": 0.4991, "step": 15707, "task_loss": 0.3599388599395752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3987544775009155, "epoch": 13.28, "learning_rate": 6.377582968065122e-06, "loss": 0.4222, "step": 15708, "task_loss": 0.4253252148628235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26684850454330444, "epoch": 13.28, "learning_rate": 6.3744520976831565e-06, "loss": 0.4303, "step": 15709, "task_loss": 0.29095685482025146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47852417826652527, "epoch": 13.28, "learning_rate": 6.37132122730119e-06, "loss": 0.3972, "step": 15710, "task_loss": 0.47620779275894165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19671741127967834, "epoch": 13.28, "learning_rate": 6.3681903569192235e-06, "loss": 0.3091, "step": 15711, "task_loss": 0.021356835961341858 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5400789976119995, "epoch": 13.28, "learning_rate": 6.365059486537257e-06, "loss": 0.4537, "step": 15712, "task_loss": 0.9988936185836792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38832205533981323, "epoch": 13.28, "learning_rate": 6.361928616155291e-06, "loss": 0.416, "step": 15713, "task_loss": 0.7801001071929932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2721688151359558, "epoch": 13.28, "learning_rate": 6.358797745773325e-06, "loss": 0.3955, "step": 15714, "task_loss": 0.38653722405433655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5898019075393677, "epoch": 13.28, "learning_rate": 6.355666875391358e-06, "loss": 0.4821, "step": 15715, "task_loss": 0.7357380390167236 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3035760521888733, "epoch": 13.28, "learning_rate": 6.352536005009393e-06, "loss": 0.4909, "step": 15716, "task_loss": 0.10131926834583282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38995957374572754, "epoch": 13.29, "learning_rate": 6.349405134627427e-06, "loss": 0.4018, "step": 15717, "task_loss": 0.11782275885343552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3187776803970337, "epoch": 13.29, "learning_rate": 6.346274264245461e-06, "loss": 0.3547, "step": 15718, "task_loss": 0.545626163482666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5170760154724121, "epoch": 13.29, "learning_rate": 6.343143393863495e-06, "loss": 0.4269, "step": 15719, "task_loss": 1.1172542572021484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2806392312049866, "epoch": 13.29, "learning_rate": 6.340012523481528e-06, "loss": 0.3751, "step": 15720, "task_loss": 0.058948814868927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37786126136779785, "epoch": 13.29, "learning_rate": 6.336881653099562e-06, "loss": 0.4361, "step": 15721, "task_loss": 0.308018296957016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3057498037815094, "epoch": 13.29, "learning_rate": 6.333750782717596e-06, "loss": 0.3894, "step": 15722, "task_loss": 0.055740319192409515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4021526575088501, "epoch": 13.29, "learning_rate": 6.330619912335629e-06, "loss": 0.3997, "step": 15723, "task_loss": 0.587271511554718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.772844672203064, "epoch": 13.29, "learning_rate": 6.327489041953663e-06, "loss": 0.5017, "step": 15724, "task_loss": 1.301624059677124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37876754999160767, "epoch": 13.29, "learning_rate": 6.324358171571697e-06, "loss": 0.555, "step": 15725, "task_loss": 0.8530471324920654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45261484384536743, "epoch": 13.29, "learning_rate": 6.321227301189731e-06, "loss": 0.4187, "step": 15726, "task_loss": 0.23782145977020264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3809325397014618, "epoch": 13.29, "learning_rate": 6.318096430807764e-06, "loss": 0.4227, "step": 15727, "task_loss": 0.22992071509361267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6192077994346619, "epoch": 13.29, "learning_rate": 6.314965560425799e-06, "loss": 0.5067, "step": 15728, "task_loss": 0.6597340703010559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2020624577999115, "epoch": 13.3, "learning_rate": 6.311834690043833e-06, "loss": 0.4078, "step": 15729, "task_loss": 0.5421299338340759 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34443986415863037, "epoch": 13.3, "learning_rate": 6.308703819661866e-06, "loss": 0.4488, "step": 15730, "task_loss": 0.869274914264679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2446242868900299, "epoch": 13.3, "learning_rate": 6.305572949279901e-06, "loss": 0.4282, "step": 15731, "task_loss": 0.5330525636672974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5520766377449036, "epoch": 13.3, "learning_rate": 6.302442078897934e-06, "loss": 0.5158, "step": 15732, "task_loss": 0.7521275877952576 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29579368233680725, "epoch": 13.3, "learning_rate": 6.2993112085159676e-06, "loss": 0.3908, "step": 15733, "task_loss": 0.3702717125415802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5816843509674072, "epoch": 13.3, "learning_rate": 6.296180338134002e-06, "loss": 0.4719, "step": 15734, "task_loss": 1.2337347269058228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5589606165885925, "epoch": 13.3, "learning_rate": 6.293049467752035e-06, "loss": 0.4326, "step": 15735, "task_loss": 0.40867942571640015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4137565791606903, "epoch": 13.3, "learning_rate": 6.289918597370069e-06, "loss": 0.4851, "step": 15736, "task_loss": 0.22056475281715393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27358001470565796, "epoch": 13.3, "learning_rate": 6.286787726988102e-06, "loss": 0.3648, "step": 15737, "task_loss": 0.7320256233215332 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5842796564102173, "epoch": 13.3, "learning_rate": 6.283656856606137e-06, "loss": 0.487, "step": 15738, "task_loss": 0.8353834748268127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5199971795082092, "epoch": 13.3, "learning_rate": 6.28052598622417e-06, "loss": 0.4835, "step": 15739, "task_loss": 0.5571466088294983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2610871195793152, "epoch": 13.3, "learning_rate": 6.277395115842205e-06, "loss": 0.2953, "step": 15740, "task_loss": 0.25401002168655396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2868635654449463, "epoch": 13.31, "learning_rate": 6.274264245460239e-06, "loss": 0.4479, "step": 15741, "task_loss": 0.5841489434242249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2939032018184662, "epoch": 13.31, "learning_rate": 6.271133375078272e-06, "loss": 0.4563, "step": 15742, "task_loss": 0.37598899006843567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3680171072483063, "epoch": 13.31, "learning_rate": 6.2680025046963065e-06, "loss": 0.4873, "step": 15743, "task_loss": 0.28088775277137756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5314181447029114, "epoch": 13.31, "learning_rate": 6.26487163431434e-06, "loss": 0.5546, "step": 15744, "task_loss": 0.267477810382843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5938636064529419, "epoch": 13.31, "learning_rate": 6.2617407639323735e-06, "loss": 0.4006, "step": 15745, "task_loss": 0.9907795190811157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39066022634506226, "epoch": 13.31, "learning_rate": 6.258609893550407e-06, "loss": 0.4891, "step": 15746, "task_loss": 0.5147544145584106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3462901711463928, "epoch": 13.31, "learning_rate": 6.255479023168441e-06, "loss": 0.383, "step": 15747, "task_loss": 0.2955489158630371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43142348527908325, "epoch": 13.31, "learning_rate": 6.252348152786475e-06, "loss": 0.6173, "step": 15748, "task_loss": 0.8246617913246155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4296700358390808, "epoch": 13.31, "learning_rate": 6.249217282404509e-06, "loss": 0.3802, "step": 15749, "task_loss": 0.49814754724502563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4068092405796051, "epoch": 13.31, "learning_rate": 6.2460864120225426e-06, "loss": 0.4431, "step": 15750, "task_loss": 0.3339289426803589 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45037373900413513, "epoch": 13.31, "learning_rate": 6.242955541640577e-06, "loss": 0.5131, "step": 15751, "task_loss": 1.4463261365890503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27705517411231995, "epoch": 13.32, "learning_rate": 6.23982467125861e-06, "loss": 0.4315, "step": 15752, "task_loss": 0.08108366280794144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9650028347969055, "epoch": 13.32, "learning_rate": 6.236693800876644e-06, "loss": 0.604, "step": 15753, "task_loss": 1.834717869758606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27609747648239136, "epoch": 13.32, "learning_rate": 6.233562930494677e-06, "loss": 0.3933, "step": 15754, "task_loss": 0.10844257473945618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31596487760543823, "epoch": 13.32, "learning_rate": 6.230432060112712e-06, "loss": 0.5061, "step": 15755, "task_loss": 0.44224733114242554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2546076774597168, "epoch": 13.32, "learning_rate": 6.227301189730746e-06, "loss": 0.473, "step": 15756, "task_loss": 0.724402129650116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35161545872688293, "epoch": 13.32, "learning_rate": 6.2241703193487794e-06, "loss": 0.4101, "step": 15757, "task_loss": 0.562107264995575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33338624238967896, "epoch": 13.32, "learning_rate": 6.221039448966813e-06, "loss": 0.3674, "step": 15758, "task_loss": 0.5329762697219849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3256211578845978, "epoch": 13.32, "learning_rate": 6.217908578584846e-06, "loss": 0.4124, "step": 15759, "task_loss": 0.4877862334251404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.304684042930603, "epoch": 13.32, "learning_rate": 6.214777708202881e-06, "loss": 0.3909, "step": 15760, "task_loss": 0.3308192789554596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.454011470079422, "epoch": 13.32, "learning_rate": 6.211646837820914e-06, "loss": 0.4526, "step": 15761, "task_loss": 0.37978702783584595 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8172845840454102, "epoch": 13.32, "learning_rate": 6.2085159674389485e-06, "loss": 0.4299, "step": 15762, "task_loss": 0.9094002842903137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3455427587032318, "epoch": 13.32, "learning_rate": 6.205385097056982e-06, "loss": 0.4363, "step": 15763, "task_loss": 0.29429084062576294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3334290385246277, "epoch": 13.33, "learning_rate": 6.202254226675016e-06, "loss": 0.4965, "step": 15764, "task_loss": 1.1051393747329712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3827548623085022, "epoch": 13.33, "learning_rate": 6.19912335629305e-06, "loss": 0.3653, "step": 15765, "task_loss": 0.16592808067798615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3540760278701782, "epoch": 13.33, "learning_rate": 6.195992485911083e-06, "loss": 0.4654, "step": 15766, "task_loss": 0.3239298164844513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.448323518037796, "epoch": 13.33, "learning_rate": 6.192861615529117e-06, "loss": 0.5373, "step": 15767, "task_loss": 0.7156351804733276 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3555176258087158, "epoch": 13.33, "learning_rate": 6.189730745147151e-06, "loss": 0.5108, "step": 15768, "task_loss": 0.42365920543670654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.643278956413269, "epoch": 13.33, "learning_rate": 6.186599874765185e-06, "loss": 0.4594, "step": 15769, "task_loss": 1.206586480140686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33331507444381714, "epoch": 13.33, "learning_rate": 6.183469004383219e-06, "loss": 0.3684, "step": 15770, "task_loss": 0.4127902686595917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3596264123916626, "epoch": 13.33, "learning_rate": 6.180338134001252e-06, "loss": 0.4249, "step": 15771, "task_loss": 0.8565923571586609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46395111083984375, "epoch": 13.33, "learning_rate": 6.177207263619287e-06, "loss": 0.4587, "step": 15772, "task_loss": 0.6925610303878784 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6244601011276245, "epoch": 13.33, "learning_rate": 6.17407639323732e-06, "loss": 0.5684, "step": 15773, "task_loss": 0.441911518573761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42718306183815, "epoch": 13.33, "learning_rate": 6.1709455228553544e-06, "loss": 0.5026, "step": 15774, "task_loss": 0.183220773935318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4553989768028259, "epoch": 13.33, "learning_rate": 6.167814652473388e-06, "loss": 0.4188, "step": 15775, "task_loss": 0.7744249105453491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44625556468963623, "epoch": 13.34, "learning_rate": 6.164683782091421e-06, "loss": 0.4659, "step": 15776, "task_loss": 1.0840792655944824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41786161065101624, "epoch": 13.34, "learning_rate": 6.161552911709456e-06, "loss": 0.5778, "step": 15777, "task_loss": 0.6549808382987976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20497415959835052, "epoch": 13.34, "learning_rate": 6.158422041327489e-06, "loss": 0.3889, "step": 15778, "task_loss": 0.3197612464427948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38486146926879883, "epoch": 13.34, "learning_rate": 6.155291170945523e-06, "loss": 0.3913, "step": 15779, "task_loss": 1.123551368713379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5766889452934265, "epoch": 13.34, "learning_rate": 6.152160300563557e-06, "loss": 0.5736, "step": 15780, "task_loss": 0.583447277545929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4823257625102997, "epoch": 13.34, "learning_rate": 6.149029430181591e-06, "loss": 0.5729, "step": 15781, "task_loss": 0.39103251695632935 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3727438151836395, "epoch": 13.34, "learning_rate": 6.145898559799625e-06, "loss": 0.4706, "step": 15782, "task_loss": 0.8827834129333496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2821478247642517, "epoch": 13.34, "learning_rate": 6.142767689417658e-06, "loss": 0.3655, "step": 15783, "task_loss": 0.6155612468719482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5025248527526855, "epoch": 13.34, "learning_rate": 6.139636819035692e-06, "loss": 0.4648, "step": 15784, "task_loss": 0.2929396629333496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44216886162757874, "epoch": 13.34, "learning_rate": 6.136505948653726e-06, "loss": 0.4387, "step": 15785, "task_loss": 0.20962442457675934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42619335651397705, "epoch": 13.34, "learning_rate": 6.13337507827176e-06, "loss": 0.4873, "step": 15786, "task_loss": 1.1981046199798584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2502480149269104, "epoch": 13.34, "learning_rate": 6.130244207889794e-06, "loss": 0.485, "step": 15787, "task_loss": 0.3547716736793518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.546096682548523, "epoch": 13.35, "learning_rate": 6.127113337507827e-06, "loss": 0.4839, "step": 15788, "task_loss": 1.3133419752120972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8590391874313354, "epoch": 13.35, "learning_rate": 6.123982467125862e-06, "loss": 0.4715, "step": 15789, "task_loss": 0.6599987149238586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6191538572311401, "epoch": 13.35, "learning_rate": 6.120851596743895e-06, "loss": 0.4947, "step": 15790, "task_loss": 0.8138296008110046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4619519114494324, "epoch": 13.35, "learning_rate": 6.117720726361929e-06, "loss": 0.3611, "step": 15791, "task_loss": 0.4124516546726227 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2634199261665344, "epoch": 13.35, "learning_rate": 6.114589855979963e-06, "loss": 0.3269, "step": 15792, "task_loss": 0.1803615540266037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2851288616657257, "epoch": 13.35, "learning_rate": 6.111458985597996e-06, "loss": 0.347, "step": 15793, "task_loss": 0.5848250389099121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.469954252243042, "epoch": 13.35, "learning_rate": 6.108328115216031e-06, "loss": 0.3435, "step": 15794, "task_loss": 0.533743143081665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.544838547706604, "epoch": 13.35, "learning_rate": 6.105197244834064e-06, "loss": 0.4145, "step": 15795, "task_loss": 0.12826567888259888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5625670552253723, "epoch": 13.35, "learning_rate": 6.102066374452098e-06, "loss": 0.464, "step": 15796, "task_loss": 0.465681254863739 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44766125082969666, "epoch": 13.35, "learning_rate": 6.098935504070131e-06, "loss": 0.4881, "step": 15797, "task_loss": 0.7040338516235352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29820677638053894, "epoch": 13.35, "learning_rate": 6.095804633688166e-06, "loss": 0.4467, "step": 15798, "task_loss": 0.32916468381881714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29405850172042847, "epoch": 13.35, "learning_rate": 6.0926737633062e-06, "loss": 0.3741, "step": 15799, "task_loss": 0.46333277225494385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.469359815120697, "epoch": 13.36, "learning_rate": 6.089542892924233e-06, "loss": 0.411, "step": 15800, "task_loss": 0.18820123374462128 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6106359958648682, "epoch": 13.36, "learning_rate": 6.086412022542267e-06, "loss": 0.3799, "step": 15801, "task_loss": 0.6667690277099609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42345112562179565, "epoch": 13.36, "learning_rate": 6.083281152160301e-06, "loss": 0.4472, "step": 15802, "task_loss": 1.2425158023834229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2502135634422302, "epoch": 13.36, "learning_rate": 6.0801502817783345e-06, "loss": 0.3928, "step": 15803, "task_loss": 0.08863341808319092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46046844124794006, "epoch": 13.36, "learning_rate": 6.077019411396369e-06, "loss": 0.4928, "step": 15804, "task_loss": 0.3421168625354767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4856645464897156, "epoch": 13.36, "learning_rate": 6.073888541014402e-06, "loss": 0.4083, "step": 15805, "task_loss": 0.7376821637153625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3863556981086731, "epoch": 13.36, "learning_rate": 6.070757670632436e-06, "loss": 0.419, "step": 15806, "task_loss": 0.4155901372432709 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41346055269241333, "epoch": 13.36, "learning_rate": 6.06762680025047e-06, "loss": 0.423, "step": 15807, "task_loss": 0.48400598764419556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5289228558540344, "epoch": 13.36, "learning_rate": 6.064495929868504e-06, "loss": 0.4615, "step": 15808, "task_loss": 0.9393970966339111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22366881370544434, "epoch": 13.36, "learning_rate": 6.061365059486537e-06, "loss": 0.3167, "step": 15809, "task_loss": 0.1216389462351799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3726382851600647, "epoch": 13.36, "learning_rate": 6.058234189104571e-06, "loss": 0.5206, "step": 15810, "task_loss": 1.1504931449890137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41248783469200134, "epoch": 13.36, "learning_rate": 6.055103318722606e-06, "loss": 0.3937, "step": 15811, "task_loss": 0.7236272096633911 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2566641569137573, "epoch": 13.37, "learning_rate": 6.051972448340639e-06, "loss": 0.3888, "step": 15812, "task_loss": 0.46169593930244446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24845390021800995, "epoch": 13.37, "learning_rate": 6.048841577958673e-06, "loss": 0.4153, "step": 15813, "task_loss": 0.3659707009792328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41013389825820923, "epoch": 13.37, "learning_rate": 6.045710707576706e-06, "loss": 0.5104, "step": 15814, "task_loss": 0.6254177093505859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5090680122375488, "epoch": 13.37, "learning_rate": 6.0425798371947405e-06, "loss": 0.4216, "step": 15815, "task_loss": 0.43950188159942627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4134365916252136, "epoch": 13.37, "learning_rate": 6.039448966812774e-06, "loss": 0.3081, "step": 15816, "task_loss": 0.2983061969280243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5817660093307495, "epoch": 13.37, "learning_rate": 6.036318096430808e-06, "loss": 0.5124, "step": 15817, "task_loss": 0.38958024978637695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40210938453674316, "epoch": 13.37, "learning_rate": 6.033187226048842e-06, "loss": 0.481, "step": 15818, "task_loss": 1.1073012351989746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.308972030878067, "epoch": 13.37, "learning_rate": 6.030056355666876e-06, "loss": 0.4785, "step": 15819, "task_loss": 0.17572645843029022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3734644651412964, "epoch": 13.37, "learning_rate": 6.0269254852849096e-06, "loss": 0.4819, "step": 15820, "task_loss": 0.34063491225242615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5829426050186157, "epoch": 13.37, "learning_rate": 6.023794614902943e-06, "loss": 0.3985, "step": 15821, "task_loss": 0.7905882000923157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2530302405357361, "epoch": 13.37, "learning_rate": 6.0206637445209765e-06, "loss": 0.4125, "step": 15822, "task_loss": 0.42546749114990234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4741145372390747, "epoch": 13.38, "learning_rate": 6.017532874139011e-06, "loss": 0.5194, "step": 15823, "task_loss": 0.9742822051048279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5338072776794434, "epoch": 13.38, "learning_rate": 6.014402003757045e-06, "loss": 0.5002, "step": 15824, "task_loss": 0.5662471055984497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2427530735731125, "epoch": 13.38, "learning_rate": 6.011271133375079e-06, "loss": 0.3579, "step": 15825, "task_loss": 0.3187355101108551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8757758140563965, "epoch": 13.38, "learning_rate": 6.008140262993112e-06, "loss": 0.5879, "step": 15826, "task_loss": 1.2668746709823608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39983826875686646, "epoch": 13.38, "learning_rate": 6.005009392611146e-06, "loss": 0.4895, "step": 15827, "task_loss": 0.9755305647850037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6021150350570679, "epoch": 13.38, "learning_rate": 6.00187852222918e-06, "loss": 0.3658, "step": 15828, "task_loss": 0.1997978836297989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9609103202819824, "epoch": 13.38, "learning_rate": 5.998747651847214e-06, "loss": 0.5577, "step": 15829, "task_loss": 1.3065621852874756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41147905588150024, "epoch": 13.38, "learning_rate": 5.995616781465248e-06, "loss": 0.4342, "step": 15830, "task_loss": 0.5156720876693726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4343652129173279, "epoch": 13.38, "learning_rate": 5.992485911083281e-06, "loss": 0.4611, "step": 15831, "task_loss": 0.7074016332626343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3060655891895294, "epoch": 13.38, "learning_rate": 5.9893550407013155e-06, "loss": 0.4381, "step": 15832, "task_loss": 0.05952662229537964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37201428413391113, "epoch": 13.38, "learning_rate": 5.986224170319349e-06, "loss": 0.3953, "step": 15833, "task_loss": 0.16968558728694916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29733791947364807, "epoch": 13.38, "learning_rate": 5.9830932999373824e-06, "loss": 0.3709, "step": 15834, "task_loss": 0.26684731245040894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26865237951278687, "epoch": 13.39, "learning_rate": 5.979962429555417e-06, "loss": 0.3145, "step": 15835, "task_loss": 0.058847054839134216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3355289697647095, "epoch": 13.39, "learning_rate": 5.976831559173451e-06, "loss": 0.4778, "step": 15836, "task_loss": 0.48457643389701843 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4865332841873169, "epoch": 13.39, "learning_rate": 5.9737006887914846e-06, "loss": 0.4234, "step": 15837, "task_loss": 1.178786277770996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39160996675491333, "epoch": 13.39, "learning_rate": 5.970569818409518e-06, "loss": 0.496, "step": 15838, "task_loss": 0.7097964286804199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3875848054885864, "epoch": 13.39, "learning_rate": 5.9674389480275515e-06, "loss": 0.3703, "step": 15839, "task_loss": 0.49873045086860657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46705615520477295, "epoch": 13.39, "learning_rate": 5.964308077645586e-06, "loss": 0.3599, "step": 15840, "task_loss": 0.7315900921821594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36896640062332153, "epoch": 13.39, "learning_rate": 5.96117720726362e-06, "loss": 0.4506, "step": 15841, "task_loss": 0.5024535059928894 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3694075644016266, "epoch": 13.39, "learning_rate": 5.958046336881654e-06, "loss": 0.5071, "step": 15842, "task_loss": 0.5632524490356445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32899045944213867, "epoch": 13.39, "learning_rate": 5.954915466499687e-06, "loss": 0.48, "step": 15843, "task_loss": 0.42378851771354675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3185833692550659, "epoch": 13.39, "learning_rate": 5.951784596117721e-06, "loss": 0.3712, "step": 15844, "task_loss": 0.14870263636112213 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23294542729854584, "epoch": 13.39, "learning_rate": 5.948653725735755e-06, "loss": 0.3569, "step": 15845, "task_loss": 0.3174562454223633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5194379091262817, "epoch": 13.39, "learning_rate": 5.945522855353788e-06, "loss": 0.6127, "step": 15846, "task_loss": 0.8854750990867615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4283796548843384, "epoch": 13.4, "learning_rate": 5.942391984971823e-06, "loss": 0.3966, "step": 15847, "task_loss": 0.3648597002029419 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5169835686683655, "epoch": 13.4, "learning_rate": 5.939261114589856e-06, "loss": 0.5018, "step": 15848, "task_loss": 0.35221850872039795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5153524279594421, "epoch": 13.4, "learning_rate": 5.9361302442078905e-06, "loss": 0.5199, "step": 15849, "task_loss": 0.4033229649066925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4983293414115906, "epoch": 13.4, "learning_rate": 5.932999373825924e-06, "loss": 0.4667, "step": 15850, "task_loss": 0.8897570967674255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3272625803947449, "epoch": 13.4, "learning_rate": 5.9298685034439575e-06, "loss": 0.3085, "step": 15851, "task_loss": 0.7025165557861328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4263741374015808, "epoch": 13.4, "learning_rate": 5.926737633061991e-06, "loss": 0.3968, "step": 15852, "task_loss": 0.4992080330848694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6625022888183594, "epoch": 13.4, "learning_rate": 5.923606762680025e-06, "loss": 0.4138, "step": 15853, "task_loss": 0.7112157940864563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33637261390686035, "epoch": 13.4, "learning_rate": 5.9204758922980596e-06, "loss": 0.4884, "step": 15854, "task_loss": 0.3007996678352356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4038006365299225, "epoch": 13.4, "learning_rate": 5.917345021916093e-06, "loss": 0.4298, "step": 15855, "task_loss": 1.1186448335647583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2709214687347412, "epoch": 13.4, "learning_rate": 5.9142141515341265e-06, "loss": 0.3201, "step": 15856, "task_loss": 0.1860659420490265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39502447843551636, "epoch": 13.4, "learning_rate": 5.911083281152161e-06, "loss": 0.4553, "step": 15857, "task_loss": 0.25085335969924927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40580281615257263, "epoch": 13.4, "learning_rate": 5.907952410770194e-06, "loss": 0.4006, "step": 15858, "task_loss": 0.9202547669410706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42234694957733154, "epoch": 13.41, "learning_rate": 5.904821540388228e-06, "loss": 0.4244, "step": 15859, "task_loss": 0.6863624453544617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6253125667572021, "epoch": 13.41, "learning_rate": 5.901690670006262e-06, "loss": 0.458, "step": 15860, "task_loss": 0.3910221755504608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3627806603908539, "epoch": 13.41, "learning_rate": 5.898559799624296e-06, "loss": 0.301, "step": 15861, "task_loss": 0.39458009600639343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49819231033325195, "epoch": 13.41, "learning_rate": 5.89542892924233e-06, "loss": 0.3867, "step": 15862, "task_loss": 0.5528027415275574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2778834402561188, "epoch": 13.41, "learning_rate": 5.892298058860363e-06, "loss": 0.3545, "step": 15863, "task_loss": 0.6147720217704773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.268973708152771, "epoch": 13.41, "learning_rate": 5.889167188478397e-06, "loss": 0.438, "step": 15864, "task_loss": 0.20359231531620026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6246383190155029, "epoch": 13.41, "learning_rate": 5.886036318096431e-06, "loss": 0.46, "step": 15865, "task_loss": 0.9312067627906799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4089069068431854, "epoch": 13.41, "learning_rate": 5.8829054477144655e-06, "loss": 0.4997, "step": 15866, "task_loss": 1.2084448337554932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46113020181655884, "epoch": 13.41, "learning_rate": 5.879774577332499e-06, "loss": 0.4502, "step": 15867, "task_loss": 1.9132471084594727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.194707989692688, "epoch": 13.41, "learning_rate": 5.8766437069505325e-06, "loss": 0.3196, "step": 15868, "task_loss": 0.3341147005558014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21020391583442688, "epoch": 13.41, "learning_rate": 5.873512836568566e-06, "loss": 0.3767, "step": 15869, "task_loss": 0.05408085137605667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21600373089313507, "epoch": 13.41, "learning_rate": 5.8703819661866e-06, "loss": 0.2261, "step": 15870, "task_loss": 0.4280610680580139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5513089299201965, "epoch": 13.42, "learning_rate": 5.867251095804634e-06, "loss": 0.4816, "step": 15871, "task_loss": 0.23537693917751312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49374666810035706, "epoch": 13.42, "learning_rate": 5.864120225422668e-06, "loss": 0.4486, "step": 15872, "task_loss": 0.3796530067920685 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3206483721733093, "epoch": 13.42, "learning_rate": 5.8609893550407015e-06, "loss": 0.4368, "step": 15873, "task_loss": 1.6495733261108398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7015609741210938, "epoch": 13.42, "learning_rate": 5.857858484658736e-06, "loss": 0.4743, "step": 15874, "task_loss": 0.30716755986213684 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5797209739685059, "epoch": 13.42, "learning_rate": 5.854727614276769e-06, "loss": 0.4762, "step": 15875, "task_loss": 0.10908312350511551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6153395175933838, "epoch": 13.42, "learning_rate": 5.851596743894803e-06, "loss": 0.4396, "step": 15876, "task_loss": 0.421658456325531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38648223876953125, "epoch": 13.42, "learning_rate": 5.848465873512836e-06, "loss": 0.3528, "step": 15877, "task_loss": 0.3646940290927887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4391847252845764, "epoch": 13.42, "learning_rate": 5.845335003130871e-06, "loss": 0.4598, "step": 15878, "task_loss": 0.5234953761100769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4352898597717285, "epoch": 13.42, "learning_rate": 5.842204132748905e-06, "loss": 0.3135, "step": 15879, "task_loss": 1.1374335289001465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4554044008255005, "epoch": 13.42, "learning_rate": 5.839073262366938e-06, "loss": 0.363, "step": 15880, "task_loss": 0.15371882915496826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45816296339035034, "epoch": 13.42, "learning_rate": 5.835942391984972e-06, "loss": 0.5106, "step": 15881, "task_loss": 0.7609803080558777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3749489188194275, "epoch": 13.42, "learning_rate": 5.832811521603005e-06, "loss": 0.3471, "step": 15882, "task_loss": 0.4382195472717285 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5171781182289124, "epoch": 13.43, "learning_rate": 5.82968065122104e-06, "loss": 0.5668, "step": 15883, "task_loss": 0.43895241618156433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3198450207710266, "epoch": 13.43, "learning_rate": 5.826549780839074e-06, "loss": 0.3766, "step": 15884, "task_loss": 0.29119059443473816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42983120679855347, "epoch": 13.43, "learning_rate": 5.8234189104571075e-06, "loss": 0.4453, "step": 15885, "task_loss": 0.7426994442939758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46943071484565735, "epoch": 13.43, "learning_rate": 5.820288040075141e-06, "loss": 0.3871, "step": 15886, "task_loss": 0.40380990505218506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40995487570762634, "epoch": 13.43, "learning_rate": 5.817157169693175e-06, "loss": 0.3985, "step": 15887, "task_loss": 1.1108518838882446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19524957239627838, "epoch": 13.43, "learning_rate": 5.814026299311209e-06, "loss": 0.3535, "step": 15888, "task_loss": 0.7192867398262024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4684205651283264, "epoch": 13.43, "learning_rate": 5.810895428929242e-06, "loss": 0.35, "step": 15889, "task_loss": 0.7574050426483154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8652693629264832, "epoch": 13.43, "learning_rate": 5.8077645585472765e-06, "loss": 0.74, "step": 15890, "task_loss": 1.5642541646957397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5885051488876343, "epoch": 13.43, "learning_rate": 5.804633688165311e-06, "loss": 0.5557, "step": 15891, "task_loss": 0.8221715688705444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5318166017532349, "epoch": 13.43, "learning_rate": 5.801502817783344e-06, "loss": 0.539, "step": 15892, "task_loss": 0.9656890630722046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36023008823394775, "epoch": 13.43, "learning_rate": 5.798371947401378e-06, "loss": 0.3605, "step": 15893, "task_loss": 0.2827311158180237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5281606912612915, "epoch": 13.44, "learning_rate": 5.795241077019411e-06, "loss": 0.5901, "step": 15894, "task_loss": 1.0547337532043457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3888595402240753, "epoch": 13.44, "learning_rate": 5.792110206637446e-06, "loss": 0.4162, "step": 15895, "task_loss": 0.1781640499830246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36913061141967773, "epoch": 13.44, "learning_rate": 5.78897933625548e-06, "loss": 0.4629, "step": 15896, "task_loss": 0.3384966552257538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31140366196632385, "epoch": 13.44, "learning_rate": 5.785848465873513e-06, "loss": 0.3897, "step": 15897, "task_loss": 0.2972017824649811 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6822065114974976, "epoch": 13.44, "learning_rate": 5.782717595491547e-06, "loss": 0.5159, "step": 15898, "task_loss": 0.6189518570899963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39749157428741455, "epoch": 13.44, "learning_rate": 5.77958672510958e-06, "loss": 0.3348, "step": 15899, "task_loss": 1.8016302585601807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.426714688539505, "epoch": 13.44, "learning_rate": 5.776455854727615e-06, "loss": 0.4748, "step": 15900, "task_loss": 0.5107586979866028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44843000173568726, "epoch": 13.44, "learning_rate": 5.773324984345648e-06, "loss": 0.3607, "step": 15901, "task_loss": 0.7449435591697693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7594591379165649, "epoch": 13.44, "learning_rate": 5.770194113963682e-06, "loss": 0.4785, "step": 15902, "task_loss": 0.4240480065345764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5649664402008057, "epoch": 13.44, "learning_rate": 5.767063243581716e-06, "loss": 0.4513, "step": 15903, "task_loss": 0.4839896857738495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2755237817764282, "epoch": 13.44, "learning_rate": 5.76393237319975e-06, "loss": 0.4217, "step": 15904, "task_loss": 0.5588721632957458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35128113627433777, "epoch": 13.44, "learning_rate": 5.760801502817784e-06, "loss": 0.4351, "step": 15905, "task_loss": 0.6146083474159241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6273253560066223, "epoch": 13.45, "learning_rate": 5.757670632435817e-06, "loss": 0.4962, "step": 15906, "task_loss": 1.2364368438720703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27345430850982666, "epoch": 13.45, "learning_rate": 5.754539762053851e-06, "loss": 0.3049, "step": 15907, "task_loss": 0.4587266147136688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20819662511348724, "epoch": 13.45, "learning_rate": 5.751408891671885e-06, "loss": 0.2995, "step": 15908, "task_loss": 0.029417166486382484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27429696917533875, "epoch": 13.45, "learning_rate": 5.748278021289919e-06, "loss": 0.3416, "step": 15909, "task_loss": 0.5042890906333923 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42059314250946045, "epoch": 13.45, "learning_rate": 5.745147150907953e-06, "loss": 0.3863, "step": 15910, "task_loss": 0.2465226650238037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33923712372779846, "epoch": 13.45, "learning_rate": 5.742016280525986e-06, "loss": 0.3392, "step": 15911, "task_loss": 0.32885023951530457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3776000142097473, "epoch": 13.45, "learning_rate": 5.738885410144021e-06, "loss": 0.4284, "step": 15912, "task_loss": 0.38084444403648376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35959962010383606, "epoch": 13.45, "learning_rate": 5.735754539762054e-06, "loss": 0.3484, "step": 15913, "task_loss": 0.8179718255996704 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2414703220129013, "epoch": 13.45, "learning_rate": 5.7326236693800876e-06, "loss": 0.4408, "step": 15914, "task_loss": 0.4633232653141022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5637484192848206, "epoch": 13.45, "learning_rate": 5.729492798998122e-06, "loss": 0.4073, "step": 15915, "task_loss": 0.8478339910507202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21066881716251373, "epoch": 13.45, "learning_rate": 5.726361928616155e-06, "loss": 0.4437, "step": 15916, "task_loss": 0.4945376515388489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3578612208366394, "epoch": 13.45, "learning_rate": 5.72323105823419e-06, "loss": 0.4562, "step": 15917, "task_loss": 1.0334341526031494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35818931460380554, "epoch": 13.46, "learning_rate": 5.720100187852223e-06, "loss": 0.3855, "step": 15918, "task_loss": 0.6400597095489502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33093202114105225, "epoch": 13.46, "learning_rate": 5.716969317470257e-06, "loss": 0.3317, "step": 15919, "task_loss": 0.2908251881599426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4372285306453705, "epoch": 13.46, "learning_rate": 5.71383844708829e-06, "loss": 0.6957, "step": 15920, "task_loss": 0.6229963302612305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3610823154449463, "epoch": 13.46, "learning_rate": 5.710707576706325e-06, "loss": 0.4185, "step": 15921, "task_loss": 0.4186477065086365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2765969932079315, "epoch": 13.46, "learning_rate": 5.707576706324359e-06, "loss": 0.4065, "step": 15922, "task_loss": 0.6640077233314514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7695496678352356, "epoch": 13.46, "learning_rate": 5.704445835942392e-06, "loss": 0.6089, "step": 15923, "task_loss": 0.6421780586242676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2505209147930145, "epoch": 13.46, "learning_rate": 5.701314965560426e-06, "loss": 0.3968, "step": 15924, "task_loss": 0.5178574323654175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4183417856693268, "epoch": 13.46, "learning_rate": 5.69818409517846e-06, "loss": 0.4314, "step": 15925, "task_loss": 1.6645570993423462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3871919512748718, "epoch": 13.46, "learning_rate": 5.6950532247964935e-06, "loss": 0.3113, "step": 15926, "task_loss": 0.4058108627796173 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3791986405849457, "epoch": 13.46, "learning_rate": 5.691922354414528e-06, "loss": 0.4819, "step": 15927, "task_loss": 0.9749367237091064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.589381217956543, "epoch": 13.46, "learning_rate": 5.688791484032561e-06, "loss": 0.4115, "step": 15928, "task_loss": 1.173983097076416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46268895268440247, "epoch": 13.46, "learning_rate": 5.685660613650596e-06, "loss": 0.4992, "step": 15929, "task_loss": 0.6565352082252502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41647738218307495, "epoch": 13.47, "learning_rate": 5.682529743268629e-06, "loss": 0.5683, "step": 15930, "task_loss": 0.3910531997680664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2654232382774353, "epoch": 13.47, "learning_rate": 5.679398872886663e-06, "loss": 0.3754, "step": 15931, "task_loss": 0.3028082549571991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5623502731323242, "epoch": 13.47, "learning_rate": 5.676268002504696e-06, "loss": 0.5103, "step": 15932, "task_loss": 0.5411068797111511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7689804434776306, "epoch": 13.47, "learning_rate": 5.67313713212273e-06, "loss": 0.4421, "step": 15933, "task_loss": 0.512871265411377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6130416393280029, "epoch": 13.47, "learning_rate": 5.670006261740765e-06, "loss": 0.4656, "step": 15934, "task_loss": 1.2822399139404297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28325706720352173, "epoch": 13.47, "learning_rate": 5.666875391358798e-06, "loss": 0.4367, "step": 15935, "task_loss": 0.06324886530637741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4883943200111389, "epoch": 13.47, "learning_rate": 5.663744520976832e-06, "loss": 0.5022, "step": 15936, "task_loss": 0.6309923529624939 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48609280586242676, "epoch": 13.47, "learning_rate": 5.660613650594865e-06, "loss": 0.4694, "step": 15937, "task_loss": 0.186183899641037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24707096815109253, "epoch": 13.47, "learning_rate": 5.6574827802128994e-06, "loss": 0.3919, "step": 15938, "task_loss": 0.43113088607788086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3173335790634155, "epoch": 13.47, "learning_rate": 5.654351909830934e-06, "loss": 0.3868, "step": 15939, "task_loss": 0.20259366929531097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5341200232505798, "epoch": 13.47, "learning_rate": 5.651221039448967e-06, "loss": 0.4251, "step": 15940, "task_loss": 0.3123184144496918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5846365690231323, "epoch": 13.47, "learning_rate": 5.648090169067001e-06, "loss": 0.5011, "step": 15941, "task_loss": 1.6794158220291138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5474283695220947, "epoch": 13.48, "learning_rate": 5.644959298685035e-06, "loss": 0.5185, "step": 15942, "task_loss": 1.2550537586212158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29959794878959656, "epoch": 13.48, "learning_rate": 5.6418284283030685e-06, "loss": 0.44, "step": 15943, "task_loss": 0.6599660515785217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21556782722473145, "epoch": 13.48, "learning_rate": 5.638697557921102e-06, "loss": 0.4113, "step": 15944, "task_loss": 0.3264596462249756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4318702816963196, "epoch": 13.48, "learning_rate": 5.6355666875391355e-06, "loss": 0.3847, "step": 15945, "task_loss": 0.39921683073043823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6222168207168579, "epoch": 13.48, "learning_rate": 5.63243581715717e-06, "loss": 0.5045, "step": 15946, "task_loss": 0.9592676758766174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.784156322479248, "epoch": 13.48, "learning_rate": 5.629304946775204e-06, "loss": 0.4721, "step": 15947, "task_loss": 0.9398033022880554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5221670269966125, "epoch": 13.48, "learning_rate": 5.626174076393238e-06, "loss": 0.4583, "step": 15948, "task_loss": 0.34551313519477844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.376197874546051, "epoch": 13.48, "learning_rate": 5.623043206011271e-06, "loss": 0.3837, "step": 15949, "task_loss": 0.2125622183084488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25767913460731506, "epoch": 13.48, "learning_rate": 5.619912335629305e-06, "loss": 0.4268, "step": 15950, "task_loss": 0.9652297496795654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5921870470046997, "epoch": 13.48, "learning_rate": 5.616781465247339e-06, "loss": 0.3981, "step": 15951, "task_loss": 0.48496609926223755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.58829665184021, "epoch": 13.48, "learning_rate": 5.613650594865373e-06, "loss": 0.4804, "step": 15952, "task_loss": 1.8568871021270752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3025396168231964, "epoch": 13.48, "learning_rate": 5.610519724483407e-06, "loss": 0.4284, "step": 15953, "task_loss": 0.39380401372909546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33757054805755615, "epoch": 13.49, "learning_rate": 5.60738885410144e-06, "loss": 0.3951, "step": 15954, "task_loss": 0.7000381946563721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45047828555107117, "epoch": 13.49, "learning_rate": 5.6042579837194745e-06, "loss": 0.4026, "step": 15955, "task_loss": 0.6926413178443909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18036480247974396, "epoch": 13.49, "learning_rate": 5.601127113337508e-06, "loss": 0.4089, "step": 15956, "task_loss": 0.23210462927818298 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2378273606300354, "epoch": 13.49, "learning_rate": 5.597996242955541e-06, "loss": 0.3917, "step": 15957, "task_loss": 0.06588597595691681 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2111419290304184, "epoch": 13.49, "learning_rate": 5.594865372573576e-06, "loss": 0.5321, "step": 15958, "task_loss": 0.04554303362965584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4552912414073944, "epoch": 13.49, "learning_rate": 5.59173450219161e-06, "loss": 0.3942, "step": 15959, "task_loss": 0.4571760296821594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4326589107513428, "epoch": 13.49, "learning_rate": 5.5886036318096435e-06, "loss": 0.3942, "step": 15960, "task_loss": 0.8159021735191345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3890712857246399, "epoch": 13.49, "learning_rate": 5.585472761427677e-06, "loss": 0.4672, "step": 15961, "task_loss": 0.7981629371643066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30421605706214905, "epoch": 13.49, "learning_rate": 5.5823418910457105e-06, "loss": 0.376, "step": 15962, "task_loss": 0.06936442106962204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4774170219898224, "epoch": 13.49, "learning_rate": 5.579211020663745e-06, "loss": 0.5197, "step": 15963, "task_loss": 0.4953659772872925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4720361828804016, "epoch": 13.49, "learning_rate": 5.576080150281779e-06, "loss": 0.3766, "step": 15964, "task_loss": 0.35131198167800903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30617374181747437, "epoch": 13.5, "learning_rate": 5.572949279899813e-06, "loss": 0.3921, "step": 15965, "task_loss": 0.28859397768974304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5155210494995117, "epoch": 13.5, "learning_rate": 5.569818409517846e-06, "loss": 0.3773, "step": 15966, "task_loss": 0.7350102663040161 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3955211937427521, "epoch": 13.5, "learning_rate": 5.56668753913588e-06, "loss": 0.4527, "step": 15967, "task_loss": 0.3294168710708618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5139592885971069, "epoch": 13.5, "learning_rate": 5.563556668753914e-06, "loss": 0.451, "step": 15968, "task_loss": 1.0541377067565918 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40722912549972534, "epoch": 13.5, "learning_rate": 5.560425798371947e-06, "loss": 0.4355, "step": 15969, "task_loss": 0.27994653582572937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40357741713523865, "epoch": 13.5, "learning_rate": 5.557294927989982e-06, "loss": 0.3681, "step": 15970, "task_loss": 0.6403274536132812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.323789119720459, "epoch": 13.5, "learning_rate": 5.554164057608015e-06, "loss": 0.376, "step": 15971, "task_loss": 0.5609753727912903 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3713567554950714, "epoch": 13.5, "learning_rate": 5.5510331872260495e-06, "loss": 0.3585, "step": 15972, "task_loss": 1.6043535470962524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6105258464813232, "epoch": 13.5, "learning_rate": 5.547902316844083e-06, "loss": 0.4464, "step": 15973, "task_loss": 0.9944821000099182 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5385831594467163, "epoch": 13.5, "learning_rate": 5.544771446462116e-06, "loss": 0.3811, "step": 15974, "task_loss": 0.600767195224762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3583395183086395, "epoch": 13.5, "learning_rate": 5.54164057608015e-06, "loss": 0.4135, "step": 15975, "task_loss": 0.2936839461326599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6555427312850952, "epoch": 13.5, "learning_rate": 5.538509705698185e-06, "loss": 0.4436, "step": 15976, "task_loss": 0.9882228374481201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.516667902469635, "epoch": 13.51, "learning_rate": 5.5353788353162185e-06, "loss": 0.5146, "step": 15977, "task_loss": 0.3304472863674164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4090024530887604, "epoch": 13.51, "learning_rate": 5.532247964934252e-06, "loss": 0.3801, "step": 15978, "task_loss": 1.3184053897857666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2626952528953552, "epoch": 13.51, "learning_rate": 5.5291170945522855e-06, "loss": 0.3424, "step": 15979, "task_loss": 0.2551988363265991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41625720262527466, "epoch": 13.51, "learning_rate": 5.52598622417032e-06, "loss": 0.4649, "step": 15980, "task_loss": 0.43568575382232666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6065245866775513, "epoch": 13.51, "learning_rate": 5.522855353788353e-06, "loss": 0.5465, "step": 15981, "task_loss": 0.725151002407074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41320696473121643, "epoch": 13.51, "learning_rate": 5.519724483406388e-06, "loss": 0.5069, "step": 15982, "task_loss": 0.6172545552253723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5004768967628479, "epoch": 13.51, "learning_rate": 5.516593613024421e-06, "loss": 0.4499, "step": 15983, "task_loss": 0.3769071698188782 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3830989599227905, "epoch": 13.51, "learning_rate": 5.5134627426424546e-06, "loss": 0.482, "step": 15984, "task_loss": 1.053480863571167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4489799439907074, "epoch": 13.51, "learning_rate": 5.510331872260489e-06, "loss": 0.4054, "step": 15985, "task_loss": 0.30505886673927307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6031147241592407, "epoch": 13.51, "learning_rate": 5.507201001878522e-06, "loss": 0.4494, "step": 15986, "task_loss": 0.8860865235328674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31175529956817627, "epoch": 13.51, "learning_rate": 5.504070131496556e-06, "loss": 0.3239, "step": 15987, "task_loss": 0.17062212526798248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5012532472610474, "epoch": 13.51, "learning_rate": 5.50093926111459e-06, "loss": 0.4203, "step": 15988, "task_loss": 1.1606543064117432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46554845571517944, "epoch": 13.52, "learning_rate": 5.4978083907326245e-06, "loss": 0.3861, "step": 15989, "task_loss": 0.6500391960144043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3853854537010193, "epoch": 13.52, "learning_rate": 5.494677520350658e-06, "loss": 0.4496, "step": 15990, "task_loss": 0.7371827363967896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34395766258239746, "epoch": 13.52, "learning_rate": 5.4915466499686914e-06, "loss": 0.453, "step": 15991, "task_loss": 0.838026225566864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31092289090156555, "epoch": 13.52, "learning_rate": 5.488415779586725e-06, "loss": 0.3582, "step": 15992, "task_loss": 0.1275496482849121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4359634518623352, "epoch": 13.52, "learning_rate": 5.485284909204759e-06, "loss": 0.5347, "step": 15993, "task_loss": 0.5505552291870117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41188615560531616, "epoch": 13.52, "learning_rate": 5.482154038822793e-06, "loss": 0.4364, "step": 15994, "task_loss": 0.7051181793212891 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39964354038238525, "epoch": 13.52, "learning_rate": 5.479023168440827e-06, "loss": 0.5039, "step": 15995, "task_loss": 0.47206196188926697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3757767677307129, "epoch": 13.52, "learning_rate": 5.4758922980588605e-06, "loss": 0.3855, "step": 15996, "task_loss": 0.17769211530685425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4432290196418762, "epoch": 13.52, "learning_rate": 5.472761427676895e-06, "loss": 0.5385, "step": 15997, "task_loss": 1.1636346578598022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7582941651344299, "epoch": 13.52, "learning_rate": 5.469630557294928e-06, "loss": 0.4764, "step": 15998, "task_loss": 1.3704617023468018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37828555703163147, "epoch": 13.52, "learning_rate": 5.466499686912962e-06, "loss": 0.3586, "step": 15999, "task_loss": 0.6716278791427612 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5392243266105652, "epoch": 13.52, "learning_rate": 5.463368816530995e-06, "loss": 0.4854, "step": 16000, "task_loss": 0.9195689558982849 }, { "epoch": 13.52, "eval_accuracy": 0.9125940594059406, "eval_loss": 0.30570128560066223, "eval_runtime": 207.8601, "eval_samples_per_second": 121.476, "eval_steps_per_second": 0.953, "step": 16000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4036414623260498, "epoch": 13.53, "learning_rate": 5.4602379461490296e-06, "loss": 0.3628, "step": 16001, "task_loss": 0.22380591928958893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49639785289764404, "epoch": 13.53, "learning_rate": 5.457107075767064e-06, "loss": 0.5087, "step": 16002, "task_loss": 0.7226376533508301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38869723677635193, "epoch": 13.53, "learning_rate": 5.453976205385097e-06, "loss": 0.5049, "step": 16003, "task_loss": 0.7327865362167358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3434619903564453, "epoch": 13.53, "learning_rate": 5.450845335003131e-06, "loss": 0.3727, "step": 16004, "task_loss": 0.49876856803894043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41989225149154663, "epoch": 13.53, "learning_rate": 5.447714464621165e-06, "loss": 0.5537, "step": 16005, "task_loss": 0.49117064476013184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5478139519691467, "epoch": 13.53, "learning_rate": 5.444583594239199e-06, "loss": 0.5083, "step": 16006, "task_loss": 0.7126954197883606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2695299983024597, "epoch": 13.53, "learning_rate": 5.441452723857233e-06, "loss": 0.36, "step": 16007, "task_loss": 0.25075727701187134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5661010146141052, "epoch": 13.53, "learning_rate": 5.4383218534752664e-06, "loss": 0.4744, "step": 16008, "task_loss": 0.425432026386261 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31012818217277527, "epoch": 13.53, "learning_rate": 5.4351909830933e-06, "loss": 0.3301, "step": 16009, "task_loss": 0.2335669845342636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39023467898368835, "epoch": 13.53, "learning_rate": 5.432060112711334e-06, "loss": 0.5054, "step": 16010, "task_loss": 0.6272222995758057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5240060091018677, "epoch": 13.53, "learning_rate": 5.428929242329368e-06, "loss": 0.4122, "step": 16011, "task_loss": 0.9964689016342163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43776267766952515, "epoch": 13.53, "learning_rate": 5.425798371947401e-06, "loss": 0.3578, "step": 16012, "task_loss": 0.5484139919281006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3412007987499237, "epoch": 13.54, "learning_rate": 5.4226675015654355e-06, "loss": 0.3645, "step": 16013, "task_loss": 0.20128020644187927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4042973220348358, "epoch": 13.54, "learning_rate": 5.41953663118347e-06, "loss": 0.4341, "step": 16014, "task_loss": 0.13655425608158112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4584958255290985, "epoch": 13.54, "learning_rate": 5.416405760801503e-06, "loss": 0.5004, "step": 16015, "task_loss": 0.0927954614162445 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3882014751434326, "epoch": 13.54, "learning_rate": 5.413274890419537e-06, "loss": 0.422, "step": 16016, "task_loss": 0.8269271850585938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3838052749633789, "epoch": 13.54, "learning_rate": 5.41014402003757e-06, "loss": 0.3765, "step": 16017, "task_loss": 0.5043097734451294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5438799858093262, "epoch": 13.54, "learning_rate": 5.4070131496556046e-06, "loss": 0.4791, "step": 16018, "task_loss": 0.8986488580703735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4536951184272766, "epoch": 13.54, "learning_rate": 5.403882279273639e-06, "loss": 0.3712, "step": 16019, "task_loss": 0.41004613041877747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9956645965576172, "epoch": 13.54, "learning_rate": 5.400751408891672e-06, "loss": 0.5569, "step": 16020, "task_loss": 0.6558770537376404 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5111361742019653, "epoch": 13.54, "learning_rate": 5.397620538509706e-06, "loss": 0.5403, "step": 16021, "task_loss": 0.6831565499305725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3215492367744446, "epoch": 13.54, "learning_rate": 5.394489668127739e-06, "loss": 0.4004, "step": 16022, "task_loss": 0.6269617676734924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35631823539733887, "epoch": 13.54, "learning_rate": 5.391358797745774e-06, "loss": 0.4993, "step": 16023, "task_loss": 0.20804768800735474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6574362516403198, "epoch": 13.54, "learning_rate": 5.388227927363807e-06, "loss": 0.5495, "step": 16024, "task_loss": 1.169173240661621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41667595505714417, "epoch": 13.55, "learning_rate": 5.3850970569818414e-06, "loss": 0.4089, "step": 16025, "task_loss": 1.0024348497390747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2702389657497406, "epoch": 13.55, "learning_rate": 5.381966186599875e-06, "loss": 0.3935, "step": 16026, "task_loss": 0.33209502696990967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6420119404792786, "epoch": 13.55, "learning_rate": 5.378835316217909e-06, "loss": 0.5726, "step": 16027, "task_loss": 0.47903090715408325 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5385780930519104, "epoch": 13.55, "learning_rate": 5.375704445835943e-06, "loss": 0.3743, "step": 16028, "task_loss": 0.42161357402801514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5079104900360107, "epoch": 13.55, "learning_rate": 5.372573575453976e-06, "loss": 0.4946, "step": 16029, "task_loss": 1.2775317430496216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2885133922100067, "epoch": 13.55, "learning_rate": 5.36944270507201e-06, "loss": 0.3657, "step": 16030, "task_loss": 0.6432491540908813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37694671750068665, "epoch": 13.55, "learning_rate": 5.366311834690045e-06, "loss": 0.4198, "step": 16031, "task_loss": 0.3991008996963501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17778049409389496, "epoch": 13.55, "learning_rate": 5.363180964308078e-06, "loss": 0.4231, "step": 16032, "task_loss": 0.21060270071029663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6272455453872681, "epoch": 13.55, "learning_rate": 5.360050093926112e-06, "loss": 0.5207, "step": 16033, "task_loss": 0.4128865599632263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5376837253570557, "epoch": 13.55, "learning_rate": 5.356919223544145e-06, "loss": 0.4646, "step": 16034, "task_loss": 1.6268870830535889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45279863476753235, "epoch": 13.55, "learning_rate": 5.35378835316218e-06, "loss": 0.368, "step": 16035, "task_loss": 0.5698357820510864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5976235866546631, "epoch": 13.56, "learning_rate": 5.350657482780213e-06, "loss": 0.5024, "step": 16036, "task_loss": 0.20599649846553802 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38407468795776367, "epoch": 13.56, "learning_rate": 5.347526612398247e-06, "loss": 0.488, "step": 16037, "task_loss": 0.10284561663866043 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.623650848865509, "epoch": 13.56, "learning_rate": 5.344395742016281e-06, "loss": 0.4816, "step": 16038, "task_loss": 0.46716663241386414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30755311250686646, "epoch": 13.56, "learning_rate": 5.341264871634314e-06, "loss": 0.3229, "step": 16039, "task_loss": 0.10337025672197342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4861069619655609, "epoch": 13.56, "learning_rate": 5.338134001252349e-06, "loss": 0.4455, "step": 16040, "task_loss": 0.48740682005882263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3405797481536865, "epoch": 13.56, "learning_rate": 5.335003130870382e-06, "loss": 0.4508, "step": 16041, "task_loss": 0.9796607494354248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1923844814300537, "epoch": 13.56, "learning_rate": 5.331872260488416e-06, "loss": 0.3025, "step": 16042, "task_loss": 0.1105889081954956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.243870809674263, "epoch": 13.56, "learning_rate": 5.32874139010645e-06, "loss": 0.367, "step": 16043, "task_loss": 0.6961446404457092 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3451387584209442, "epoch": 13.56, "learning_rate": 5.325610519724484e-06, "loss": 0.5094, "step": 16044, "task_loss": 0.4197939336299896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5090410113334656, "epoch": 13.56, "learning_rate": 5.322479649342518e-06, "loss": 0.364, "step": 16045, "task_loss": 0.25697454810142517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4236462712287903, "epoch": 13.56, "learning_rate": 5.319348778960551e-06, "loss": 0.3144, "step": 16046, "task_loss": 0.4348564147949219 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.294482558965683, "epoch": 13.56, "learning_rate": 5.316217908578585e-06, "loss": 0.3164, "step": 16047, "task_loss": 0.47599852085113525 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3845425248146057, "epoch": 13.57, "learning_rate": 5.313087038196619e-06, "loss": 0.3683, "step": 16048, "task_loss": 0.6894305348396301 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6612570285797119, "epoch": 13.57, "learning_rate": 5.3099561678146525e-06, "loss": 0.6435, "step": 16049, "task_loss": 1.3233530521392822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4505549669265747, "epoch": 13.57, "learning_rate": 5.306825297432687e-06, "loss": 0.5209, "step": 16050, "task_loss": 0.9803087711334229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32036644220352173, "epoch": 13.57, "learning_rate": 5.30369442705072e-06, "loss": 0.4524, "step": 16051, "task_loss": 0.5422289967536926 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6499855518341064, "epoch": 13.57, "learning_rate": 5.300563556668755e-06, "loss": 0.4982, "step": 16052, "task_loss": 0.30145034193992615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5617957711219788, "epoch": 13.57, "learning_rate": 5.297432686286788e-06, "loss": 0.654, "step": 16053, "task_loss": 1.2619727849960327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44667714834213257, "epoch": 13.57, "learning_rate": 5.2943018159048215e-06, "loss": 0.5555, "step": 16054, "task_loss": 1.0651010274887085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3172281086444855, "epoch": 13.57, "learning_rate": 5.291170945522855e-06, "loss": 0.3503, "step": 16055, "task_loss": 0.6316465139389038 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5245024561882019, "epoch": 13.57, "learning_rate": 5.288040075140889e-06, "loss": 0.4981, "step": 16056, "task_loss": 1.0780495405197144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19772961735725403, "epoch": 13.57, "learning_rate": 5.284909204758924e-06, "loss": 0.4415, "step": 16057, "task_loss": 0.036752596497535706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.266097754240036, "epoch": 13.57, "learning_rate": 5.281778334376957e-06, "loss": 0.3683, "step": 16058, "task_loss": 0.7201486229896545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3713662028312683, "epoch": 13.57, "learning_rate": 5.278647463994991e-06, "loss": 0.3911, "step": 16059, "task_loss": 0.8085359334945679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42447954416275024, "epoch": 13.58, "learning_rate": 5.275516593613024e-06, "loss": 0.3017, "step": 16060, "task_loss": 0.9788237810134888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5023958683013916, "epoch": 13.58, "learning_rate": 5.272385723231058e-06, "loss": 0.4223, "step": 16061, "task_loss": 1.0490739345550537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2807427644729614, "epoch": 13.58, "learning_rate": 5.269254852849093e-06, "loss": 0.4472, "step": 16062, "task_loss": 0.08507099002599716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5737848281860352, "epoch": 13.58, "learning_rate": 5.266123982467126e-06, "loss": 0.4368, "step": 16063, "task_loss": 1.0913604497909546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.514731764793396, "epoch": 13.58, "learning_rate": 5.26299311208516e-06, "loss": 0.4831, "step": 16064, "task_loss": 0.43338340520858765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3279547691345215, "epoch": 13.58, "learning_rate": 5.259862241703194e-06, "loss": 0.4374, "step": 16065, "task_loss": 0.11337442696094513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40529075264930725, "epoch": 13.58, "learning_rate": 5.2567313713212275e-06, "loss": 0.4447, "step": 16066, "task_loss": 0.9912667870521545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49674004316329956, "epoch": 13.58, "learning_rate": 5.253600500939261e-06, "loss": 0.4344, "step": 16067, "task_loss": 0.7025578022003174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.311653733253479, "epoch": 13.58, "learning_rate": 5.250469630557295e-06, "loss": 0.3733, "step": 16068, "task_loss": 0.08689505606889725 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6140755414962769, "epoch": 13.58, "learning_rate": 5.24733876017533e-06, "loss": 0.4815, "step": 16069, "task_loss": 1.5242189168930054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20747864246368408, "epoch": 13.58, "learning_rate": 5.244207889793363e-06, "loss": 0.4627, "step": 16070, "task_loss": 0.3524254262447357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22503653168678284, "epoch": 13.58, "learning_rate": 5.2410770194113966e-06, "loss": 0.3965, "step": 16071, "task_loss": 0.4335981607437134 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2684359848499298, "epoch": 13.59, "learning_rate": 5.23794614902943e-06, "loss": 0.4786, "step": 16072, "task_loss": 0.6504862904548645 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4025508761405945, "epoch": 13.59, "learning_rate": 5.234815278647464e-06, "loss": 0.3882, "step": 16073, "task_loss": 0.3353551924228668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3488007187843323, "epoch": 13.59, "learning_rate": 5.231684408265499e-06, "loss": 0.5614, "step": 16074, "task_loss": 0.29208970069885254 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6023778915405273, "epoch": 13.59, "learning_rate": 5.228553537883532e-06, "loss": 0.4274, "step": 16075, "task_loss": 0.39236828684806824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.326160728931427, "epoch": 13.59, "learning_rate": 5.225422667501566e-06, "loss": 0.3962, "step": 16076, "task_loss": 0.36135122179985046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3914017677307129, "epoch": 13.59, "learning_rate": 5.222291797119599e-06, "loss": 0.4353, "step": 16077, "task_loss": 0.17125388979911804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6713229417800903, "epoch": 13.59, "learning_rate": 5.219160926737633e-06, "loss": 0.4779, "step": 16078, "task_loss": 1.5789567232131958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3441070020198822, "epoch": 13.59, "learning_rate": 5.216030056355667e-06, "loss": 0.4769, "step": 16079, "task_loss": 0.4025875926017761 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5170281529426575, "epoch": 13.59, "learning_rate": 5.212899185973701e-06, "loss": 0.4751, "step": 16080, "task_loss": 0.7047117948532104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.493792325258255, "epoch": 13.59, "learning_rate": 5.209768315591735e-06, "loss": 0.3653, "step": 16081, "task_loss": 0.982862114906311 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38152605295181274, "epoch": 13.59, "learning_rate": 5.206637445209769e-06, "loss": 0.4408, "step": 16082, "task_loss": 0.7838725447654724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5364397168159485, "epoch": 13.59, "learning_rate": 5.2035065748278025e-06, "loss": 0.5682, "step": 16083, "task_loss": 0.3777877390384674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7218778133392334, "epoch": 13.6, "learning_rate": 5.200375704445836e-06, "loss": 0.5497, "step": 16084, "task_loss": 1.60286545753479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1819290816783905, "epoch": 13.6, "learning_rate": 5.1972448340638694e-06, "loss": 0.3926, "step": 16085, "task_loss": 0.2527207136154175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47794321179389954, "epoch": 13.6, "learning_rate": 5.194113963681904e-06, "loss": 0.4637, "step": 16086, "task_loss": 1.2586846351623535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3859628438949585, "epoch": 13.6, "learning_rate": 5.190983093299938e-06, "loss": 0.5339, "step": 16087, "task_loss": 0.5405428409576416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3302335739135742, "epoch": 13.6, "learning_rate": 5.1878522229179716e-06, "loss": 0.3492, "step": 16088, "task_loss": 0.12915953993797302 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4460911750793457, "epoch": 13.6, "learning_rate": 5.184721352536005e-06, "loss": 0.3794, "step": 16089, "task_loss": 1.1942734718322754 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25631725788116455, "epoch": 13.6, "learning_rate": 5.181590482154039e-06, "loss": 0.3726, "step": 16090, "task_loss": 1.2074253559112549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20054759085178375, "epoch": 13.6, "learning_rate": 5.178459611772073e-06, "loss": 0.3595, "step": 16091, "task_loss": 0.37506580352783203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6808284521102905, "epoch": 13.6, "learning_rate": 5.175328741390106e-06, "loss": 0.522, "step": 16092, "task_loss": 0.28050923347473145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40247011184692383, "epoch": 13.6, "learning_rate": 5.172197871008141e-06, "loss": 0.3988, "step": 16093, "task_loss": 0.796729564666748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2846023440361023, "epoch": 13.6, "learning_rate": 5.169067000626174e-06, "loss": 0.4222, "step": 16094, "task_loss": 0.4401085674762726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20824560523033142, "epoch": 13.6, "learning_rate": 5.1659361302442084e-06, "loss": 0.3345, "step": 16095, "task_loss": 0.22011639177799225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36511164903640747, "epoch": 13.61, "learning_rate": 5.162805259862242e-06, "loss": 0.3161, "step": 16096, "task_loss": 0.5056166648864746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6216379404067993, "epoch": 13.61, "learning_rate": 5.159674389480275e-06, "loss": 0.4985, "step": 16097, "task_loss": 1.7628753185272217 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32389137148857117, "epoch": 13.61, "learning_rate": 5.156543519098309e-06, "loss": 0.4526, "step": 16098, "task_loss": 0.2882223129272461 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.463140070438385, "epoch": 13.61, "learning_rate": 5.153412648716344e-06, "loss": 0.4599, "step": 16099, "task_loss": 0.6913072466850281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23803076148033142, "epoch": 13.61, "learning_rate": 5.1502817783343775e-06, "loss": 0.3913, "step": 16100, "task_loss": 0.2460981160402298 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7458738684654236, "epoch": 13.61, "learning_rate": 5.147150907952411e-06, "loss": 0.5194, "step": 16101, "task_loss": 0.40965813398361206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.546964704990387, "epoch": 13.61, "learning_rate": 5.1440200375704445e-06, "loss": 0.367, "step": 16102, "task_loss": 0.2484133243560791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3314015865325928, "epoch": 13.61, "learning_rate": 5.140889167188479e-06, "loss": 0.321, "step": 16103, "task_loss": 0.4047233462333679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2780044972896576, "epoch": 13.61, "learning_rate": 5.137758296806512e-06, "loss": 0.3917, "step": 16104, "task_loss": 0.10848265141248703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5568065643310547, "epoch": 13.61, "learning_rate": 5.1346274264245466e-06, "loss": 0.4127, "step": 16105, "task_loss": 0.993651807308197 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29191163182258606, "epoch": 13.61, "learning_rate": 5.13149655604258e-06, "loss": 0.3605, "step": 16106, "task_loss": 0.33100348711013794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6408013105392456, "epoch": 13.61, "learning_rate": 5.128365685660614e-06, "loss": 0.3488, "step": 16107, "task_loss": 0.475220263004303 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42563390731811523, "epoch": 13.62, "learning_rate": 5.125234815278648e-06, "loss": 0.3513, "step": 16108, "task_loss": 0.7149974703788757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5422083735466003, "epoch": 13.62, "learning_rate": 5.122103944896681e-06, "loss": 0.4313, "step": 16109, "task_loss": 0.42790859937667847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20298467576503754, "epoch": 13.62, "learning_rate": 5.118973074514715e-06, "loss": 0.3334, "step": 16110, "task_loss": 0.1809345781803131 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23293530941009521, "epoch": 13.62, "learning_rate": 5.115842204132749e-06, "loss": 0.3778, "step": 16111, "task_loss": 0.4095805287361145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4372689127922058, "epoch": 13.62, "learning_rate": 5.1127113337507834e-06, "loss": 0.5029, "step": 16112, "task_loss": 0.8448078036308289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36935216188430786, "epoch": 13.62, "learning_rate": 5.109580463368817e-06, "loss": 0.5092, "step": 16113, "task_loss": 0.5782462358474731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.421531617641449, "epoch": 13.62, "learning_rate": 5.10644959298685e-06, "loss": 0.5077, "step": 16114, "task_loss": 0.07390519976615906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4274154007434845, "epoch": 13.62, "learning_rate": 5.103318722604884e-06, "loss": 0.4391, "step": 16115, "task_loss": 0.5172128081321716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40625545382499695, "epoch": 13.62, "learning_rate": 5.100187852222918e-06, "loss": 0.4315, "step": 16116, "task_loss": 0.20484404265880585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6973296999931335, "epoch": 13.62, "learning_rate": 5.0970569818409525e-06, "loss": 0.5393, "step": 16117, "task_loss": 0.5558215379714966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39443355798721313, "epoch": 13.62, "learning_rate": 5.093926111458986e-06, "loss": 0.4304, "step": 16118, "task_loss": 0.9654121994972229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6403842568397522, "epoch": 13.63, "learning_rate": 5.0907952410770195e-06, "loss": 0.4167, "step": 16119, "task_loss": 0.8495772480964661 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3976494073867798, "epoch": 13.63, "learning_rate": 5.087664370695054e-06, "loss": 0.369, "step": 16120, "task_loss": 0.4482710659503937 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2671700417995453, "epoch": 13.63, "learning_rate": 5.084533500313087e-06, "loss": 0.4554, "step": 16121, "task_loss": 0.8403947353363037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35987919569015503, "epoch": 13.63, "learning_rate": 5.081402629931121e-06, "loss": 0.5462, "step": 16122, "task_loss": 0.5547089576721191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6281348466873169, "epoch": 13.63, "learning_rate": 5.078271759549155e-06, "loss": 0.4, "step": 16123, "task_loss": 0.37879863381385803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2932904362678528, "epoch": 13.63, "learning_rate": 5.0751408891671885e-06, "loss": 0.3702, "step": 16124, "task_loss": 0.3616836369037628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.332045316696167, "epoch": 13.63, "learning_rate": 5.072010018785223e-06, "loss": 0.3416, "step": 16125, "task_loss": 0.7503412365913391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46532684564590454, "epoch": 13.63, "learning_rate": 5.068879148403256e-06, "loss": 0.3953, "step": 16126, "task_loss": 0.25388994812965393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38975849747657776, "epoch": 13.63, "learning_rate": 5.06574827802129e-06, "loss": 0.4962, "step": 16127, "task_loss": 0.47701019048690796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5208578109741211, "epoch": 13.63, "learning_rate": 5.062617407639324e-06, "loss": 0.4801, "step": 16128, "task_loss": 0.3705202341079712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2587718367576599, "epoch": 13.63, "learning_rate": 5.0594865372573584e-06, "loss": 0.3947, "step": 16129, "task_loss": 0.40447908639907837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30361804366111755, "epoch": 13.63, "learning_rate": 5.056355666875392e-06, "loss": 0.4136, "step": 16130, "task_loss": 0.41957351565361023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46618396043777466, "epoch": 13.64, "learning_rate": 5.053224796493425e-06, "loss": 0.6024, "step": 16131, "task_loss": 0.6535351276397705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3248808979988098, "epoch": 13.64, "learning_rate": 5.050093926111459e-06, "loss": 0.4048, "step": 16132, "task_loss": 0.32404181361198425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3602469861507416, "epoch": 13.64, "learning_rate": 5.046963055729493e-06, "loss": 0.3243, "step": 16133, "task_loss": 0.5457889437675476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3629108965396881, "epoch": 13.64, "learning_rate": 5.043832185347527e-06, "loss": 0.3815, "step": 16134, "task_loss": 0.9886918663978577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4289121627807617, "epoch": 13.64, "learning_rate": 5.04070131496556e-06, "loss": 0.4363, "step": 16135, "task_loss": 0.6204341650009155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5344879627227783, "epoch": 13.64, "learning_rate": 5.0375704445835945e-06, "loss": 0.4565, "step": 16136, "task_loss": 0.9210548400878906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3039117455482483, "epoch": 13.64, "learning_rate": 5.034439574201629e-06, "loss": 0.3152, "step": 16137, "task_loss": 0.49400952458381653 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5426570773124695, "epoch": 13.64, "learning_rate": 5.031308703819662e-06, "loss": 0.4768, "step": 16138, "task_loss": 0.6366143226623535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5794742107391357, "epoch": 13.64, "learning_rate": 5.028177833437696e-06, "loss": 0.4217, "step": 16139, "task_loss": 0.6998840570449829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20152977108955383, "epoch": 13.64, "learning_rate": 5.025046963055729e-06, "loss": 0.4219, "step": 16140, "task_loss": 0.7252890467643738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3293999433517456, "epoch": 13.64, "learning_rate": 5.0219160926737635e-06, "loss": 0.4459, "step": 16141, "task_loss": 0.8488249778747559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35429391264915466, "epoch": 13.64, "learning_rate": 5.018785222291798e-06, "loss": 0.3683, "step": 16142, "task_loss": 0.20388488471508026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4712735712528229, "epoch": 13.65, "learning_rate": 5.015654351909831e-06, "loss": 0.3398, "step": 16143, "task_loss": 0.5039381384849548 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3660573363304138, "epoch": 13.65, "learning_rate": 5.012523481527865e-06, "loss": 0.3671, "step": 16144, "task_loss": 0.6971773505210876 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.314670205116272, "epoch": 13.65, "learning_rate": 5.009392611145899e-06, "loss": 0.3059, "step": 16145, "task_loss": 0.776339054107666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.410808265209198, "epoch": 13.65, "learning_rate": 5.006261740763933e-06, "loss": 0.474, "step": 16146, "task_loss": 0.3339221477508545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5091865062713623, "epoch": 13.65, "learning_rate": 5.003130870381966e-06, "loss": 0.5411, "step": 16147, "task_loss": 0.7648942470550537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19494396448135376, "epoch": 13.65, "learning_rate": 5e-06, "loss": 0.3102, "step": 16148, "task_loss": 0.21059070527553558 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2954481244087219, "epoch": 13.65, "learning_rate": 4.996869129618034e-06, "loss": 0.4125, "step": 16149, "task_loss": 0.8240920305252075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3258790075778961, "epoch": 13.65, "learning_rate": 4.993738259236068e-06, "loss": 0.4168, "step": 16150, "task_loss": 0.581048309803009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.252301424741745, "epoch": 13.65, "learning_rate": 4.990607388854102e-06, "loss": 0.3425, "step": 16151, "task_loss": 0.3207944631576538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3277853727340698, "epoch": 13.65, "learning_rate": 4.987476518472135e-06, "loss": 0.4148, "step": 16152, "task_loss": 0.494858056306839 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46667081117630005, "epoch": 13.65, "learning_rate": 4.984345648090169e-06, "loss": 0.3599, "step": 16153, "task_loss": 0.6346244812011719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5440918803215027, "epoch": 13.65, "learning_rate": 4.981214777708204e-06, "loss": 0.3984, "step": 16154, "task_loss": 0.579608142375946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2424282729625702, "epoch": 13.66, "learning_rate": 4.978083907326237e-06, "loss": 0.4038, "step": 16155, "task_loss": 0.38494113087654114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3418683111667633, "epoch": 13.66, "learning_rate": 4.974953036944271e-06, "loss": 0.2798, "step": 16156, "task_loss": 0.14834898710250854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4941818416118622, "epoch": 13.66, "learning_rate": 4.971822166562304e-06, "loss": 0.475, "step": 16157, "task_loss": 0.368724524974823 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.453991174697876, "epoch": 13.66, "learning_rate": 4.9686912961803386e-06, "loss": 0.4877, "step": 16158, "task_loss": 1.0746623277664185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5080498456954956, "epoch": 13.66, "learning_rate": 4.965560425798372e-06, "loss": 0.3811, "step": 16159, "task_loss": 0.8313778042793274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2683848440647125, "epoch": 13.66, "learning_rate": 4.962429555416406e-06, "loss": 0.4437, "step": 16160, "task_loss": 0.23468562960624695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5286918878555298, "epoch": 13.66, "learning_rate": 4.95929868503444e-06, "loss": 0.5583, "step": 16161, "task_loss": 0.7216743230819702 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3213794529438019, "epoch": 13.66, "learning_rate": 4.956167814652473e-06, "loss": 0.453, "step": 16162, "task_loss": 0.1379069834947586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45332956314086914, "epoch": 13.66, "learning_rate": 4.953036944270508e-06, "loss": 0.3791, "step": 16163, "task_loss": 0.3055698871612549 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43561747670173645, "epoch": 13.66, "learning_rate": 4.949906073888541e-06, "loss": 0.4168, "step": 16164, "task_loss": 0.650242805480957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6163140535354614, "epoch": 13.66, "learning_rate": 4.9467752035065746e-06, "loss": 0.5221, "step": 16165, "task_loss": 0.7045388221740723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26115912199020386, "epoch": 13.66, "learning_rate": 4.943644333124609e-06, "loss": 0.4271, "step": 16166, "task_loss": 0.3606419861316681 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5674836039543152, "epoch": 13.67, "learning_rate": 4.940513462742643e-06, "loss": 0.4181, "step": 16167, "task_loss": 0.21306125819683075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42815250158309937, "epoch": 13.67, "learning_rate": 4.937382592360677e-06, "loss": 0.4992, "step": 16168, "task_loss": 0.5018365383148193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8114582896232605, "epoch": 13.67, "learning_rate": 4.93425172197871e-06, "loss": 0.4644, "step": 16169, "task_loss": 0.8349753022193909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7057942152023315, "epoch": 13.67, "learning_rate": 4.931120851596744e-06, "loss": 0.5257, "step": 16170, "task_loss": 1.091231346130371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.511738121509552, "epoch": 13.67, "learning_rate": 4.927989981214778e-06, "loss": 0.3706, "step": 16171, "task_loss": 0.2357608526945114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5422924757003784, "epoch": 13.67, "learning_rate": 4.924859110832812e-06, "loss": 0.5161, "step": 16172, "task_loss": 0.32568761706352234 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6464292407035828, "epoch": 13.67, "learning_rate": 4.921728240450846e-06, "loss": 0.4749, "step": 16173, "task_loss": 0.4286509156227112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4423980712890625, "epoch": 13.67, "learning_rate": 4.918597370068879e-06, "loss": 0.4892, "step": 16174, "task_loss": 1.1492811441421509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16982021927833557, "epoch": 13.67, "learning_rate": 4.9154664996869136e-06, "loss": 0.366, "step": 16175, "task_loss": 0.5682809948921204 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42734140157699585, "epoch": 13.67, "learning_rate": 4.912335629304947e-06, "loss": 0.4707, "step": 16176, "task_loss": 0.8059224486351013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21392759680747986, "epoch": 13.67, "learning_rate": 4.9092047589229805e-06, "loss": 0.4246, "step": 16177, "task_loss": 0.4320198595523834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2101873904466629, "epoch": 13.67, "learning_rate": 4.906073888541014e-06, "loss": 0.5097, "step": 16178, "task_loss": 0.44220227003097534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21425077319145203, "epoch": 13.68, "learning_rate": 4.902943018159048e-06, "loss": 0.3655, "step": 16179, "task_loss": 0.14982491731643677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3197284936904907, "epoch": 13.68, "learning_rate": 4.899812147777083e-06, "loss": 0.3685, "step": 16180, "task_loss": 0.483199805021286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49256670475006104, "epoch": 13.68, "learning_rate": 4.896681277395116e-06, "loss": 0.4356, "step": 16181, "task_loss": 0.41316017508506775 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38196229934692383, "epoch": 13.68, "learning_rate": 4.89355040701315e-06, "loss": 0.3995, "step": 16182, "task_loss": 0.05831027030944824 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3232452869415283, "epoch": 13.68, "learning_rate": 4.890419536631184e-06, "loss": 0.3498, "step": 16183, "task_loss": 0.845448911190033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39209267497062683, "epoch": 13.68, "learning_rate": 4.887288666249217e-06, "loss": 0.3465, "step": 16184, "task_loss": 0.36503222584724426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5453774929046631, "epoch": 13.68, "learning_rate": 4.884157795867252e-06, "loss": 0.3552, "step": 16185, "task_loss": 0.7832910418510437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37433922290802, "epoch": 13.68, "learning_rate": 4.881026925485285e-06, "loss": 0.4171, "step": 16186, "task_loss": 0.2768784463405609 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37164610624313354, "epoch": 13.68, "learning_rate": 4.877896055103319e-06, "loss": 0.3552, "step": 16187, "task_loss": 0.4543629288673401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40140819549560547, "epoch": 13.68, "learning_rate": 4.874765184721353e-06, "loss": 0.4358, "step": 16188, "task_loss": 0.6504811644554138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5983971357345581, "epoch": 13.68, "learning_rate": 4.8716343143393864e-06, "loss": 0.4692, "step": 16189, "task_loss": 1.1767786741256714 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5717809200286865, "epoch": 13.69, "learning_rate": 4.86850344395742e-06, "loss": 0.503, "step": 16190, "task_loss": 0.4889140725135803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4132276177406311, "epoch": 13.69, "learning_rate": 4.865372573575454e-06, "loss": 0.4934, "step": 16191, "task_loss": 1.4911704063415527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39464372396469116, "epoch": 13.69, "learning_rate": 4.8622417031934886e-06, "loss": 0.3361, "step": 16192, "task_loss": 0.4080425798892975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7211109399795532, "epoch": 13.69, "learning_rate": 4.859110832811522e-06, "loss": 0.445, "step": 16193, "task_loss": 0.41585633158683777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3020575940608978, "epoch": 13.69, "learning_rate": 4.8559799624295555e-06, "loss": 0.4402, "step": 16194, "task_loss": 0.3905579447746277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40286117792129517, "epoch": 13.69, "learning_rate": 4.852849092047589e-06, "loss": 0.457, "step": 16195, "task_loss": 0.7491313219070435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38009482622146606, "epoch": 13.69, "learning_rate": 4.849718221665623e-06, "loss": 0.503, "step": 16196, "task_loss": 0.8006600737571716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5070353746414185, "epoch": 13.69, "learning_rate": 4.846587351283658e-06, "loss": 0.4985, "step": 16197, "task_loss": 0.5589599013328552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26326242089271545, "epoch": 13.69, "learning_rate": 4.843456480901691e-06, "loss": 0.3721, "step": 16198, "task_loss": 1.0333603620529175 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42275547981262207, "epoch": 13.69, "learning_rate": 4.840325610519725e-06, "loss": 0.4537, "step": 16199, "task_loss": 0.7546346187591553 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6580424308776855, "epoch": 13.69, "learning_rate": 4.837194740137758e-06, "loss": 0.4719, "step": 16200, "task_loss": 0.3598799705505371 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.543779730796814, "epoch": 13.69, "learning_rate": 4.834063869755792e-06, "loss": 0.3916, "step": 16201, "task_loss": 0.7323631048202515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48661765456199646, "epoch": 13.7, "learning_rate": 4.830932999373826e-06, "loss": 0.5175, "step": 16202, "task_loss": 0.37324294447898865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.430931031703949, "epoch": 13.7, "learning_rate": 4.82780212899186e-06, "loss": 0.3957, "step": 16203, "task_loss": 0.8337973356246948 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34346166253089905, "epoch": 13.7, "learning_rate": 4.824671258609894e-06, "loss": 0.3699, "step": 16204, "task_loss": 0.9088588356971741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44018521904945374, "epoch": 13.7, "learning_rate": 4.821540388227928e-06, "loss": 0.4886, "step": 16205, "task_loss": 1.227975845336914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3458050489425659, "epoch": 13.7, "learning_rate": 4.8184095178459615e-06, "loss": 0.4435, "step": 16206, "task_loss": 0.5419281125068665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5529592633247375, "epoch": 13.7, "learning_rate": 4.815278647463995e-06, "loss": 0.3686, "step": 16207, "task_loss": 0.7819405198097229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5445743203163147, "epoch": 13.7, "learning_rate": 4.812147777082028e-06, "loss": 0.4461, "step": 16208, "task_loss": 1.0574854612350464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35288044810295105, "epoch": 13.7, "learning_rate": 4.809016906700064e-06, "loss": 0.4258, "step": 16209, "task_loss": 0.340836763381958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16124124825000763, "epoch": 13.7, "learning_rate": 4.805886036318097e-06, "loss": 0.4702, "step": 16210, "task_loss": 0.10292944312095642 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44651293754577637, "epoch": 13.7, "learning_rate": 4.8027551659361305e-06, "loss": 0.4233, "step": 16211, "task_loss": 0.41115978360176086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36049288511276245, "epoch": 13.7, "learning_rate": 4.799624295554164e-06, "loss": 0.3909, "step": 16212, "task_loss": 0.2390042245388031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6214162707328796, "epoch": 13.7, "learning_rate": 4.796493425172198e-06, "loss": 0.5119, "step": 16213, "task_loss": 0.5809746384620667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3237493336200714, "epoch": 13.71, "learning_rate": 4.793362554790232e-06, "loss": 0.5169, "step": 16214, "task_loss": 0.33680030703544617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3941512405872345, "epoch": 13.71, "learning_rate": 4.790231684408266e-06, "loss": 0.4123, "step": 16215, "task_loss": 0.37735259532928467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.13910944759845734, "epoch": 13.71, "learning_rate": 4.7871008140263e-06, "loss": 0.3014, "step": 16216, "task_loss": 0.005849303677678108 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48596489429473877, "epoch": 13.71, "learning_rate": 4.783969943644333e-06, "loss": 0.4678, "step": 16217, "task_loss": 0.915682315826416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3852992653846741, "epoch": 13.71, "learning_rate": 4.780839073262367e-06, "loss": 0.4526, "step": 16218, "task_loss": 0.52094566822052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5256748199462891, "epoch": 13.71, "learning_rate": 4.777708202880401e-06, "loss": 0.4152, "step": 16219, "task_loss": 1.9011647701263428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18933573365211487, "epoch": 13.71, "learning_rate": 4.774577332498434e-06, "loss": 0.4043, "step": 16220, "task_loss": 0.7629954814910889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26953208446502686, "epoch": 13.71, "learning_rate": 4.771446462116469e-06, "loss": 0.4158, "step": 16221, "task_loss": 0.8448173403739929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5176976919174194, "epoch": 13.71, "learning_rate": 4.768315591734503e-06, "loss": 0.5824, "step": 16222, "task_loss": 0.8014988899230957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2955090403556824, "epoch": 13.71, "learning_rate": 4.7651847213525365e-06, "loss": 0.3463, "step": 16223, "task_loss": 0.8149003982543945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3695441484451294, "epoch": 13.71, "learning_rate": 4.76205385097057e-06, "loss": 0.4316, "step": 16224, "task_loss": 0.39532291889190674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7606106996536255, "epoch": 13.71, "learning_rate": 4.758922980588603e-06, "loss": 0.5841, "step": 16225, "task_loss": 1.8312904834747314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2746717929840088, "epoch": 13.72, "learning_rate": 4.755792110206638e-06, "loss": 0.367, "step": 16226, "task_loss": 0.6109702587127686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6886279582977295, "epoch": 13.72, "learning_rate": 4.752661239824671e-06, "loss": 0.4387, "step": 16227, "task_loss": 0.7532926797866821 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2690354287624359, "epoch": 13.72, "learning_rate": 4.7495303694427055e-06, "loss": 0.3288, "step": 16228, "task_loss": 0.2842206358909607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4309701919555664, "epoch": 13.72, "learning_rate": 4.746399499060739e-06, "loss": 0.4368, "step": 16229, "task_loss": 0.6715153455734253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22362777590751648, "epoch": 13.72, "learning_rate": 4.743268628678773e-06, "loss": 0.331, "step": 16230, "task_loss": 0.275934636592865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4094560742378235, "epoch": 13.72, "learning_rate": 4.740137758296807e-06, "loss": 0.4585, "step": 16231, "task_loss": 0.0867387056350708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4482269883155823, "epoch": 13.72, "learning_rate": 4.73700688791484e-06, "loss": 0.3133, "step": 16232, "task_loss": 0.5083227157592773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40505921840667725, "epoch": 13.72, "learning_rate": 4.733876017532874e-06, "loss": 0.427, "step": 16233, "task_loss": 0.2645764946937561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.370595246553421, "epoch": 13.72, "learning_rate": 4.730745147150908e-06, "loss": 0.4039, "step": 16234, "task_loss": 0.3144499361515045 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22448313236236572, "epoch": 13.72, "learning_rate": 4.727614276768942e-06, "loss": 0.3375, "step": 16235, "task_loss": 0.6536279916763306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34510645270347595, "epoch": 13.72, "learning_rate": 4.724483406386976e-06, "loss": 0.3795, "step": 16236, "task_loss": 0.5949620604515076 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36564162373542786, "epoch": 13.72, "learning_rate": 4.721352536005009e-06, "loss": 0.3887, "step": 16237, "task_loss": 0.27179402112960815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3006308376789093, "epoch": 13.73, "learning_rate": 4.718221665623043e-06, "loss": 0.397, "step": 16238, "task_loss": 0.3179163336753845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4897891581058502, "epoch": 13.73, "learning_rate": 4.715090795241077e-06, "loss": 0.5527, "step": 16239, "task_loss": 0.754901111125946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4589666724205017, "epoch": 13.73, "learning_rate": 4.7119599248591115e-06, "loss": 0.4569, "step": 16240, "task_loss": 0.46471288800239563 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35707151889801025, "epoch": 13.73, "learning_rate": 4.708829054477145e-06, "loss": 0.3673, "step": 16241, "task_loss": 0.4906906485557556 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30933552980422974, "epoch": 13.73, "learning_rate": 4.7056981840951784e-06, "loss": 0.5132, "step": 16242, "task_loss": 0.4966990649700165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32958680391311646, "epoch": 13.73, "learning_rate": 4.702567313713213e-06, "loss": 0.4665, "step": 16243, "task_loss": 0.2664606273174286 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43378758430480957, "epoch": 13.73, "learning_rate": 4.699436443331246e-06, "loss": 0.4033, "step": 16244, "task_loss": 0.8499917984008789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3890218734741211, "epoch": 13.73, "learning_rate": 4.69630557294928e-06, "loss": 0.3719, "step": 16245, "task_loss": 0.9232933521270752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5595427751541138, "epoch": 13.73, "learning_rate": 4.693174702567314e-06, "loss": 0.508, "step": 16246, "task_loss": 0.5690313577651978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45122167468070984, "epoch": 13.73, "learning_rate": 4.690043832185348e-06, "loss": 0.4175, "step": 16247, "task_loss": 0.41517582535743713 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48059722781181335, "epoch": 13.73, "learning_rate": 4.686912961803382e-06, "loss": 0.4113, "step": 16248, "task_loss": 1.5121381282806396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3209276497364044, "epoch": 13.73, "learning_rate": 4.683782091421415e-06, "loss": 0.3909, "step": 16249, "task_loss": 0.4494827389717102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6911919116973877, "epoch": 13.74, "learning_rate": 4.680651221039449e-06, "loss": 0.5534, "step": 16250, "task_loss": 2.043896198272705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24825331568717957, "epoch": 13.74, "learning_rate": 4.677520350657483e-06, "loss": 0.3636, "step": 16251, "task_loss": 0.6797268390655518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3824015259742737, "epoch": 13.74, "learning_rate": 4.674389480275517e-06, "loss": 0.5934, "step": 16252, "task_loss": 0.48538655042648315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24030788242816925, "epoch": 13.74, "learning_rate": 4.671258609893551e-06, "loss": 0.3327, "step": 16253, "task_loss": 0.4447020888328552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3268396854400635, "epoch": 13.74, "learning_rate": 4.668127739511584e-06, "loss": 0.4487, "step": 16254, "task_loss": 0.8719024658203125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.492361843585968, "epoch": 13.74, "learning_rate": 4.664996869129618e-06, "loss": 0.4174, "step": 16255, "task_loss": 0.2093188762664795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3042285144329071, "epoch": 13.74, "learning_rate": 4.661865998747652e-06, "loss": 0.3942, "step": 16256, "task_loss": 0.10577571392059326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3332201838493347, "epoch": 13.74, "learning_rate": 4.658735128365686e-06, "loss": 0.5236, "step": 16257, "task_loss": 1.042197346687317 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.379880428314209, "epoch": 13.74, "learning_rate": 4.65560425798372e-06, "loss": 0.371, "step": 16258, "task_loss": 0.6299658417701721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6156659126281738, "epoch": 13.74, "learning_rate": 4.6524733876017534e-06, "loss": 0.4797, "step": 16259, "task_loss": 0.9281580448150635 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5926430225372314, "epoch": 13.74, "learning_rate": 4.649342517219788e-06, "loss": 0.5383, "step": 16260, "task_loss": 0.6555206775665283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3797641396522522, "epoch": 13.75, "learning_rate": 4.646211646837821e-06, "loss": 0.4025, "step": 16261, "task_loss": 0.6484031677246094 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6282318830490112, "epoch": 13.75, "learning_rate": 4.643080776455855e-06, "loss": 0.4896, "step": 16262, "task_loss": 0.42293086647987366 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25777095556259155, "epoch": 13.75, "learning_rate": 4.639949906073888e-06, "loss": 0.4579, "step": 16263, "task_loss": 0.22711826860904694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43133556842803955, "epoch": 13.75, "learning_rate": 4.6368190356919225e-06, "loss": 0.3703, "step": 16264, "task_loss": 0.4601871371269226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15309998393058777, "epoch": 13.75, "learning_rate": 4.633688165309957e-06, "loss": 0.4111, "step": 16265, "task_loss": 0.40187782049179077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7548449039459229, "epoch": 13.75, "learning_rate": 4.63055729492799e-06, "loss": 0.6086, "step": 16266, "task_loss": 1.5011258125305176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42930132150650024, "epoch": 13.75, "learning_rate": 4.627426424546024e-06, "loss": 0.5322, "step": 16267, "task_loss": 1.4493515491485596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42985057830810547, "epoch": 13.75, "learning_rate": 4.624295554164058e-06, "loss": 0.5484, "step": 16268, "task_loss": 0.4303368628025055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2582542896270752, "epoch": 13.75, "learning_rate": 4.6211646837820916e-06, "loss": 0.402, "step": 16269, "task_loss": 0.07665666937828064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3800501525402069, "epoch": 13.75, "learning_rate": 4.618033813400125e-06, "loss": 0.3859, "step": 16270, "task_loss": 0.2850613594055176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2522447407245636, "epoch": 13.75, "learning_rate": 4.614902943018159e-06, "loss": 0.3865, "step": 16271, "task_loss": 0.4284813702106476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3580248951911926, "epoch": 13.75, "learning_rate": 4.611772072636193e-06, "loss": 0.4183, "step": 16272, "task_loss": 0.5364897847175598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3602379560470581, "epoch": 13.76, "learning_rate": 4.608641202254227e-06, "loss": 0.401, "step": 16273, "task_loss": 0.6071609258651733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3158413767814636, "epoch": 13.76, "learning_rate": 4.605510331872261e-06, "loss": 0.3544, "step": 16274, "task_loss": 0.611625611782074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34263819456100464, "epoch": 13.76, "learning_rate": 4.602379461490294e-06, "loss": 0.4109, "step": 16275, "task_loss": 0.7881128787994385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44034016132354736, "epoch": 13.76, "learning_rate": 4.599248591108328e-06, "loss": 0.3824, "step": 16276, "task_loss": 0.37123656272888184 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4127587676048279, "epoch": 13.76, "learning_rate": 4.596117720726363e-06, "loss": 0.4153, "step": 16277, "task_loss": 0.21245639026165009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48851561546325684, "epoch": 13.76, "learning_rate": 4.592986850344396e-06, "loss": 0.4099, "step": 16278, "task_loss": 0.7268454432487488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6395609378814697, "epoch": 13.76, "learning_rate": 4.58985597996243e-06, "loss": 0.5659, "step": 16279, "task_loss": 1.0998003482818604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27689045667648315, "epoch": 13.76, "learning_rate": 4.586725109580463e-06, "loss": 0.3981, "step": 16280, "task_loss": 0.658164918422699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16983315348625183, "epoch": 13.76, "learning_rate": 4.5835942391984975e-06, "loss": 0.3813, "step": 16281, "task_loss": 0.1236528679728508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42601388692855835, "epoch": 13.76, "learning_rate": 4.580463368816531e-06, "loss": 0.5139, "step": 16282, "task_loss": 1.0860702991485596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31995338201522827, "epoch": 13.76, "learning_rate": 4.577332498434565e-06, "loss": 0.3865, "step": 16283, "task_loss": 0.3562820851802826 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21619242429733276, "epoch": 13.76, "learning_rate": 4.574201628052599e-06, "loss": 0.3862, "step": 16284, "task_loss": 0.43613240122795105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4683340787887573, "epoch": 13.77, "learning_rate": 4.571070757670633e-06, "loss": 0.4584, "step": 16285, "task_loss": 0.8397165536880493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32805922627449036, "epoch": 13.77, "learning_rate": 4.567939887288667e-06, "loss": 0.4066, "step": 16286, "task_loss": 0.23632140457630157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21004608273506165, "epoch": 13.77, "learning_rate": 4.5648090169067e-06, "loss": 0.3992, "step": 16287, "task_loss": 0.4270586371421814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2419108748435974, "epoch": 13.77, "learning_rate": 4.5616781465247335e-06, "loss": 0.4146, "step": 16288, "task_loss": 0.2625471353530884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42344632744789124, "epoch": 13.77, "learning_rate": 4.558547276142768e-06, "loss": 0.4237, "step": 16289, "task_loss": 1.3555259704589844 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38177263736724854, "epoch": 13.77, "learning_rate": 4.555416405760802e-06, "loss": 0.4302, "step": 16290, "task_loss": 0.7238706350326538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30893319845199585, "epoch": 13.77, "learning_rate": 4.552285535378836e-06, "loss": 0.376, "step": 16291, "task_loss": 0.35587137937545776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3455009162425995, "epoch": 13.77, "learning_rate": 4.549154664996869e-06, "loss": 0.506, "step": 16292, "task_loss": 0.6384402513504028 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28443068265914917, "epoch": 13.77, "learning_rate": 4.546023794614903e-06, "loss": 0.3075, "step": 16293, "task_loss": 1.304082989692688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3282212018966675, "epoch": 13.77, "learning_rate": 4.542892924232937e-06, "loss": 0.5044, "step": 16294, "task_loss": 0.0666128471493721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40870821475982666, "epoch": 13.77, "learning_rate": 4.539762053850971e-06, "loss": 0.4279, "step": 16295, "task_loss": 0.3762693703174591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3825893998146057, "epoch": 13.77, "learning_rate": 4.536631183469005e-06, "loss": 0.479, "step": 16296, "task_loss": 0.6652783155441284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40028125047683716, "epoch": 13.78, "learning_rate": 4.533500313087038e-06, "loss": 0.3676, "step": 16297, "task_loss": 0.7624350786209106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43324822187423706, "epoch": 13.78, "learning_rate": 4.5303694427050725e-06, "loss": 0.5924, "step": 16298, "task_loss": 0.6737279295921326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41089633107185364, "epoch": 13.78, "learning_rate": 4.527238572323106e-06, "loss": 0.4246, "step": 16299, "task_loss": 0.4042293429374695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31671813130378723, "epoch": 13.78, "learning_rate": 4.5241077019411395e-06, "loss": 0.3828, "step": 16300, "task_loss": 0.41226062178611755 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3562116026878357, "epoch": 13.78, "learning_rate": 4.520976831559174e-06, "loss": 0.3635, "step": 16301, "task_loss": 0.1951063722372055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19929440319538116, "epoch": 13.78, "learning_rate": 4.517845961177207e-06, "loss": 0.4476, "step": 16302, "task_loss": 0.3112215995788574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6966243386268616, "epoch": 13.78, "learning_rate": 4.514715090795242e-06, "loss": 0.4229, "step": 16303, "task_loss": 0.4202461242675781 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3780598044395447, "epoch": 13.78, "learning_rate": 4.511584220413275e-06, "loss": 0.4622, "step": 16304, "task_loss": 0.7207561731338501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6188704967498779, "epoch": 13.78, "learning_rate": 4.5084533500313085e-06, "loss": 0.4778, "step": 16305, "task_loss": 0.12382490187883377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32635563611984253, "epoch": 13.78, "learning_rate": 4.505322479649343e-06, "loss": 0.4311, "step": 16306, "task_loss": 0.5978676676750183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43723464012145996, "epoch": 13.78, "learning_rate": 4.502191609267377e-06, "loss": 0.47, "step": 16307, "task_loss": 0.18434610962867737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39524805545806885, "epoch": 13.78, "learning_rate": 4.499060738885411e-06, "loss": 0.4515, "step": 16308, "task_loss": 0.22699369490146637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.806010365486145, "epoch": 13.79, "learning_rate": 4.495929868503444e-06, "loss": 0.4183, "step": 16309, "task_loss": 0.7803768515586853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30192315578460693, "epoch": 13.79, "learning_rate": 4.492798998121478e-06, "loss": 0.3358, "step": 16310, "task_loss": 0.3026004135608673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2981984615325928, "epoch": 13.79, "learning_rate": 4.489668127739512e-06, "loss": 0.5249, "step": 16311, "task_loss": 0.877020537853241 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.296659916639328, "epoch": 13.79, "learning_rate": 4.486537257357545e-06, "loss": 0.3586, "step": 16312, "task_loss": 0.43806710839271545 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3488882780075073, "epoch": 13.79, "learning_rate": 4.48340638697558e-06, "loss": 0.6832, "step": 16313, "task_loss": 0.9376505017280579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5263966917991638, "epoch": 13.79, "learning_rate": 4.480275516593613e-06, "loss": 0.4924, "step": 16314, "task_loss": 1.0111852884292603 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.383121520280838, "epoch": 13.79, "learning_rate": 4.4771446462116475e-06, "loss": 0.4695, "step": 16315, "task_loss": 0.14831015467643738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36970895528793335, "epoch": 13.79, "learning_rate": 4.474013775829681e-06, "loss": 0.4383, "step": 16316, "task_loss": 1.0382665395736694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16455930471420288, "epoch": 13.79, "learning_rate": 4.4708829054477145e-06, "loss": 0.3554, "step": 16317, "task_loss": 0.090693399310112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15017689764499664, "epoch": 13.79, "learning_rate": 4.467752035065748e-06, "loss": 0.2764, "step": 16318, "task_loss": 0.4882739782333374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24622760713100433, "epoch": 13.79, "learning_rate": 4.464621164683782e-06, "loss": 0.4974, "step": 16319, "task_loss": 0.12966950237751007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3979654312133789, "epoch": 13.79, "learning_rate": 4.461490294301817e-06, "loss": 0.4912, "step": 16320, "task_loss": 0.5665307641029358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32580843567848206, "epoch": 13.8, "learning_rate": 4.45835942391985e-06, "loss": 0.4569, "step": 16321, "task_loss": 1.164804220199585 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34584277868270874, "epoch": 13.8, "learning_rate": 4.4552285535378836e-06, "loss": 0.3939, "step": 16322, "task_loss": 0.8228781819343567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42886191606521606, "epoch": 13.8, "learning_rate": 4.452097683155918e-06, "loss": 0.3349, "step": 16323, "task_loss": 0.558988094329834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3005717098712921, "epoch": 13.8, "learning_rate": 4.448966812773951e-06, "loss": 0.4388, "step": 16324, "task_loss": 0.7317487001419067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48223260045051575, "epoch": 13.8, "learning_rate": 4.445835942391985e-06, "loss": 0.5876, "step": 16325, "task_loss": 0.644068717956543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4997795522212982, "epoch": 13.8, "learning_rate": 4.442705072010019e-06, "loss": 0.5436, "step": 16326, "task_loss": 1.0641396045684814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5281206965446472, "epoch": 13.8, "learning_rate": 4.439574201628053e-06, "loss": 0.3874, "step": 16327, "task_loss": 0.5422359704971313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35814982652664185, "epoch": 13.8, "learning_rate": 4.436443331246087e-06, "loss": 0.3596, "step": 16328, "task_loss": 0.5393711924552917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4398115575313568, "epoch": 13.8, "learning_rate": 4.43331246086412e-06, "loss": 0.4524, "step": 16329, "task_loss": 0.9489027261734009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41511067748069763, "epoch": 13.8, "learning_rate": 4.430181590482154e-06, "loss": 0.4274, "step": 16330, "task_loss": 0.6533553600311279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38103121519088745, "epoch": 13.8, "learning_rate": 4.427050720100187e-06, "loss": 0.3419, "step": 16331, "task_loss": 0.39549365639686584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44513756036758423, "epoch": 13.81, "learning_rate": 4.4239198497182225e-06, "loss": 0.4781, "step": 16332, "task_loss": 0.6131216883659363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31854620575904846, "epoch": 13.81, "learning_rate": 4.420788979336256e-06, "loss": 0.4695, "step": 16333, "task_loss": 0.3867679238319397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3313009440898895, "epoch": 13.81, "learning_rate": 4.4176581089542895e-06, "loss": 0.5095, "step": 16334, "task_loss": 0.8542637825012207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32061439752578735, "epoch": 13.81, "learning_rate": 4.414527238572323e-06, "loss": 0.5139, "step": 16335, "task_loss": 0.9973844885826111 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31087014079093933, "epoch": 13.81, "learning_rate": 4.411396368190357e-06, "loss": 0.4257, "step": 16336, "task_loss": 0.6797540187835693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2234594076871872, "epoch": 13.81, "learning_rate": 4.408265497808391e-06, "loss": 0.3815, "step": 16337, "task_loss": 0.50302654504776 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34813979268074036, "epoch": 13.81, "learning_rate": 4.405134627426425e-06, "loss": 0.3422, "step": 16338, "task_loss": 0.25605860352516174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44020378589630127, "epoch": 13.81, "learning_rate": 4.4020037570444586e-06, "loss": 0.4061, "step": 16339, "task_loss": 0.9086729288101196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35023432970046997, "epoch": 13.81, "learning_rate": 4.398872886662492e-06, "loss": 0.3685, "step": 16340, "task_loss": 0.40507280826568604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27398842573165894, "epoch": 13.81, "learning_rate": 4.395742016280526e-06, "loss": 0.4892, "step": 16341, "task_loss": 0.3791852593421936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39471882581710815, "epoch": 13.81, "learning_rate": 4.39261114589856e-06, "loss": 0.4076, "step": 16342, "task_loss": 0.30691850185394287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44588950276374817, "epoch": 13.81, "learning_rate": 4.389480275516593e-06, "loss": 0.5092, "step": 16343, "task_loss": 1.1221429109573364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37715113162994385, "epoch": 13.82, "learning_rate": 4.386349405134628e-06, "loss": 0.3843, "step": 16344, "task_loss": 0.3244728147983551 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5009396076202393, "epoch": 13.82, "learning_rate": 4.383218534752662e-06, "loss": 0.4573, "step": 16345, "task_loss": 0.5720166563987732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40646758675575256, "epoch": 13.82, "learning_rate": 4.3800876643706954e-06, "loss": 0.4773, "step": 16346, "task_loss": 0.3576285243034363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5641888976097107, "epoch": 13.82, "learning_rate": 4.376956793988729e-06, "loss": 0.5419, "step": 16347, "task_loss": 0.5168100595474243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.213420569896698, "epoch": 13.82, "learning_rate": 4.373825923606762e-06, "loss": 0.379, "step": 16348, "task_loss": 1.0574709177017212 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3075288236141205, "epoch": 13.82, "learning_rate": 4.370695053224797e-06, "loss": 0.421, "step": 16349, "task_loss": 0.19175289571285248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4347866177558899, "epoch": 13.82, "learning_rate": 4.367564182842831e-06, "loss": 0.368, "step": 16350, "task_loss": 0.4078078269958496 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3799450695514679, "epoch": 13.82, "learning_rate": 4.3644333124608645e-06, "loss": 0.4782, "step": 16351, "task_loss": 0.25562557578086853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43248450756073, "epoch": 13.82, "learning_rate": 4.361302442078898e-06, "loss": 0.3809, "step": 16352, "task_loss": 1.5408353805541992 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6356050968170166, "epoch": 13.82, "learning_rate": 4.358171571696932e-06, "loss": 0.5664, "step": 16353, "task_loss": 1.0911341905593872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4800657331943512, "epoch": 13.82, "learning_rate": 4.355040701314966e-06, "loss": 0.4599, "step": 16354, "task_loss": 1.2980082035064697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5321740508079529, "epoch": 13.82, "learning_rate": 4.351909830932999e-06, "loss": 0.547, "step": 16355, "task_loss": 0.6969826221466064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2589637041091919, "epoch": 13.83, "learning_rate": 4.3487789605510336e-06, "loss": 0.5362, "step": 16356, "task_loss": 0.3361469507217407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26116102933883667, "epoch": 13.83, "learning_rate": 4.345648090169067e-06, "loss": 0.3537, "step": 16357, "task_loss": 0.366035521030426 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24583424627780914, "epoch": 13.83, "learning_rate": 4.342517219787101e-06, "loss": 0.3076, "step": 16358, "task_loss": 0.08042329549789429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2815743684768677, "epoch": 13.83, "learning_rate": 4.339386349405135e-06, "loss": 0.3462, "step": 16359, "task_loss": 0.17922364175319672 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3644930124282837, "epoch": 13.83, "learning_rate": 4.336255479023168e-06, "loss": 0.3812, "step": 16360, "task_loss": 0.4728233218193054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41069239377975464, "epoch": 13.83, "learning_rate": 4.333124608641203e-06, "loss": 0.449, "step": 16361, "task_loss": 0.5288141369819641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30040180683135986, "epoch": 13.83, "learning_rate": 4.329993738259236e-06, "loss": 0.4305, "step": 16362, "task_loss": 0.48470956087112427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38320159912109375, "epoch": 13.83, "learning_rate": 4.3268628678772704e-06, "loss": 0.4219, "step": 16363, "task_loss": 0.37542587518692017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5715799331665039, "epoch": 13.83, "learning_rate": 4.323731997495304e-06, "loss": 0.4601, "step": 16364, "task_loss": 0.7925469279289246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4784640669822693, "epoch": 13.83, "learning_rate": 4.320601127113337e-06, "loss": 0.3973, "step": 16365, "task_loss": 0.42777585983276367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31515640020370483, "epoch": 13.83, "learning_rate": 4.317470256731372e-06, "loss": 0.3928, "step": 16366, "task_loss": 0.448232501745224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4145112931728363, "epoch": 13.83, "learning_rate": 4.314339386349405e-06, "loss": 0.4225, "step": 16367, "task_loss": 1.0592350959777832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7358428239822388, "epoch": 13.84, "learning_rate": 4.311208515967439e-06, "loss": 0.4222, "step": 16368, "task_loss": 0.9035835266113281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.374377578496933, "epoch": 13.84, "learning_rate": 4.308077645585473e-06, "loss": 0.4742, "step": 16369, "task_loss": 0.7284926772117615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47864553332328796, "epoch": 13.84, "learning_rate": 4.304946775203507e-06, "loss": 0.4418, "step": 16370, "task_loss": 0.5733485221862793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3952482342720032, "epoch": 13.84, "learning_rate": 4.301815904821541e-06, "loss": 0.4138, "step": 16371, "task_loss": 0.40103816986083984 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4724307358264923, "epoch": 13.84, "learning_rate": 4.298685034439574e-06, "loss": 0.4537, "step": 16372, "task_loss": 1.5754636526107788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7828083038330078, "epoch": 13.84, "learning_rate": 4.295554164057608e-06, "loss": 0.5173, "step": 16373, "task_loss": 0.571670651435852 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20353356003761292, "epoch": 13.84, "learning_rate": 4.292423293675642e-06, "loss": 0.4332, "step": 16374, "task_loss": 0.15196658670902252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5156834721565247, "epoch": 13.84, "learning_rate": 4.289292423293676e-06, "loss": 0.5454, "step": 16375, "task_loss": 0.8081533312797546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37526577711105347, "epoch": 13.84, "learning_rate": 4.28616155291171e-06, "loss": 0.4387, "step": 16376, "task_loss": 0.07030580937862396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4734323024749756, "epoch": 13.84, "learning_rate": 4.283030682529743e-06, "loss": 0.4227, "step": 16377, "task_loss": 1.4422837495803833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49870431423187256, "epoch": 13.84, "learning_rate": 4.279899812147777e-06, "loss": 0.4738, "step": 16378, "task_loss": 0.4868346154689789 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2777520418167114, "epoch": 13.84, "learning_rate": 4.276768941765811e-06, "loss": 0.2963, "step": 16379, "task_loss": 0.607688307762146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5048487186431885, "epoch": 13.85, "learning_rate": 4.273638071383845e-06, "loss": 0.4523, "step": 16380, "task_loss": 0.7202870845794678 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2595664858818054, "epoch": 13.85, "learning_rate": 4.270507201001879e-06, "loss": 0.4645, "step": 16381, "task_loss": 1.3175216913223267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3094375729560852, "epoch": 13.85, "learning_rate": 4.267376330619912e-06, "loss": 0.3247, "step": 16382, "task_loss": 0.5183826088905334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45456463098526, "epoch": 13.85, "learning_rate": 4.264245460237947e-06, "loss": 0.5026, "step": 16383, "task_loss": 1.0426607131958008 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.279690682888031, "epoch": 13.85, "learning_rate": 4.26111458985598e-06, "loss": 0.4948, "step": 16384, "task_loss": 0.3233312964439392 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5191485285758972, "epoch": 13.85, "learning_rate": 4.257983719474014e-06, "loss": 0.4543, "step": 16385, "task_loss": 0.41420117020606995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5587066411972046, "epoch": 13.85, "learning_rate": 4.254852849092047e-06, "loss": 0.4378, "step": 16386, "task_loss": 0.5393715500831604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6819663047790527, "epoch": 13.85, "learning_rate": 4.251721978710082e-06, "loss": 0.513, "step": 16387, "task_loss": 1.1558575630187988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7216792106628418, "epoch": 13.85, "learning_rate": 4.248591108328116e-06, "loss": 0.5798, "step": 16388, "task_loss": 0.49170175194740295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.284203439950943, "epoch": 13.85, "learning_rate": 4.245460237946149e-06, "loss": 0.3856, "step": 16389, "task_loss": 0.15931914746761322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5674620866775513, "epoch": 13.85, "learning_rate": 4.242329367564183e-06, "loss": 0.5153, "step": 16390, "task_loss": 0.8125632405281067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46802452206611633, "epoch": 13.85, "learning_rate": 4.239198497182217e-06, "loss": 0.4984, "step": 16391, "task_loss": 0.5672407150268555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5708935856819153, "epoch": 13.86, "learning_rate": 4.2360676268002505e-06, "loss": 0.4774, "step": 16392, "task_loss": 1.192036509513855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5251849889755249, "epoch": 13.86, "learning_rate": 4.232936756418285e-06, "loss": 0.4463, "step": 16393, "task_loss": 1.0511521100997925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3022526800632477, "epoch": 13.86, "learning_rate": 4.229805886036318e-06, "loss": 0.4, "step": 16394, "task_loss": 0.08798215538263321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4452265501022339, "epoch": 13.86, "learning_rate": 4.226675015654352e-06, "loss": 0.3752, "step": 16395, "task_loss": 0.4679177403450012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36952778697013855, "epoch": 13.86, "learning_rate": 4.223544145272386e-06, "loss": 0.4353, "step": 16396, "task_loss": 0.44175487756729126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1903570592403412, "epoch": 13.86, "learning_rate": 4.22041327489042e-06, "loss": 0.3884, "step": 16397, "task_loss": 0.6545342206954956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5070984959602356, "epoch": 13.86, "learning_rate": 4.217282404508453e-06, "loss": 0.4656, "step": 16398, "task_loss": 1.0730788707733154 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4292685389518738, "epoch": 13.86, "learning_rate": 4.214151534126487e-06, "loss": 0.4392, "step": 16399, "task_loss": 0.34573793411254883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6943142414093018, "epoch": 13.86, "learning_rate": 4.211020663744522e-06, "loss": 0.6633, "step": 16400, "task_loss": 0.7107977271080017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41859251260757446, "epoch": 13.86, "learning_rate": 4.207889793362555e-06, "loss": 0.3518, "step": 16401, "task_loss": 1.1358773708343506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4092459976673126, "epoch": 13.86, "learning_rate": 4.204758922980589e-06, "loss": 0.4456, "step": 16402, "task_loss": 0.8761306405067444 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31295037269592285, "epoch": 13.87, "learning_rate": 4.201628052598622e-06, "loss": 0.3983, "step": 16403, "task_loss": 0.6453520655632019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34894609451293945, "epoch": 13.87, "learning_rate": 4.1984971822166565e-06, "loss": 0.3495, "step": 16404, "task_loss": 0.51539146900177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7710849046707153, "epoch": 13.87, "learning_rate": 4.195366311834691e-06, "loss": 0.5835, "step": 16405, "task_loss": 1.6490339040756226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5197281241416931, "epoch": 13.87, "learning_rate": 4.192235441452724e-06, "loss": 0.4688, "step": 16406, "task_loss": 0.4887680113315582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31838738918304443, "epoch": 13.87, "learning_rate": 4.189104571070758e-06, "loss": 0.3681, "step": 16407, "task_loss": 0.33342036604881287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38988184928894043, "epoch": 13.87, "learning_rate": 4.185973700688792e-06, "loss": 0.3805, "step": 16408, "task_loss": 0.5421009659767151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23000100255012512, "epoch": 13.87, "learning_rate": 4.1828428303068255e-06, "loss": 0.415, "step": 16409, "task_loss": 0.10358544439077377 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3967306613922119, "epoch": 13.87, "learning_rate": 4.179711959924859e-06, "loss": 0.4525, "step": 16410, "task_loss": 0.41531574726104736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3480069637298584, "epoch": 13.87, "learning_rate": 4.1765810895428925e-06, "loss": 0.3271, "step": 16411, "task_loss": 0.4528176188468933 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.594963550567627, "epoch": 13.87, "learning_rate": 4.173450219160927e-06, "loss": 0.5086, "step": 16412, "task_loss": 0.7242487072944641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3708975911140442, "epoch": 13.87, "learning_rate": 4.170319348778961e-06, "loss": 0.3852, "step": 16413, "task_loss": 0.37582170963287354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5560828447341919, "epoch": 13.87, "learning_rate": 4.167188478396995e-06, "loss": 0.4752, "step": 16414, "task_loss": 0.8141415119171143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2634339928627014, "epoch": 13.88, "learning_rate": 4.164057608015028e-06, "loss": 0.4742, "step": 16415, "task_loss": 0.02555954083800316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25132572650909424, "epoch": 13.88, "learning_rate": 4.1609267376330616e-06, "loss": 0.3879, "step": 16416, "task_loss": 0.8732253909111023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21217580139636993, "epoch": 13.88, "learning_rate": 4.157795867251096e-06, "loss": 0.3854, "step": 16417, "task_loss": 0.243172749876976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5211279392242432, "epoch": 13.88, "learning_rate": 4.15466499686913e-06, "loss": 0.4773, "step": 16418, "task_loss": 0.7412530183792114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3816518485546112, "epoch": 13.88, "learning_rate": 4.151534126487164e-06, "loss": 0.3567, "step": 16419, "task_loss": 0.5757445693016052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5031436085700989, "epoch": 13.88, "learning_rate": 4.148403256105197e-06, "loss": 0.3918, "step": 16420, "task_loss": 0.7691917419433594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4246860146522522, "epoch": 13.88, "learning_rate": 4.1452723857232315e-06, "loss": 0.4397, "step": 16421, "task_loss": 0.4163333773612976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26309505105018616, "epoch": 13.88, "learning_rate": 4.142141515341265e-06, "loss": 0.5483, "step": 16422, "task_loss": 0.4224494695663452 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4790506660938263, "epoch": 13.88, "learning_rate": 4.1390106449592984e-06, "loss": 0.3991, "step": 16423, "task_loss": 0.3746579885482788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42022189497947693, "epoch": 13.88, "learning_rate": 4.135879774577333e-06, "loss": 0.4041, "step": 16424, "task_loss": 0.4441361725330353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40214312076568604, "epoch": 13.88, "learning_rate": 4.132748904195367e-06, "loss": 0.5048, "step": 16425, "task_loss": 0.5997905135154724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26863881945610046, "epoch": 13.88, "learning_rate": 4.1296180338134006e-06, "loss": 0.3389, "step": 16426, "task_loss": 0.037727512419223785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6993637084960938, "epoch": 13.89, "learning_rate": 4.126487163431434e-06, "loss": 0.4959, "step": 16427, "task_loss": 0.7847298383712769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48367661237716675, "epoch": 13.89, "learning_rate": 4.1233562930494675e-06, "loss": 0.4845, "step": 16428, "task_loss": 0.6269140839576721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45781227946281433, "epoch": 13.89, "learning_rate": 4.120225422667502e-06, "loss": 0.479, "step": 16429, "task_loss": 0.33854055404663086 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6184714436531067, "epoch": 13.89, "learning_rate": 4.117094552285536e-06, "loss": 0.4974, "step": 16430, "task_loss": 0.8755799531936646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2719985842704773, "epoch": 13.89, "learning_rate": 4.11396368190357e-06, "loss": 0.418, "step": 16431, "task_loss": 0.46740278601646423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22278369963169098, "epoch": 13.89, "learning_rate": 4.110832811521603e-06, "loss": 0.4172, "step": 16432, "task_loss": 0.4937216341495514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5941270589828491, "epoch": 13.89, "learning_rate": 4.107701941139637e-06, "loss": 0.4397, "step": 16433, "task_loss": 0.8595921397209167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29090672731399536, "epoch": 13.89, "learning_rate": 4.104571070757671e-06, "loss": 0.4364, "step": 16434, "task_loss": 0.16140012443065643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29076820611953735, "epoch": 13.89, "learning_rate": 4.101440200375704e-06, "loss": 0.368, "step": 16435, "task_loss": 0.34101539850234985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.666029691696167, "epoch": 13.89, "learning_rate": 4.098309329993739e-06, "loss": 0.5121, "step": 16436, "task_loss": 0.9235875010490417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.518000602722168, "epoch": 13.89, "learning_rate": 4.095178459611772e-06, "loss": 0.4492, "step": 16437, "task_loss": 0.5727716684341431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5692358613014221, "epoch": 13.89, "learning_rate": 4.0920475892298065e-06, "loss": 0.4103, "step": 16438, "task_loss": 0.6539928913116455 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3381466269493103, "epoch": 13.9, "learning_rate": 4.08891671884784e-06, "loss": 0.3204, "step": 16439, "task_loss": 0.07513851672410965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29208195209503174, "epoch": 13.9, "learning_rate": 4.0857858484658734e-06, "loss": 0.344, "step": 16440, "task_loss": 0.20675575733184814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2953278720378876, "epoch": 13.9, "learning_rate": 4.082654978083907e-06, "loss": 0.4478, "step": 16441, "task_loss": 0.3635328710079193 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6570925712585449, "epoch": 13.9, "learning_rate": 4.079524107701941e-06, "loss": 0.3777, "step": 16442, "task_loss": 0.7093407511711121 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23915383219718933, "epoch": 13.9, "learning_rate": 4.0763932373199756e-06, "loss": 0.3568, "step": 16443, "task_loss": 0.30426958203315735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36913424730300903, "epoch": 13.9, "learning_rate": 4.073262366938009e-06, "loss": 0.4566, "step": 16444, "task_loss": 0.4190140664577484 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2664641737937927, "epoch": 13.9, "learning_rate": 4.0701314965560425e-06, "loss": 0.3012, "step": 16445, "task_loss": 0.07489842921495438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35399144887924194, "epoch": 13.9, "learning_rate": 4.067000626174077e-06, "loss": 0.4122, "step": 16446, "task_loss": 0.8306644558906555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44773802161216736, "epoch": 13.9, "learning_rate": 4.06386975579211e-06, "loss": 0.433, "step": 16447, "task_loss": 1.1180543899536133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42489832639694214, "epoch": 13.9, "learning_rate": 4.060738885410145e-06, "loss": 0.4244, "step": 16448, "task_loss": 0.36702316999435425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22299036383628845, "epoch": 13.9, "learning_rate": 4.057608015028178e-06, "loss": 0.3843, "step": 16449, "task_loss": 0.3584781885147095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.572941780090332, "epoch": 13.9, "learning_rate": 4.054477144646212e-06, "loss": 0.5273, "step": 16450, "task_loss": 0.9274557828903198 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2975853383541107, "epoch": 13.91, "learning_rate": 4.051346274264246e-06, "loss": 0.3699, "step": 16451, "task_loss": 0.563959538936615 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27018871903419495, "epoch": 13.91, "learning_rate": 4.048215403882279e-06, "loss": 0.3597, "step": 16452, "task_loss": 0.7062305808067322 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5194599628448486, "epoch": 13.91, "learning_rate": 4.045084533500313e-06, "loss": 0.379, "step": 16453, "task_loss": 0.45913466811180115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5736343264579773, "epoch": 13.91, "learning_rate": 4.041953663118346e-06, "loss": 0.4053, "step": 16454, "task_loss": 1.0039979219436646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38389283418655396, "epoch": 13.91, "learning_rate": 4.0388227927363815e-06, "loss": 0.3554, "step": 16455, "task_loss": 0.06669417768716812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6706545352935791, "epoch": 13.91, "learning_rate": 4.035691922354415e-06, "loss": 0.4858, "step": 16456, "task_loss": 0.9371312856674194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30056464672088623, "epoch": 13.91, "learning_rate": 4.0325610519724485e-06, "loss": 0.4105, "step": 16457, "task_loss": 0.17415913939476013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33330732583999634, "epoch": 13.91, "learning_rate": 4.029430181590482e-06, "loss": 0.4222, "step": 16458, "task_loss": 0.4060627222061157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32789161801338196, "epoch": 13.91, "learning_rate": 4.026299311208516e-06, "loss": 0.3559, "step": 16459, "task_loss": 0.31982916593551636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3982716202735901, "epoch": 13.91, "learning_rate": 4.02316844082655e-06, "loss": 0.395, "step": 16460, "task_loss": 0.6655773520469666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27375710010528564, "epoch": 13.91, "learning_rate": 4.020037570444584e-06, "loss": 0.4235, "step": 16461, "task_loss": 0.7664551138877869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36737364530563354, "epoch": 13.91, "learning_rate": 4.0169067000626175e-06, "loss": 0.3351, "step": 16462, "task_loss": 0.27944961190223694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4980902075767517, "epoch": 13.92, "learning_rate": 4.013775829680652e-06, "loss": 0.4748, "step": 16463, "task_loss": 0.5740262866020203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2667354643344879, "epoch": 13.92, "learning_rate": 4.010644959298685e-06, "loss": 0.4886, "step": 16464, "task_loss": 0.7052434682846069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3154122531414032, "epoch": 13.92, "learning_rate": 4.007514088916719e-06, "loss": 0.4562, "step": 16465, "task_loss": 0.5833793878555298 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5064323544502258, "epoch": 13.92, "learning_rate": 4.004383218534752e-06, "loss": 0.5355, "step": 16466, "task_loss": 0.660574734210968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8409949541091919, "epoch": 13.92, "learning_rate": 4.001252348152787e-06, "loss": 0.5747, "step": 16467, "task_loss": 1.1016498804092407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3607318103313446, "epoch": 13.92, "learning_rate": 3.998121477770821e-06, "loss": 0.4592, "step": 16468, "task_loss": 0.06994148343801498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5086803436279297, "epoch": 13.92, "learning_rate": 3.994990607388854e-06, "loss": 0.3887, "step": 16469, "task_loss": 0.6707781553268433 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40276798605918884, "epoch": 13.92, "learning_rate": 3.991859737006888e-06, "loss": 0.3785, "step": 16470, "task_loss": 1.6727322340011597 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4985622763633728, "epoch": 13.92, "learning_rate": 3.988728866624921e-06, "loss": 0.4575, "step": 16471, "task_loss": 0.059138067066669464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5977975130081177, "epoch": 13.92, "learning_rate": 3.985597996242956e-06, "loss": 0.4677, "step": 16472, "task_loss": 0.7886539101600647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37768712639808655, "epoch": 13.92, "learning_rate": 3.98246712586099e-06, "loss": 0.4553, "step": 16473, "task_loss": 1.3670716285705566 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3611014485359192, "epoch": 13.93, "learning_rate": 3.9793362554790235e-06, "loss": 0.3443, "step": 16474, "task_loss": 0.5837692022323608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5468282699584961, "epoch": 13.93, "learning_rate": 3.976205385097057e-06, "loss": 0.5484, "step": 16475, "task_loss": 0.4883539080619812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42070135474205017, "epoch": 13.93, "learning_rate": 3.973074514715091e-06, "loss": 0.3951, "step": 16476, "task_loss": 0.08386661857366562 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5144878625869751, "epoch": 13.93, "learning_rate": 3.969943644333125e-06, "loss": 0.4239, "step": 16477, "task_loss": 0.39162734150886536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5920753479003906, "epoch": 13.93, "learning_rate": 3.966812773951158e-06, "loss": 0.5283, "step": 16478, "task_loss": 1.3184375762939453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7936543226242065, "epoch": 13.93, "learning_rate": 3.9636819035691925e-06, "loss": 0.4602, "step": 16479, "task_loss": 0.2932255268096924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5187294483184814, "epoch": 13.93, "learning_rate": 3.960551033187226e-06, "loss": 0.3306, "step": 16480, "task_loss": 1.1131318807601929 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18763676285743713, "epoch": 13.93, "learning_rate": 3.95742016280526e-06, "loss": 0.3094, "step": 16481, "task_loss": 0.5515356659889221 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3793075680732727, "epoch": 13.93, "learning_rate": 3.954289292423294e-06, "loss": 0.3369, "step": 16482, "task_loss": 0.27310505509376526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42201298475265503, "epoch": 13.93, "learning_rate": 3.951158422041327e-06, "loss": 0.486, "step": 16483, "task_loss": 0.674584150314331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4883400797843933, "epoch": 13.93, "learning_rate": 3.948027551659362e-06, "loss": 0.5081, "step": 16484, "task_loss": 1.082183837890625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2836020886898041, "epoch": 13.93, "learning_rate": 3.944896681277396e-06, "loss": 0.3302, "step": 16485, "task_loss": 0.4353885352611542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3587155342102051, "epoch": 13.94, "learning_rate": 3.941765810895429e-06, "loss": 0.4612, "step": 16486, "task_loss": 0.3200996518135071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3966309726238251, "epoch": 13.94, "learning_rate": 3.938634940513463e-06, "loss": 0.5157, "step": 16487, "task_loss": 0.4173533022403717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6567266583442688, "epoch": 13.94, "learning_rate": 3.935504070131496e-06, "loss": 0.4928, "step": 16488, "task_loss": 1.1690829992294312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.13989828526973724, "epoch": 13.94, "learning_rate": 3.932373199749531e-06, "loss": 0.3816, "step": 16489, "task_loss": 0.030914224684238434 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5096784234046936, "epoch": 13.94, "learning_rate": 3.929242329367564e-06, "loss": 0.3761, "step": 16490, "task_loss": 1.2862626314163208 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37493646144866943, "epoch": 13.94, "learning_rate": 3.9261114589855985e-06, "loss": 0.3835, "step": 16491, "task_loss": 1.4589217901229858 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5190935134887695, "epoch": 13.94, "learning_rate": 3.922980588603632e-06, "loss": 0.4367, "step": 16492, "task_loss": 0.9683103561401367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34295403957366943, "epoch": 13.94, "learning_rate": 3.919849718221666e-06, "loss": 0.4409, "step": 16493, "task_loss": 0.6002787947654724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47701555490493774, "epoch": 13.94, "learning_rate": 3.9167188478397e-06, "loss": 0.4094, "step": 16494, "task_loss": 0.26211801171302795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4160897135734558, "epoch": 13.94, "learning_rate": 3.913587977457733e-06, "loss": 0.4071, "step": 16495, "task_loss": 0.5472652912139893 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2972728908061981, "epoch": 13.94, "learning_rate": 3.910457107075767e-06, "loss": 0.4027, "step": 16496, "task_loss": 0.3433762192726135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5884740948677063, "epoch": 13.94, "learning_rate": 3.907326236693801e-06, "loss": 0.3662, "step": 16497, "task_loss": 0.9965288639068604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3567471206188202, "epoch": 13.95, "learning_rate": 3.904195366311835e-06, "loss": 0.4291, "step": 16498, "task_loss": 0.7490228414535522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2886922359466553, "epoch": 13.95, "learning_rate": 3.901064495929869e-06, "loss": 0.4088, "step": 16499, "task_loss": 0.5343332886695862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5600884556770325, "epoch": 13.95, "learning_rate": 3.897933625547902e-06, "loss": 0.5472, "step": 16500, "task_loss": 0.36970171332359314 }, { "epoch": 13.95, "eval_accuracy": 0.9134257425742575, "eval_loss": 0.303165465593338, "eval_runtime": 206.1169, "eval_samples_per_second": 122.503, "eval_steps_per_second": 0.961, "step": 16500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3510187268257141, "epoch": 13.95, "learning_rate": 3.894802755165937e-06, "loss": 0.3965, "step": 16501, "task_loss": 0.5386896729469299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2904305160045624, "epoch": 13.95, "learning_rate": 3.89167188478397e-06, "loss": 0.3947, "step": 16502, "task_loss": 0.35616934299468994 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5954498052597046, "epoch": 13.95, "learning_rate": 3.8885410144020036e-06, "loss": 0.4583, "step": 16503, "task_loss": 0.86827152967453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.565775990486145, "epoch": 13.95, "learning_rate": 3.885410144020038e-06, "loss": 0.4416, "step": 16504, "task_loss": 0.5518581867218018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5285365581512451, "epoch": 13.95, "learning_rate": 3.882279273638071e-06, "loss": 0.6178, "step": 16505, "task_loss": 1.21390700340271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5828710198402405, "epoch": 13.95, "learning_rate": 3.879148403256106e-06, "loss": 0.486, "step": 16506, "task_loss": 0.9648109078407288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26787102222442627, "epoch": 13.95, "learning_rate": 3.876017532874139e-06, "loss": 0.3052, "step": 16507, "task_loss": 0.3682158291339874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33363527059555054, "epoch": 13.95, "learning_rate": 3.872886662492173e-06, "loss": 0.4577, "step": 16508, "task_loss": 0.3041831851005554 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31259801983833313, "epoch": 13.95, "learning_rate": 3.869755792110206e-06, "loss": 0.4708, "step": 16509, "task_loss": 0.28872066736221313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5248394012451172, "epoch": 13.96, "learning_rate": 3.866624921728241e-06, "loss": 0.3653, "step": 16510, "task_loss": 0.30287033319473267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38704460859298706, "epoch": 13.96, "learning_rate": 3.863494051346275e-06, "loss": 0.34, "step": 16511, "task_loss": 0.2172536849975586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2280166745185852, "epoch": 13.96, "learning_rate": 3.860363180964308e-06, "loss": 0.3367, "step": 16512, "task_loss": 0.6053559184074402 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5263766646385193, "epoch": 13.96, "learning_rate": 3.857232310582342e-06, "loss": 0.5527, "step": 16513, "task_loss": 0.7904815673828125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.513543426990509, "epoch": 13.96, "learning_rate": 3.854101440200376e-06, "loss": 0.4037, "step": 16514, "task_loss": 0.12018617987632751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40749165415763855, "epoch": 13.96, "learning_rate": 3.8509705698184095e-06, "loss": 0.4392, "step": 16515, "task_loss": 0.588556706905365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4315241575241089, "epoch": 13.96, "learning_rate": 3.847839699436444e-06, "loss": 0.3632, "step": 16516, "task_loss": 1.120820164680481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48848769068717957, "epoch": 13.96, "learning_rate": 3.844708829054477e-06, "loss": 0.4751, "step": 16517, "task_loss": 0.2652299702167511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3446184992790222, "epoch": 13.96, "learning_rate": 3.841577958672511e-06, "loss": 0.3804, "step": 16518, "task_loss": 0.5100160241127014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46127164363861084, "epoch": 13.96, "learning_rate": 3.838447088290545e-06, "loss": 0.4278, "step": 16519, "task_loss": 0.5272716879844666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5187241435050964, "epoch": 13.96, "learning_rate": 3.8353162179085786e-06, "loss": 0.4835, "step": 16520, "task_loss": 0.987303614616394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46849220991134644, "epoch": 13.96, "learning_rate": 3.832185347526612e-06, "loss": 0.6267, "step": 16521, "task_loss": 0.5234523415565491 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5844963788986206, "epoch": 13.97, "learning_rate": 3.829054477144646e-06, "loss": 0.4931, "step": 16522, "task_loss": 0.5933325886726379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5102123022079468, "epoch": 13.97, "learning_rate": 3.825923606762681e-06, "loss": 0.5696, "step": 16523, "task_loss": 0.5118111371994019 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33932697772979736, "epoch": 13.97, "learning_rate": 3.822792736380714e-06, "loss": 0.4579, "step": 16524, "task_loss": 0.6113328337669373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4034292995929718, "epoch": 13.97, "learning_rate": 3.819661865998748e-06, "loss": 0.4254, "step": 16525, "task_loss": 0.462442547082901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6585119962692261, "epoch": 13.97, "learning_rate": 3.816530995616781e-06, "loss": 0.5524, "step": 16526, "task_loss": 0.983272910118103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5623289346694946, "epoch": 13.97, "learning_rate": 3.813400125234815e-06, "loss": 0.4588, "step": 16527, "task_loss": 1.0010493993759155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42672839760780334, "epoch": 13.97, "learning_rate": 3.8102692548528498e-06, "loss": 0.5814, "step": 16528, "task_loss": 1.2609426975250244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3716003894805908, "epoch": 13.97, "learning_rate": 3.8071383844708832e-06, "loss": 0.4841, "step": 16529, "task_loss": 0.9383043050765991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42990797758102417, "epoch": 13.97, "learning_rate": 3.804007514088917e-06, "loss": 0.6128, "step": 16530, "task_loss": 0.13007552921772003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21334561705589294, "epoch": 13.97, "learning_rate": 3.8008766437069506e-06, "loss": 0.2809, "step": 16531, "task_loss": 0.7723675966262817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38261154294013977, "epoch": 13.97, "learning_rate": 3.7977457733249845e-06, "loss": 0.3647, "step": 16532, "task_loss": 0.5644190311431885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25132542848587036, "epoch": 13.97, "learning_rate": 3.794614902943018e-06, "loss": 0.3815, "step": 16533, "task_loss": 1.2010051012039185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6533140540122986, "epoch": 13.98, "learning_rate": 3.7914840325610523e-06, "loss": 0.4724, "step": 16534, "task_loss": 0.3099665641784668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21690885722637177, "epoch": 13.98, "learning_rate": 3.788353162179086e-06, "loss": 0.4445, "step": 16535, "task_loss": 0.17909343540668488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3880302309989929, "epoch": 13.98, "learning_rate": 3.7852222917971197e-06, "loss": 0.5118, "step": 16536, "task_loss": 0.3290344476699829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4691469669342041, "epoch": 13.98, "learning_rate": 3.7820914214151536e-06, "loss": 0.5483, "step": 16537, "task_loss": 0.17622515559196472 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4048526883125305, "epoch": 13.98, "learning_rate": 3.7789605510331875e-06, "loss": 0.4387, "step": 16538, "task_loss": 0.78774094581604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30605828762054443, "epoch": 13.98, "learning_rate": 3.775829680651221e-06, "loss": 0.3393, "step": 16539, "task_loss": 0.6667405366897583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.534171462059021, "epoch": 13.98, "learning_rate": 3.7726988102692553e-06, "loss": 0.4061, "step": 16540, "task_loss": 0.1649617850780487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3081711530685425, "epoch": 13.98, "learning_rate": 3.769567939887289e-06, "loss": 0.4681, "step": 16541, "task_loss": 0.1653508096933365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5681308507919312, "epoch": 13.98, "learning_rate": 3.7664370695053227e-06, "loss": 0.6457, "step": 16542, "task_loss": 1.4154589176177979 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5178776383399963, "epoch": 13.98, "learning_rate": 3.7633061991233566e-06, "loss": 0.3898, "step": 16543, "task_loss": 0.7426899075508118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4773732125759125, "epoch": 13.98, "learning_rate": 3.76017532874139e-06, "loss": 0.3904, "step": 16544, "task_loss": 0.5238459706306458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3057432770729065, "epoch": 13.99, "learning_rate": 3.757044458359424e-06, "loss": 0.3086, "step": 16545, "task_loss": 0.8700815439224243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3759247064590454, "epoch": 13.99, "learning_rate": 3.7539135879774574e-06, "loss": 0.396, "step": 16546, "task_loss": 0.310603529214859 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3285183012485504, "epoch": 13.99, "learning_rate": 3.750782717595492e-06, "loss": 0.3324, "step": 16547, "task_loss": 0.4550637900829315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5848722457885742, "epoch": 13.99, "learning_rate": 3.7476518472135256e-06, "loss": 0.4044, "step": 16548, "task_loss": 1.561119556427002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.11638437211513519, "epoch": 13.99, "learning_rate": 3.7445209768315595e-06, "loss": 0.3512, "step": 16549, "task_loss": 0.4469057321548462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2752820551395416, "epoch": 13.99, "learning_rate": 3.741390106449593e-06, "loss": 0.3577, "step": 16550, "task_loss": 0.1615479290485382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5525320172309875, "epoch": 13.99, "learning_rate": 3.738259236067627e-06, "loss": 0.6001, "step": 16551, "task_loss": 0.6760501265525818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5011394023895264, "epoch": 13.99, "learning_rate": 3.7351283656856604e-06, "loss": 0.3946, "step": 16552, "task_loss": 0.35647785663604736 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2375747114419937, "epoch": 13.99, "learning_rate": 3.7319974953036947e-06, "loss": 0.4729, "step": 16553, "task_loss": 0.26266565918922424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42811957001686096, "epoch": 13.99, "learning_rate": 3.7288666249217286e-06, "loss": 0.4135, "step": 16554, "task_loss": 0.1813298910856247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4091813266277313, "epoch": 13.99, "learning_rate": 3.725735754539762e-06, "loss": 0.4324, "step": 16555, "task_loss": 1.5524423122406006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4683910608291626, "epoch": 13.99, "learning_rate": 3.722604884157796e-06, "loss": 0.5354, "step": 16556, "task_loss": 1.3643735647201538 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22099092602729797, "epoch": 14.0, "learning_rate": 3.71947401377583e-06, "loss": 0.4098, "step": 16557, "task_loss": 0.510031521320343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3890569806098938, "epoch": 14.0, "learning_rate": 3.7163431433938633e-06, "loss": 0.3508, "step": 16558, "task_loss": 0.6692176461219788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2878648638725281, "epoch": 14.0, "learning_rate": 3.7132122730118977e-06, "loss": 0.3767, "step": 16559, "task_loss": 0.22526881098747253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4476144313812256, "epoch": 14.0, "learning_rate": 3.7100814026299316e-06, "loss": 0.4211, "step": 16560, "task_loss": 0.7798353433609009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3899519443511963, "epoch": 14.0, "learning_rate": 3.706950532247965e-06, "loss": 0.4344, "step": 16561, "task_loss": 0.14298665523529053 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39945894479751587, "epoch": 14.0, "learning_rate": 3.703819661865999e-06, "loss": 0.4087, "step": 16562, "task_loss": 0.9455318450927734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5649905204772949, "epoch": 14.0, "learning_rate": 3.7006887914840324e-06, "loss": 0.8504, "step": 16563, "task_loss": 0.7508898377418518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3512151539325714, "epoch": 14.0, "learning_rate": 3.6975579211020663e-06, "loss": 0.4178, "step": 16564, "task_loss": 0.359850138425827 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4854937791824341, "epoch": 14.0, "learning_rate": 3.6944270507201006e-06, "loss": 0.5599, "step": 16565, "task_loss": 0.458648681640625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3373410701751709, "epoch": 14.0, "learning_rate": 3.6912961803381345e-06, "loss": 0.4326, "step": 16566, "task_loss": 0.42862969636917114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.446760892868042, "epoch": 14.0, "learning_rate": 3.688165309956168e-06, "loss": 0.3931, "step": 16567, "task_loss": 0.24998527765274048 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4060378670692444, "epoch": 14.01, "learning_rate": 3.685034439574202e-06, "loss": 0.4226, "step": 16568, "task_loss": 0.5132647156715393 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5328637361526489, "epoch": 14.01, "learning_rate": 3.6819035691922354e-06, "loss": 0.4598, "step": 16569, "task_loss": 0.7107442617416382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3708736300468445, "epoch": 14.01, "learning_rate": 3.6787726988102693e-06, "loss": 0.4289, "step": 16570, "task_loss": 0.13718156516551971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19045455753803253, "epoch": 14.01, "learning_rate": 3.6756418284283036e-06, "loss": 0.3399, "step": 16571, "task_loss": 0.26186665892601013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38661885261535645, "epoch": 14.01, "learning_rate": 3.672510958046337e-06, "loss": 0.4418, "step": 16572, "task_loss": 0.09336303919553757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46085309982299805, "epoch": 14.01, "learning_rate": 3.669380087664371e-06, "loss": 0.4912, "step": 16573, "task_loss": 0.6248362064361572 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39125993847846985, "epoch": 14.01, "learning_rate": 3.6662492172824045e-06, "loss": 0.4129, "step": 16574, "task_loss": 0.3576470613479614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35053446888923645, "epoch": 14.01, "learning_rate": 3.6631183469004384e-06, "loss": 0.4245, "step": 16575, "task_loss": 0.6774932146072388 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37130701541900635, "epoch": 14.01, "learning_rate": 3.6599874765184723e-06, "loss": 0.3231, "step": 16576, "task_loss": 1.2671937942504883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3145853877067566, "epoch": 14.01, "learning_rate": 3.6568566061365066e-06, "loss": 0.336, "step": 16577, "task_loss": 0.3732157349586487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5715894103050232, "epoch": 14.01, "learning_rate": 3.65372573575454e-06, "loss": 0.4397, "step": 16578, "task_loss": 1.2308374643325806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3797394037246704, "epoch": 14.01, "learning_rate": 3.650594865372574e-06, "loss": 0.4008, "step": 16579, "task_loss": 0.6664953827857971 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36662253737449646, "epoch": 14.02, "learning_rate": 3.6474639949906074e-06, "loss": 0.3229, "step": 16580, "task_loss": 0.36640363931655884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8118773698806763, "epoch": 14.02, "learning_rate": 3.6443331246086413e-06, "loss": 0.5229, "step": 16581, "task_loss": 0.9511626362800598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.350055456161499, "epoch": 14.02, "learning_rate": 3.641202254226675e-06, "loss": 0.4366, "step": 16582, "task_loss": 0.8448593020439148 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49182185530662537, "epoch": 14.02, "learning_rate": 3.638071383844709e-06, "loss": 0.421, "step": 16583, "task_loss": 0.5719929933547974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32098549604415894, "epoch": 14.02, "learning_rate": 3.634940513462743e-06, "loss": 0.4233, "step": 16584, "task_loss": 0.2322402149438858 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6126773357391357, "epoch": 14.02, "learning_rate": 3.631809643080777e-06, "loss": 0.4716, "step": 16585, "task_loss": 1.228800892829895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48058149218559265, "epoch": 14.02, "learning_rate": 3.6286787726988104e-06, "loss": 0.4471, "step": 16586, "task_loss": 1.1898515224456787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41392695903778076, "epoch": 14.02, "learning_rate": 3.6255479023168443e-06, "loss": 0.431, "step": 16587, "task_loss": 0.305397629737854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4891948997974396, "epoch": 14.02, "learning_rate": 3.6224170319348778e-06, "loss": 0.4421, "step": 16588, "task_loss": 0.16767841577529907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2990739941596985, "epoch": 14.02, "learning_rate": 3.619286161552912e-06, "loss": 0.4174, "step": 16589, "task_loss": 0.08373275399208069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2723061144351959, "epoch": 14.02, "learning_rate": 3.616155291170946e-06, "loss": 0.4133, "step": 16590, "task_loss": 0.450976699590683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35926249623298645, "epoch": 14.02, "learning_rate": 3.6130244207889795e-06, "loss": 0.4545, "step": 16591, "task_loss": 0.15174129605293274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6667042970657349, "epoch": 14.03, "learning_rate": 3.6098935504070134e-06, "loss": 0.6059, "step": 16592, "task_loss": 0.07348207384347916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2947467565536499, "epoch": 14.03, "learning_rate": 3.606762680025047e-06, "loss": 0.4536, "step": 16593, "task_loss": 0.9610814452171326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.15723592042922974, "epoch": 14.03, "learning_rate": 3.6036318096430807e-06, "loss": 0.321, "step": 16594, "task_loss": 0.039662010967731476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45034438371658325, "epoch": 14.03, "learning_rate": 3.6005009392611146e-06, "loss": 0.3961, "step": 16595, "task_loss": 0.5883641839027405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38538676500320435, "epoch": 14.03, "learning_rate": 3.597370068879149e-06, "loss": 0.4214, "step": 16596, "task_loss": 0.07230599969625473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1545482873916626, "epoch": 14.03, "learning_rate": 3.5942391984971824e-06, "loss": 0.3553, "step": 16597, "task_loss": 0.12961719930171967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44013333320617676, "epoch": 14.03, "learning_rate": 3.5911083281152163e-06, "loss": 0.53, "step": 16598, "task_loss": 0.883078396320343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34822484850883484, "epoch": 14.03, "learning_rate": 3.58797745773325e-06, "loss": 0.42, "step": 16599, "task_loss": 0.18502286076545715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3347959518432617, "epoch": 14.03, "learning_rate": 3.5848465873512837e-06, "loss": 0.4468, "step": 16600, "task_loss": 0.7294427752494812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3928414583206177, "epoch": 14.03, "learning_rate": 3.581715716969317e-06, "loss": 0.4402, "step": 16601, "task_loss": 0.9170352220535278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6119402050971985, "epoch": 14.03, "learning_rate": 3.578584846587352e-06, "loss": 0.54, "step": 16602, "task_loss": 0.89034104347229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36253711581230164, "epoch": 14.03, "learning_rate": 3.5754539762053854e-06, "loss": 0.3215, "step": 16603, "task_loss": 0.7398703098297119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8320568203926086, "epoch": 14.04, "learning_rate": 3.5723231058234193e-06, "loss": 0.5245, "step": 16604, "task_loss": 2.339993953704834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40631651878356934, "epoch": 14.04, "learning_rate": 3.5691922354414528e-06, "loss": 0.4659, "step": 16605, "task_loss": 0.5000026226043701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5401800274848938, "epoch": 14.04, "learning_rate": 3.5660613650594867e-06, "loss": 0.5105, "step": 16606, "task_loss": 1.213660478591919 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3401307165622711, "epoch": 14.04, "learning_rate": 3.56293049467752e-06, "loss": 0.4441, "step": 16607, "task_loss": 0.5574635863304138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6184453964233398, "epoch": 14.04, "learning_rate": 3.5597996242955545e-06, "loss": 0.5188, "step": 16608, "task_loss": 1.7591652870178223 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37417489290237427, "epoch": 14.04, "learning_rate": 3.5566687539135884e-06, "loss": 0.439, "step": 16609, "task_loss": 0.8494524955749512 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44436246156692505, "epoch": 14.04, "learning_rate": 3.553537883531622e-06, "loss": 0.3633, "step": 16610, "task_loss": 0.16162975132465363 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3073500394821167, "epoch": 14.04, "learning_rate": 3.5504070131496557e-06, "loss": 0.434, "step": 16611, "task_loss": 0.36215853691101074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31966811418533325, "epoch": 14.04, "learning_rate": 3.5472761427676892e-06, "loss": 0.4148, "step": 16612, "task_loss": 0.7491605281829834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4083715081214905, "epoch": 14.04, "learning_rate": 3.544145272385723e-06, "loss": 0.4635, "step": 16613, "task_loss": 1.0630096197128296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28911328315734863, "epoch": 14.04, "learning_rate": 3.5410144020037574e-06, "loss": 0.4562, "step": 16614, "task_loss": 0.3713776171207428 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2602614164352417, "epoch": 14.04, "learning_rate": 3.5378835316217913e-06, "loss": 0.3296, "step": 16615, "task_loss": 0.9033509492874146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31547701358795166, "epoch": 14.05, "learning_rate": 3.534752661239825e-06, "loss": 0.3698, "step": 16616, "task_loss": 0.4864380955696106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37545210123062134, "epoch": 14.05, "learning_rate": 3.5316217908578587e-06, "loss": 0.5186, "step": 16617, "task_loss": 0.21208709478378296 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44650042057037354, "epoch": 14.05, "learning_rate": 3.528490920475892e-06, "loss": 0.3965, "step": 16618, "task_loss": 0.2483750730752945 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36038994789123535, "epoch": 14.05, "learning_rate": 3.525360050093926e-06, "loss": 0.4026, "step": 16619, "task_loss": 0.10904853045940399 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5164042711257935, "epoch": 14.05, "learning_rate": 3.5222291797119604e-06, "loss": 0.3632, "step": 16620, "task_loss": 0.6200109720230103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6653776168823242, "epoch": 14.05, "learning_rate": 3.5190983093299943e-06, "loss": 0.5283, "step": 16621, "task_loss": 1.219690203666687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5635845065116882, "epoch": 14.05, "learning_rate": 3.5159674389480278e-06, "loss": 0.4382, "step": 16622, "task_loss": 0.525129497051239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2994413375854492, "epoch": 14.05, "learning_rate": 3.5128365685660617e-06, "loss": 0.4216, "step": 16623, "task_loss": 1.1369531154632568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3216892182826996, "epoch": 14.05, "learning_rate": 3.509705698184095e-06, "loss": 0.4349, "step": 16624, "task_loss": 0.08935043960809708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29794955253601074, "epoch": 14.05, "learning_rate": 3.506574827802129e-06, "loss": 0.4382, "step": 16625, "task_loss": 0.4666251242160797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23371165990829468, "epoch": 14.05, "learning_rate": 3.5034439574201634e-06, "loss": 0.4245, "step": 16626, "task_loss": 0.33100077509880066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.351892352104187, "epoch": 14.05, "learning_rate": 3.500313087038197e-06, "loss": 0.3932, "step": 16627, "task_loss": 0.5006498098373413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41794055700302124, "epoch": 14.06, "learning_rate": 3.4971822166562308e-06, "loss": 0.4432, "step": 16628, "task_loss": 0.47702282667160034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27873730659484863, "epoch": 14.06, "learning_rate": 3.4940513462742642e-06, "loss": 0.4239, "step": 16629, "task_loss": 0.4012645483016968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3918468952178955, "epoch": 14.06, "learning_rate": 3.490920475892298e-06, "loss": 0.4396, "step": 16630, "task_loss": 0.2825097441673279 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7786844372749329, "epoch": 14.06, "learning_rate": 3.4877896055103316e-06, "loss": 0.6531, "step": 16631, "task_loss": 1.358902096748352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34678030014038086, "epoch": 14.06, "learning_rate": 3.4846587351283663e-06, "loss": 0.4423, "step": 16632, "task_loss": 0.2599804997444153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41035202145576477, "epoch": 14.06, "learning_rate": 3.4815278647464e-06, "loss": 0.4615, "step": 16633, "task_loss": 0.3048070967197418 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33393394947052, "epoch": 14.06, "learning_rate": 3.4783969943644337e-06, "loss": 0.4279, "step": 16634, "task_loss": 0.5837187170982361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32838720083236694, "epoch": 14.06, "learning_rate": 3.475266123982467e-06, "loss": 0.4051, "step": 16635, "task_loss": 0.05288040265440941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3729170560836792, "epoch": 14.06, "learning_rate": 3.472135253600501e-06, "loss": 0.4788, "step": 16636, "task_loss": 1.1828525066375732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25883448123931885, "epoch": 14.06, "learning_rate": 3.4690043832185346e-06, "loss": 0.3533, "step": 16637, "task_loss": 0.4162878096103668 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5855430364608765, "epoch": 14.06, "learning_rate": 3.4658735128365685e-06, "loss": 0.4556, "step": 16638, "task_loss": 0.22995835542678833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44196832180023193, "epoch": 14.07, "learning_rate": 3.462742642454603e-06, "loss": 0.3432, "step": 16639, "task_loss": 0.641635000705719 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3725370764732361, "epoch": 14.07, "learning_rate": 3.4596117720726367e-06, "loss": 0.3961, "step": 16640, "task_loss": 0.736556887626648 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3294290006160736, "epoch": 14.07, "learning_rate": 3.45648090169067e-06, "loss": 0.5334, "step": 16641, "task_loss": 0.4173990488052368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34436678886413574, "epoch": 14.07, "learning_rate": 3.453350031308704e-06, "loss": 0.4876, "step": 16642, "task_loss": 0.890171468257904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3581418991088867, "epoch": 14.07, "learning_rate": 3.4502191609267375e-06, "loss": 0.4376, "step": 16643, "task_loss": 0.9581742882728577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2894332706928253, "epoch": 14.07, "learning_rate": 3.4470882905447714e-06, "loss": 0.3766, "step": 16644, "task_loss": 1.108647108078003 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.12074895203113556, "epoch": 14.07, "learning_rate": 3.4439574201628058e-06, "loss": 0.3936, "step": 16645, "task_loss": 0.4089471995830536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27118566632270813, "epoch": 14.07, "learning_rate": 3.4408265497808392e-06, "loss": 0.3392, "step": 16646, "task_loss": 0.18898004293441772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32885342836380005, "epoch": 14.07, "learning_rate": 3.437695679398873e-06, "loss": 0.3627, "step": 16647, "task_loss": 0.7766087651252747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5416033267974854, "epoch": 14.07, "learning_rate": 3.4345648090169066e-06, "loss": 0.398, "step": 16648, "task_loss": 0.49229249358177185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31389716267585754, "epoch": 14.07, "learning_rate": 3.4314339386349405e-06, "loss": 0.3666, "step": 16649, "task_loss": 0.2705325484275818 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4094359576702118, "epoch": 14.07, "learning_rate": 3.428303068252974e-06, "loss": 0.5005, "step": 16650, "task_loss": 1.0979394912719727 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5237434506416321, "epoch": 14.08, "learning_rate": 3.4251721978710087e-06, "loss": 0.432, "step": 16651, "task_loss": 1.4004660844802856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5491586923599243, "epoch": 14.08, "learning_rate": 3.422041327489042e-06, "loss": 0.5394, "step": 16652, "task_loss": 0.7955853939056396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3890582025051117, "epoch": 14.08, "learning_rate": 3.418910457107076e-06, "loss": 0.4479, "step": 16653, "task_loss": 0.9333439469337463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7368963956832886, "epoch": 14.08, "learning_rate": 3.4157795867251096e-06, "loss": 0.348, "step": 16654, "task_loss": 1.130610704421997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6728568077087402, "epoch": 14.08, "learning_rate": 3.4126487163431435e-06, "loss": 0.4429, "step": 16655, "task_loss": 1.0777437686920166 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.331524133682251, "epoch": 14.08, "learning_rate": 3.409517845961177e-06, "loss": 0.391, "step": 16656, "task_loss": 1.0006953477859497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3774237036705017, "epoch": 14.08, "learning_rate": 3.4063869755792113e-06, "loss": 0.3603, "step": 16657, "task_loss": 0.5431119799613953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.424228310585022, "epoch": 14.08, "learning_rate": 3.403256105197245e-06, "loss": 0.4686, "step": 16658, "task_loss": 0.31099969148635864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47960782051086426, "epoch": 14.08, "learning_rate": 3.400125234815279e-06, "loss": 0.4864, "step": 16659, "task_loss": 0.4054742157459259 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6612704992294312, "epoch": 14.08, "learning_rate": 3.3969943644333125e-06, "loss": 0.535, "step": 16660, "task_loss": 1.1338762044906616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5819136500358582, "epoch": 14.08, "learning_rate": 3.3938634940513464e-06, "loss": 0.3382, "step": 16661, "task_loss": 0.40725505352020264 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5189341902732849, "epoch": 14.08, "learning_rate": 3.39073262366938e-06, "loss": 0.4257, "step": 16662, "task_loss": 0.5260877013206482 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6165244579315186, "epoch": 14.09, "learning_rate": 3.3876017532874142e-06, "loss": 0.3786, "step": 16663, "task_loss": 0.8416500091552734 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.530254602432251, "epoch": 14.09, "learning_rate": 3.384470882905448e-06, "loss": 0.5454, "step": 16664, "task_loss": 0.4529018700122833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29551127552986145, "epoch": 14.09, "learning_rate": 3.3813400125234816e-06, "loss": 0.3325, "step": 16665, "task_loss": 0.10524212568998337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27661803364753723, "epoch": 14.09, "learning_rate": 3.3782091421415155e-06, "loss": 0.4631, "step": 16666, "task_loss": 0.44427767395973206 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37128740549087524, "epoch": 14.09, "learning_rate": 3.375078271759549e-06, "loss": 0.4703, "step": 16667, "task_loss": 0.43555766344070435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8361345529556274, "epoch": 14.09, "learning_rate": 3.371947401377583e-06, "loss": 0.5759, "step": 16668, "task_loss": 1.0943480730056763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21265538036823273, "epoch": 14.09, "learning_rate": 3.3688165309956172e-06, "loss": 0.2762, "step": 16669, "task_loss": 0.26338082551956177 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3816104829311371, "epoch": 14.09, "learning_rate": 3.365685660613651e-06, "loss": 0.4801, "step": 16670, "task_loss": 0.26315563917160034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2475188821554184, "epoch": 14.09, "learning_rate": 3.3625547902316846e-06, "loss": 0.4376, "step": 16671, "task_loss": 0.8411150574684143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.506128191947937, "epoch": 14.09, "learning_rate": 3.3594239198497185e-06, "loss": 0.4643, "step": 16672, "task_loss": 1.3404780626296997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3440450429916382, "epoch": 14.09, "learning_rate": 3.356293049467752e-06, "loss": 0.4014, "step": 16673, "task_loss": 0.464986115694046 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4350292682647705, "epoch": 14.09, "learning_rate": 3.353162179085786e-06, "loss": 0.5536, "step": 16674, "task_loss": 0.4171915650367737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3717454671859741, "epoch": 14.1, "learning_rate": 3.35003130870382e-06, "loss": 0.392, "step": 16675, "task_loss": 0.41278350353240967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44366833567619324, "epoch": 14.1, "learning_rate": 3.3469004383218537e-06, "loss": 0.433, "step": 16676, "task_loss": 0.5942737460136414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2817080318927765, "epoch": 14.1, "learning_rate": 3.3437695679398876e-06, "loss": 0.3832, "step": 16677, "task_loss": 0.6162261962890625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4779600501060486, "epoch": 14.1, "learning_rate": 3.3406386975579215e-06, "loss": 0.4874, "step": 16678, "task_loss": 0.845024824142456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3418300151824951, "epoch": 14.1, "learning_rate": 3.337507827175955e-06, "loss": 0.4323, "step": 16679, "task_loss": 1.2598177194595337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3183099627494812, "epoch": 14.1, "learning_rate": 3.334376956793989e-06, "loss": 0.4356, "step": 16680, "task_loss": 0.24696582555770874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5110447406768799, "epoch": 14.1, "learning_rate": 3.331246086412023e-06, "loss": 0.4355, "step": 16681, "task_loss": 1.3576562404632568 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22315004467964172, "epoch": 14.1, "learning_rate": 3.3281152160300566e-06, "loss": 0.4199, "step": 16682, "task_loss": 0.07477252930402756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4650924801826477, "epoch": 14.1, "learning_rate": 3.3249843456480905e-06, "loss": 0.4853, "step": 16683, "task_loss": 1.2031347751617432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44078773260116577, "epoch": 14.1, "learning_rate": 3.321853475266124e-06, "loss": 0.5242, "step": 16684, "task_loss": 1.0352847576141357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5508905053138733, "epoch": 14.1, "learning_rate": 3.318722604884158e-06, "loss": 0.4348, "step": 16685, "task_loss": 0.3995194733142853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4097152054309845, "epoch": 14.1, "learning_rate": 3.3155917345021914e-06, "loss": 0.4495, "step": 16686, "task_loss": 1.0202393531799316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5102778673171997, "epoch": 14.11, "learning_rate": 3.3124608641202253e-06, "loss": 0.463, "step": 16687, "task_loss": 0.15414008498191833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6029987335205078, "epoch": 14.11, "learning_rate": 3.3093299937382596e-06, "loss": 0.5421, "step": 16688, "task_loss": 0.2968686521053314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45070236921310425, "epoch": 14.11, "learning_rate": 3.3061991233562935e-06, "loss": 0.3496, "step": 16689, "task_loss": 0.6418491005897522 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2973948121070862, "epoch": 14.11, "learning_rate": 3.303068252974327e-06, "loss": 0.3682, "step": 16690, "task_loss": 0.5368298292160034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29722297191619873, "epoch": 14.11, "learning_rate": 3.299937382592361e-06, "loss": 0.3355, "step": 16691, "task_loss": 0.7067745923995972 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5541505813598633, "epoch": 14.11, "learning_rate": 3.2968065122103943e-06, "loss": 0.4253, "step": 16692, "task_loss": 1.1107912063598633 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4288604259490967, "epoch": 14.11, "learning_rate": 3.2936756418284282e-06, "loss": 0.4526, "step": 16693, "task_loss": 0.25848525762557983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47651001811027527, "epoch": 14.11, "learning_rate": 3.2905447714464626e-06, "loss": 0.4195, "step": 16694, "task_loss": 0.37270018458366394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3540933132171631, "epoch": 14.11, "learning_rate": 3.287413901064496e-06, "loss": 0.476, "step": 16695, "task_loss": 1.5523518323898315 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6425036191940308, "epoch": 14.11, "learning_rate": 3.28428303068253e-06, "loss": 0.4721, "step": 16696, "task_loss": 0.9533817768096924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4575101137161255, "epoch": 14.11, "learning_rate": 3.281152160300564e-06, "loss": 0.4398, "step": 16697, "task_loss": 0.2596123516559601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37098291516304016, "epoch": 14.11, "learning_rate": 3.2780212899185973e-06, "loss": 0.3638, "step": 16698, "task_loss": 0.25670933723449707 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6039205193519592, "epoch": 14.12, "learning_rate": 3.2748904195366312e-06, "loss": 0.3754, "step": 16699, "task_loss": 0.9416992664337158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5350862741470337, "epoch": 14.12, "learning_rate": 3.2717595491546655e-06, "loss": 0.4499, "step": 16700, "task_loss": 0.8137832283973694 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5729490518569946, "epoch": 14.12, "learning_rate": 3.268628678772699e-06, "loss": 0.4649, "step": 16701, "task_loss": 0.6728530526161194 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.759671151638031, "epoch": 14.12, "learning_rate": 3.265497808390733e-06, "loss": 0.5294, "step": 16702, "task_loss": 0.3539232313632965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3813536465167999, "epoch": 14.12, "learning_rate": 3.2623669380087664e-06, "loss": 0.4642, "step": 16703, "task_loss": 1.1901003122329712 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4491240978240967, "epoch": 14.12, "learning_rate": 3.2592360676268003e-06, "loss": 0.483, "step": 16704, "task_loss": 0.8476523756980896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3718731999397278, "epoch": 14.12, "learning_rate": 3.2561051972448338e-06, "loss": 0.3414, "step": 16705, "task_loss": 0.6647421717643738 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4109448790550232, "epoch": 14.12, "learning_rate": 3.2529743268628685e-06, "loss": 0.4175, "step": 16706, "task_loss": 0.41014549136161804 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4031458795070648, "epoch": 14.12, "learning_rate": 3.249843456480902e-06, "loss": 0.377, "step": 16707, "task_loss": 0.6447562575340271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4080791175365448, "epoch": 14.12, "learning_rate": 3.246712586098936e-06, "loss": 0.3189, "step": 16708, "task_loss": 0.22831475734710693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5118553638458252, "epoch": 14.12, "learning_rate": 3.2435817157169694e-06, "loss": 0.3617, "step": 16709, "task_loss": 0.19311045110225677 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23838666081428528, "epoch": 14.13, "learning_rate": 3.2404508453350033e-06, "loss": 0.5287, "step": 16710, "task_loss": 0.4069368243217468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33573824167251587, "epoch": 14.13, "learning_rate": 3.2373199749530367e-06, "loss": 0.3114, "step": 16711, "task_loss": 0.41208821535110474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37269479036331177, "epoch": 14.13, "learning_rate": 3.234189104571071e-06, "loss": 0.5015, "step": 16712, "task_loss": 0.6381068825721741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47478315234184265, "epoch": 14.13, "learning_rate": 3.231058234189105e-06, "loss": 0.4454, "step": 16713, "task_loss": 0.4108310639858246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6401984095573425, "epoch": 14.13, "learning_rate": 3.2279273638071384e-06, "loss": 0.4554, "step": 16714, "task_loss": 0.4813010096549988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20808406174182892, "epoch": 14.13, "learning_rate": 3.2247964934251723e-06, "loss": 0.3632, "step": 16715, "task_loss": 0.04833192750811577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27220702171325684, "epoch": 14.13, "learning_rate": 3.2216656230432062e-06, "loss": 0.3548, "step": 16716, "task_loss": 0.183022141456604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4017159044742584, "epoch": 14.13, "learning_rate": 3.2185347526612397e-06, "loss": 0.4298, "step": 16717, "task_loss": 0.22892899811267853 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1925036609172821, "epoch": 14.13, "learning_rate": 3.215403882279274e-06, "loss": 0.392, "step": 16718, "task_loss": 0.6011527180671692 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25135478377342224, "epoch": 14.13, "learning_rate": 3.212273011897308e-06, "loss": 0.4261, "step": 16719, "task_loss": 0.6498263478279114 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39042648673057556, "epoch": 14.13, "learning_rate": 3.2091421415153414e-06, "loss": 0.3249, "step": 16720, "task_loss": 0.5287468433380127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4795578420162201, "epoch": 14.13, "learning_rate": 3.2060112711333753e-06, "loss": 0.4927, "step": 16721, "task_loss": 0.4578231871128082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6027464270591736, "epoch": 14.14, "learning_rate": 3.2028804007514088e-06, "loss": 0.4807, "step": 16722, "task_loss": 1.0429420471191406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2549973130226135, "epoch": 14.14, "learning_rate": 3.1997495303694427e-06, "loss": 0.315, "step": 16723, "task_loss": 0.8940004706382751 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3379175066947937, "epoch": 14.14, "learning_rate": 3.196618659987477e-06, "loss": 0.2883, "step": 16724, "task_loss": 0.7091282606124878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4868648648262024, "epoch": 14.14, "learning_rate": 3.193487789605511e-06, "loss": 0.4765, "step": 16725, "task_loss": 0.9675384163856506 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4753872752189636, "epoch": 14.14, "learning_rate": 3.1903569192235444e-06, "loss": 0.3752, "step": 16726, "task_loss": 0.18052351474761963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4290948510169983, "epoch": 14.14, "learning_rate": 3.1872260488415783e-06, "loss": 0.2995, "step": 16727, "task_loss": 0.1252918690443039 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2899719774723053, "epoch": 14.14, "learning_rate": 3.1840951784596117e-06, "loss": 0.4234, "step": 16728, "task_loss": 0.24042710661888123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3868565559387207, "epoch": 14.14, "learning_rate": 3.1809643080776456e-06, "loss": 0.4791, "step": 16729, "task_loss": 0.23231185972690582 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30586880445480347, "epoch": 14.14, "learning_rate": 3.177833437695679e-06, "loss": 0.3613, "step": 16730, "task_loss": 0.2420385479927063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33693718910217285, "epoch": 14.14, "learning_rate": 3.1747025673137134e-06, "loss": 0.3466, "step": 16731, "task_loss": 0.5448337197303772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3863060772418976, "epoch": 14.14, "learning_rate": 3.1715716969317473e-06, "loss": 0.4569, "step": 16732, "task_loss": 0.6408237218856812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2766968905925751, "epoch": 14.14, "learning_rate": 3.168440826549781e-06, "loss": 0.385, "step": 16733, "task_loss": 0.21599163115024567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3304632902145386, "epoch": 14.15, "learning_rate": 3.1653099561678147e-06, "loss": 0.4899, "step": 16734, "task_loss": 0.4432154893875122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38324683904647827, "epoch": 14.15, "learning_rate": 3.1621790857858486e-06, "loss": 0.4282, "step": 16735, "task_loss": 1.0474220514297485 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29569655656814575, "epoch": 14.15, "learning_rate": 3.159048215403882e-06, "loss": 0.3481, "step": 16736, "task_loss": 0.6255205273628235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.610668420791626, "epoch": 14.15, "learning_rate": 3.1559173450219164e-06, "loss": 0.4365, "step": 16737, "task_loss": 0.27214300632476807 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44577112793922424, "epoch": 14.15, "learning_rate": 3.1527864746399503e-06, "loss": 0.6539, "step": 16738, "task_loss": 0.3203915059566498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5434094071388245, "epoch": 14.15, "learning_rate": 3.1496556042579838e-06, "loss": 0.4456, "step": 16739, "task_loss": 0.7115404605865479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2543061375617981, "epoch": 14.15, "learning_rate": 3.1465247338760177e-06, "loss": 0.4458, "step": 16740, "task_loss": 0.051698874682188034 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18559342622756958, "epoch": 14.15, "learning_rate": 3.143393863494051e-06, "loss": 0.4082, "step": 16741, "task_loss": 0.43008047342300415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1842956244945526, "epoch": 14.15, "learning_rate": 3.140262993112085e-06, "loss": 0.3718, "step": 16742, "task_loss": 0.09390833228826523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2516264319419861, "epoch": 14.15, "learning_rate": 3.1371321227301194e-06, "loss": 0.4097, "step": 16743, "task_loss": 0.6436243653297424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17793160676956177, "epoch": 14.15, "learning_rate": 3.1340012523481533e-06, "loss": 0.3575, "step": 16744, "task_loss": 0.20801731944084167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2504723072052002, "epoch": 14.15, "learning_rate": 3.1308703819661867e-06, "loss": 0.3369, "step": 16745, "task_loss": 0.3262992799282074 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4681759476661682, "epoch": 14.16, "learning_rate": 3.1277395115842206e-06, "loss": 0.5047, "step": 16746, "task_loss": 0.5157175064086914 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26004043221473694, "epoch": 14.16, "learning_rate": 3.1246086412022545e-06, "loss": 0.4007, "step": 16747, "task_loss": 0.1870332509279251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37750813364982605, "epoch": 14.16, "learning_rate": 3.1214777708202884e-06, "loss": 0.468, "step": 16748, "task_loss": 0.644827663898468 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20619633793830872, "epoch": 14.16, "learning_rate": 3.118346900438322e-06, "loss": 0.5623, "step": 16749, "task_loss": 0.2467750757932663 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23240235447883606, "epoch": 14.16, "learning_rate": 3.115216030056356e-06, "loss": 0.2574, "step": 16750, "task_loss": 0.31303632259368896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3519785702228546, "epoch": 14.16, "learning_rate": 3.1120851596743897e-06, "loss": 0.3828, "step": 16751, "task_loss": 0.5880220532417297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43124425411224365, "epoch": 14.16, "learning_rate": 3.108954289292423e-06, "loss": 0.4281, "step": 16752, "task_loss": 0.48011505603790283 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22745707631111145, "epoch": 14.16, "learning_rate": 3.105823418910457e-06, "loss": 0.4509, "step": 16753, "task_loss": 0.1687062382698059 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19888520240783691, "epoch": 14.16, "learning_rate": 3.102692548528491e-06, "loss": 0.3551, "step": 16754, "task_loss": 0.29573020339012146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4134036898612976, "epoch": 14.16, "learning_rate": 3.099561678146525e-06, "loss": 0.364, "step": 16755, "task_loss": 0.35861414670944214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29533708095550537, "epoch": 14.16, "learning_rate": 3.0964308077645584e-06, "loss": 0.4775, "step": 16756, "task_loss": 0.9049550890922546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2561722993850708, "epoch": 14.16, "learning_rate": 3.0932999373825927e-06, "loss": 0.4119, "step": 16757, "task_loss": 1.323047399520874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.478391170501709, "epoch": 14.17, "learning_rate": 3.090169067000626e-06, "loss": 0.368, "step": 16758, "task_loss": 0.4114754796028137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5475700497627258, "epoch": 14.17, "learning_rate": 3.08703819661866e-06, "loss": 0.4381, "step": 16759, "task_loss": 0.22574780881404877 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28307443857192993, "epoch": 14.17, "learning_rate": 3.083907326236694e-06, "loss": 0.4249, "step": 16760, "task_loss": 0.772895336151123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5513944625854492, "epoch": 14.17, "learning_rate": 3.080776455854728e-06, "loss": 0.3734, "step": 16761, "task_loss": 0.1971167027950287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4763180911540985, "epoch": 14.17, "learning_rate": 3.0776455854727613e-06, "loss": 0.4366, "step": 16762, "task_loss": 0.2293403595685959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3988131880760193, "epoch": 14.17, "learning_rate": 3.0745147150907957e-06, "loss": 0.4182, "step": 16763, "task_loss": 0.2211586833000183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31873440742492676, "epoch": 14.17, "learning_rate": 3.071383844708829e-06, "loss": 0.4506, "step": 16764, "task_loss": 0.3412633240222931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40190982818603516, "epoch": 14.17, "learning_rate": 3.068252974326863e-06, "loss": 0.3931, "step": 16765, "task_loss": 1.0317198038101196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43424710631370544, "epoch": 14.17, "learning_rate": 3.065122103944897e-06, "loss": 0.4253, "step": 16766, "task_loss": 1.0506731271743774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37289243936538696, "epoch": 14.17, "learning_rate": 3.061991233562931e-06, "loss": 0.555, "step": 16767, "task_loss": 0.4150778651237488 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37710604071617126, "epoch": 14.17, "learning_rate": 3.0588603631809643e-06, "loss": 0.5766, "step": 16768, "task_loss": 1.0937278270721436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27137109637260437, "epoch": 14.17, "learning_rate": 3.055729492798998e-06, "loss": 0.3489, "step": 16769, "task_loss": 0.8881827592849731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.296378493309021, "epoch": 14.18, "learning_rate": 3.052598622417032e-06, "loss": 0.4357, "step": 16770, "task_loss": 0.6819801330566406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34909719228744507, "epoch": 14.18, "learning_rate": 3.0494677520350656e-06, "loss": 0.4682, "step": 16771, "task_loss": 0.3444351553916931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5647115707397461, "epoch": 14.18, "learning_rate": 3.0463368816531e-06, "loss": 0.4425, "step": 16772, "task_loss": 1.3047960996627808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17982083559036255, "epoch": 14.18, "learning_rate": 3.0432060112711334e-06, "loss": 0.2491, "step": 16773, "task_loss": 0.038089267909526825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5155758857727051, "epoch": 14.18, "learning_rate": 3.0400751408891673e-06, "loss": 0.3567, "step": 16774, "task_loss": 1.4581151008605957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4634135961532593, "epoch": 14.18, "learning_rate": 3.036944270507201e-06, "loss": 0.5091, "step": 16775, "task_loss": 0.3943212330341339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5898382067680359, "epoch": 14.18, "learning_rate": 3.033813400125235e-06, "loss": 0.4626, "step": 16776, "task_loss": 0.9010851979255676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24743294715881348, "epoch": 14.18, "learning_rate": 3.0306825297432685e-06, "loss": 0.4038, "step": 16777, "task_loss": 0.41517046093940735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3121030032634735, "epoch": 14.18, "learning_rate": 3.027551659361303e-06, "loss": 0.3696, "step": 16778, "task_loss": 0.303835391998291 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33394214510917664, "epoch": 14.18, "learning_rate": 3.0244207889793363e-06, "loss": 0.4454, "step": 16779, "task_loss": 1.0437358617782593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30445945262908936, "epoch": 14.18, "learning_rate": 3.0212899185973702e-06, "loss": 0.4116, "step": 16780, "task_loss": 0.8258463144302368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28568026423454285, "epoch": 14.19, "learning_rate": 3.018159048215404e-06, "loss": 0.4285, "step": 16781, "task_loss": 0.4263773262500763 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3190816640853882, "epoch": 14.19, "learning_rate": 3.015028177833438e-06, "loss": 0.3626, "step": 16782, "task_loss": 0.2758204936981201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33364254236221313, "epoch": 14.19, "learning_rate": 3.0118973074514715e-06, "loss": 0.5025, "step": 16783, "task_loss": 0.5587963461875916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3371988534927368, "epoch": 14.19, "learning_rate": 3.0087664370695054e-06, "loss": 0.3446, "step": 16784, "task_loss": 0.16406987607479095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3068675100803375, "epoch": 14.19, "learning_rate": 3.0056355666875393e-06, "loss": 0.3882, "step": 16785, "task_loss": 0.5557737350463867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1928108036518097, "epoch": 14.19, "learning_rate": 3.002504696305573e-06, "loss": 0.3029, "step": 16786, "task_loss": 0.26465144753456116 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4728195071220398, "epoch": 14.19, "learning_rate": 2.999373825923607e-06, "loss": 0.366, "step": 16787, "task_loss": 0.7223328351974487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6131080985069275, "epoch": 14.19, "learning_rate": 2.9962429555416406e-06, "loss": 0.4911, "step": 16788, "task_loss": 1.1513460874557495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.254392147064209, "epoch": 14.19, "learning_rate": 2.9931120851596745e-06, "loss": 0.3586, "step": 16789, "task_loss": 0.09545312076807022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28153884410858154, "epoch": 14.19, "learning_rate": 2.9899812147777084e-06, "loss": 0.3547, "step": 16790, "task_loss": 1.0480576753616333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39610928297042847, "epoch": 14.19, "learning_rate": 2.9868503443957423e-06, "loss": 0.3562, "step": 16791, "task_loss": 0.6943846940994263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34122443199157715, "epoch": 14.19, "learning_rate": 2.9837194740137758e-06, "loss": 0.4235, "step": 16792, "task_loss": 0.313327819108963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5018352270126343, "epoch": 14.2, "learning_rate": 2.98058860363181e-06, "loss": 0.5112, "step": 16793, "task_loss": 0.9160400629043579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22860071063041687, "epoch": 14.2, "learning_rate": 2.9774577332498436e-06, "loss": 0.4538, "step": 16794, "task_loss": 0.6695484519004822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35037317872047424, "epoch": 14.2, "learning_rate": 2.9743268628678775e-06, "loss": 0.4349, "step": 16795, "task_loss": 0.5359184145927429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17649665474891663, "epoch": 14.2, "learning_rate": 2.9711959924859114e-06, "loss": 0.3086, "step": 16796, "task_loss": 0.05295177549123764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5803205966949463, "epoch": 14.2, "learning_rate": 2.9680651221039453e-06, "loss": 0.3934, "step": 16797, "task_loss": 0.6233041882514954 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45203152298927307, "epoch": 14.2, "learning_rate": 2.9649342517219787e-06, "loss": 0.531, "step": 16798, "task_loss": 0.7007414102554321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4238870441913605, "epoch": 14.2, "learning_rate": 2.9618033813400126e-06, "loss": 0.3832, "step": 16799, "task_loss": 0.6086696982383728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31659209728240967, "epoch": 14.2, "learning_rate": 2.9586725109580465e-06, "loss": 0.4209, "step": 16800, "task_loss": 0.5097706317901611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2423413246870041, "epoch": 14.2, "learning_rate": 2.9555416405760804e-06, "loss": 0.3823, "step": 16801, "task_loss": 0.08062314987182617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3674079179763794, "epoch": 14.2, "learning_rate": 2.952410770194114e-06, "loss": 0.5086, "step": 16802, "task_loss": 0.11048687994480133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48295778036117554, "epoch": 14.2, "learning_rate": 2.949279899812148e-06, "loss": 0.4877, "step": 16803, "task_loss": 0.40487203001976013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5174657106399536, "epoch": 14.2, "learning_rate": 2.9461490294301817e-06, "loss": 0.4482, "step": 16804, "task_loss": 0.35262587666511536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6155754327774048, "epoch": 14.21, "learning_rate": 2.9430181590482156e-06, "loss": 0.5237, "step": 16805, "task_loss": 0.5806806087493896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43616509437561035, "epoch": 14.21, "learning_rate": 2.9398872886662495e-06, "loss": 0.4291, "step": 16806, "task_loss": 0.5118290185928345 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27535125613212585, "epoch": 14.21, "learning_rate": 2.936756418284283e-06, "loss": 0.4704, "step": 16807, "task_loss": 0.14014218747615814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6773364543914795, "epoch": 14.21, "learning_rate": 2.933625547902317e-06, "loss": 0.502, "step": 16808, "task_loss": 1.076874017715454 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19959509372711182, "epoch": 14.21, "learning_rate": 2.9304946775203508e-06, "loss": 0.3132, "step": 16809, "task_loss": 0.7073841691017151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3998625576496124, "epoch": 14.21, "learning_rate": 2.9273638071383847e-06, "loss": 0.4874, "step": 16810, "task_loss": 0.6735832691192627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3473345637321472, "epoch": 14.21, "learning_rate": 2.924232936756418e-06, "loss": 0.4259, "step": 16811, "task_loss": 1.1879584789276123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26906806230545044, "epoch": 14.21, "learning_rate": 2.9211020663744525e-06, "loss": 0.39, "step": 16812, "task_loss": 1.556945562362671 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2862372398376465, "epoch": 14.21, "learning_rate": 2.917971195992486e-06, "loss": 0.327, "step": 16813, "task_loss": 0.3970416486263275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.339463472366333, "epoch": 14.21, "learning_rate": 2.91484032561052e-06, "loss": 0.428, "step": 16814, "task_loss": 0.4205631613731384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3750857710838318, "epoch": 14.21, "learning_rate": 2.9117094552285537e-06, "loss": 0.3406, "step": 16815, "task_loss": 0.6083822846412659 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6159992218017578, "epoch": 14.21, "learning_rate": 2.9085785848465876e-06, "loss": 0.4075, "step": 16816, "task_loss": 0.7434397339820862 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5176740884780884, "epoch": 14.22, "learning_rate": 2.905447714464621e-06, "loss": 0.4866, "step": 16817, "task_loss": 0.26146024465560913 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4890359044075012, "epoch": 14.22, "learning_rate": 2.9023168440826554e-06, "loss": 0.4595, "step": 16818, "task_loss": 0.8235184550285339 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5972027778625488, "epoch": 14.22, "learning_rate": 2.899185973700689e-06, "loss": 0.3969, "step": 16819, "task_loss": 0.2290230691432953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6192071437835693, "epoch": 14.22, "learning_rate": 2.896055103318723e-06, "loss": 0.3621, "step": 16820, "task_loss": 0.4904252290725708 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2917618155479431, "epoch": 14.22, "learning_rate": 2.8929242329367567e-06, "loss": 0.3166, "step": 16821, "task_loss": 0.20978212356567383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32286420464515686, "epoch": 14.22, "learning_rate": 2.88979336255479e-06, "loss": 0.4103, "step": 16822, "task_loss": 0.4527497887611389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2589210867881775, "epoch": 14.22, "learning_rate": 2.886662492172824e-06, "loss": 0.3247, "step": 16823, "task_loss": 0.05127372592687607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4971885085105896, "epoch": 14.22, "learning_rate": 2.883531621790858e-06, "loss": 0.3918, "step": 16824, "task_loss": 0.9979428052902222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5150735974311829, "epoch": 14.22, "learning_rate": 2.880400751408892e-06, "loss": 0.3564, "step": 16825, "task_loss": 0.7046076059341431 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5191441774368286, "epoch": 14.22, "learning_rate": 2.8772698810269254e-06, "loss": 0.5678, "step": 16826, "task_loss": 1.2800184488296509 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2315463125705719, "epoch": 14.22, "learning_rate": 2.8741390106449597e-06, "loss": 0.4287, "step": 16827, "task_loss": 0.4074268937110901 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6254975199699402, "epoch": 14.22, "learning_rate": 2.871008140262993e-06, "loss": 0.4608, "step": 16828, "task_loss": 0.5892733335494995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38801679015159607, "epoch": 14.23, "learning_rate": 2.867877269881027e-06, "loss": 0.264, "step": 16829, "task_loss": 0.4810626208782196 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3277725577354431, "epoch": 14.23, "learning_rate": 2.864746399499061e-06, "loss": 0.5092, "step": 16830, "task_loss": 0.6325255632400513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40789395570755005, "epoch": 14.23, "learning_rate": 2.861615529117095e-06, "loss": 0.4768, "step": 16831, "task_loss": 0.6437469124794006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3851545453071594, "epoch": 14.23, "learning_rate": 2.8584846587351283e-06, "loss": 0.4544, "step": 16832, "task_loss": 0.6051374673843384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5259397029876709, "epoch": 14.23, "learning_rate": 2.8553537883531626e-06, "loss": 0.3345, "step": 16833, "task_loss": 0.16826188564300537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3280130922794342, "epoch": 14.23, "learning_rate": 2.852222917971196e-06, "loss": 0.4064, "step": 16834, "task_loss": 0.4911932349205017 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4984583258628845, "epoch": 14.23, "learning_rate": 2.84909204758923e-06, "loss": 0.4468, "step": 16835, "task_loss": 0.7974283695220947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36193686723709106, "epoch": 14.23, "learning_rate": 2.845961177207264e-06, "loss": 0.292, "step": 16836, "task_loss": 0.8045521974563599 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4395745098590851, "epoch": 14.23, "learning_rate": 2.842830306825298e-06, "loss": 0.4364, "step": 16837, "task_loss": 1.222826600074768 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30270472168922424, "epoch": 14.23, "learning_rate": 2.8396994364433313e-06, "loss": 0.352, "step": 16838, "task_loss": 1.0981618165969849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4132867455482483, "epoch": 14.23, "learning_rate": 2.836568566061365e-06, "loss": 0.481, "step": 16839, "task_loss": 0.7292972803115845 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5127169489860535, "epoch": 14.23, "learning_rate": 2.833437695679399e-06, "loss": 0.4301, "step": 16840, "task_loss": 0.738821268081665 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30517300963401794, "epoch": 14.24, "learning_rate": 2.8303068252974326e-06, "loss": 0.4197, "step": 16841, "task_loss": 1.2851279973983765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.12287113815546036, "epoch": 14.24, "learning_rate": 2.827175954915467e-06, "loss": 0.3604, "step": 16842, "task_loss": 0.07723229378461838 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4904131293296814, "epoch": 14.24, "learning_rate": 2.8240450845335004e-06, "loss": 0.4612, "step": 16843, "task_loss": 0.5380504131317139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3374950885772705, "epoch": 14.24, "learning_rate": 2.8209142141515343e-06, "loss": 0.4996, "step": 16844, "task_loss": 1.1050177812576294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38051676750183105, "epoch": 14.24, "learning_rate": 2.8177833437695677e-06, "loss": 0.4001, "step": 16845, "task_loss": 0.4026775360107422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4248877465724945, "epoch": 14.24, "learning_rate": 2.814652473387602e-06, "loss": 0.4488, "step": 16846, "task_loss": 0.4150923490524292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49452078342437744, "epoch": 14.24, "learning_rate": 2.8115216030056355e-06, "loss": 0.4568, "step": 16847, "task_loss": 1.2653331756591797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3455706536769867, "epoch": 14.24, "learning_rate": 2.8083907326236694e-06, "loss": 0.3089, "step": 16848, "task_loss": 0.5102758407592773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5419859886169434, "epoch": 14.24, "learning_rate": 2.8052598622417033e-06, "loss": 0.525, "step": 16849, "task_loss": 0.4211790859699249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40407103300094604, "epoch": 14.24, "learning_rate": 2.8021289918597372e-06, "loss": 0.4242, "step": 16850, "task_loss": 0.8957080841064453 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2637454867362976, "epoch": 14.24, "learning_rate": 2.7989981214777707e-06, "loss": 0.3866, "step": 16851, "task_loss": 0.4431425631046295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5033305287361145, "epoch": 14.24, "learning_rate": 2.795867251095805e-06, "loss": 0.404, "step": 16852, "task_loss": 0.832224428653717 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31082355976104736, "epoch": 14.25, "learning_rate": 2.7927363807138385e-06, "loss": 0.4291, "step": 16853, "task_loss": 0.6217875480651855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.521175742149353, "epoch": 14.25, "learning_rate": 2.7896055103318724e-06, "loss": 0.4303, "step": 16854, "task_loss": 0.7254982590675354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3093973994255066, "epoch": 14.25, "learning_rate": 2.7864746399499063e-06, "loss": 0.438, "step": 16855, "task_loss": 0.6640552282333374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41099095344543457, "epoch": 14.25, "learning_rate": 2.78334376956794e-06, "loss": 0.4837, "step": 16856, "task_loss": 0.4437297284603119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35916227102279663, "epoch": 14.25, "learning_rate": 2.7802128991859737e-06, "loss": 0.544, "step": 16857, "task_loss": 0.5664665102958679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3819722831249237, "epoch": 14.25, "learning_rate": 2.7770820288040076e-06, "loss": 0.4142, "step": 16858, "task_loss": 0.37226223945617676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26928871870040894, "epoch": 14.25, "learning_rate": 2.7739511584220415e-06, "loss": 0.3237, "step": 16859, "task_loss": 0.348271906375885 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44315940141677856, "epoch": 14.25, "learning_rate": 2.770820288040075e-06, "loss": 0.4402, "step": 16860, "task_loss": 0.33715564012527466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4261418282985687, "epoch": 14.25, "learning_rate": 2.7676894176581093e-06, "loss": 0.4985, "step": 16861, "task_loss": 1.0327155590057373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3361784815788269, "epoch": 14.25, "learning_rate": 2.7645585472761427e-06, "loss": 0.4244, "step": 16862, "task_loss": 0.8237431645393372 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27203312516212463, "epoch": 14.25, "learning_rate": 2.7614276768941766e-06, "loss": 0.3936, "step": 16863, "task_loss": 0.48978227376937866 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37870103120803833, "epoch": 14.26, "learning_rate": 2.7582968065122105e-06, "loss": 0.4338, "step": 16864, "task_loss": 0.8035740852355957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44729921221733093, "epoch": 14.26, "learning_rate": 2.7551659361302444e-06, "loss": 0.4914, "step": 16865, "task_loss": 0.424022912979126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43188321590423584, "epoch": 14.26, "learning_rate": 2.752035065748278e-06, "loss": 0.4179, "step": 16866, "task_loss": 0.5804831981658936 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32825666666030884, "epoch": 14.26, "learning_rate": 2.7489041953663122e-06, "loss": 0.3485, "step": 16867, "task_loss": 0.6519280672073364 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47014832496643066, "epoch": 14.26, "learning_rate": 2.7457733249843457e-06, "loss": 0.4144, "step": 16868, "task_loss": 0.9441032409667969 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35128432512283325, "epoch": 14.26, "learning_rate": 2.7426424546023796e-06, "loss": 0.4161, "step": 16869, "task_loss": 0.7342578172683716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5216079354286194, "epoch": 14.26, "learning_rate": 2.7395115842204135e-06, "loss": 0.4482, "step": 16870, "task_loss": 1.0270495414733887 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3423991799354553, "epoch": 14.26, "learning_rate": 2.7363807138384474e-06, "loss": 0.5024, "step": 16871, "task_loss": 0.6569247841835022 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28084254264831543, "epoch": 14.26, "learning_rate": 2.733249843456481e-06, "loss": 0.3929, "step": 16872, "task_loss": 1.2246469259262085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8179463744163513, "epoch": 14.26, "learning_rate": 2.7301189730745148e-06, "loss": 0.4258, "step": 16873, "task_loss": 1.632295846939087 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30140256881713867, "epoch": 14.26, "learning_rate": 2.7269881026925487e-06, "loss": 0.3688, "step": 16874, "task_loss": 0.32597672939300537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6076879501342773, "epoch": 14.26, "learning_rate": 2.7238572323105826e-06, "loss": 0.4878, "step": 16875, "task_loss": 1.2495604753494263 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4158540964126587, "epoch": 14.27, "learning_rate": 2.7207263619286165e-06, "loss": 0.4278, "step": 16876, "task_loss": 0.414215087890625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7365249991416931, "epoch": 14.27, "learning_rate": 2.71759549154665e-06, "loss": 0.4425, "step": 16877, "task_loss": 0.7719587087631226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.312755823135376, "epoch": 14.27, "learning_rate": 2.714464621164684e-06, "loss": 0.324, "step": 16878, "task_loss": 0.6069991588592529 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4606669545173645, "epoch": 14.27, "learning_rate": 2.7113337507827178e-06, "loss": 0.5014, "step": 16879, "task_loss": 0.7246982455253601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30392974615097046, "epoch": 14.27, "learning_rate": 2.7082028804007517e-06, "loss": 0.3387, "step": 16880, "task_loss": 0.41444331407546997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24765431880950928, "epoch": 14.27, "learning_rate": 2.705072010018785e-06, "loss": 0.3021, "step": 16881, "task_loss": 0.04889160394668579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4233742356300354, "epoch": 14.27, "learning_rate": 2.7019411396368194e-06, "loss": 0.4021, "step": 16882, "task_loss": 0.20455434918403625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.512436032295227, "epoch": 14.27, "learning_rate": 2.698810269254853e-06, "loss": 0.4052, "step": 16883, "task_loss": 0.6191865801811218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38527968525886536, "epoch": 14.27, "learning_rate": 2.695679398872887e-06, "loss": 0.4586, "step": 16884, "task_loss": 0.5359143614768982 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36162298917770386, "epoch": 14.27, "learning_rate": 2.6925485284909207e-06, "loss": 0.5297, "step": 16885, "task_loss": 1.0327314138412476 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28886717557907104, "epoch": 14.27, "learning_rate": 2.6894176581089546e-06, "loss": 0.3025, "step": 16886, "task_loss": 0.29749780893325806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42549076676368713, "epoch": 14.27, "learning_rate": 2.686286787726988e-06, "loss": 0.42, "step": 16887, "task_loss": 0.28640952706336975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6049543023109436, "epoch": 14.28, "learning_rate": 2.6831559173450224e-06, "loss": 0.6136, "step": 16888, "task_loss": 0.6840921640396118 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40304461121559143, "epoch": 14.28, "learning_rate": 2.680025046963056e-06, "loss": 0.4085, "step": 16889, "task_loss": 1.0584402084350586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2732566297054291, "epoch": 14.28, "learning_rate": 2.67689417658109e-06, "loss": 0.4827, "step": 16890, "task_loss": 0.6670897603034973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5541156530380249, "epoch": 14.28, "learning_rate": 2.6737633061991237e-06, "loss": 0.5415, "step": 16891, "task_loss": 0.7721157670021057 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7756498456001282, "epoch": 14.28, "learning_rate": 2.670632435817157e-06, "loss": 0.506, "step": 16892, "task_loss": 0.9605322480201721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1764499545097351, "epoch": 14.28, "learning_rate": 2.667501565435191e-06, "loss": 0.3219, "step": 16893, "task_loss": 0.9979633688926697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3354296088218689, "epoch": 14.28, "learning_rate": 2.664370695053225e-06, "loss": 0.4873, "step": 16894, "task_loss": 0.5367278456687927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2875908613204956, "epoch": 14.28, "learning_rate": 2.661239824671259e-06, "loss": 0.3845, "step": 16895, "task_loss": 0.84916752576828 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4107818007469177, "epoch": 14.28, "learning_rate": 2.6581089542892923e-06, "loss": 0.4621, "step": 16896, "task_loss": 1.2823753356933594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2969222962856293, "epoch": 14.28, "learning_rate": 2.6549780839073262e-06, "loss": 0.4129, "step": 16897, "task_loss": 0.4624881148338318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42139026522636414, "epoch": 14.28, "learning_rate": 2.65184721352536e-06, "loss": 0.5254, "step": 16898, "task_loss": 0.7346590757369995 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34942880272865295, "epoch": 14.28, "learning_rate": 2.648716343143394e-06, "loss": 0.4529, "step": 16899, "task_loss": 1.5853722095489502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41321784257888794, "epoch": 14.29, "learning_rate": 2.6455854727614275e-06, "loss": 0.4278, "step": 16900, "task_loss": 0.1384279578924179 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4330812096595764, "epoch": 14.29, "learning_rate": 2.642454602379462e-06, "loss": 0.4613, "step": 16901, "task_loss": 0.1388973742723465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4742439091205597, "epoch": 14.29, "learning_rate": 2.6393237319974953e-06, "loss": 0.4315, "step": 16902, "task_loss": 0.5763947367668152 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29920631647109985, "epoch": 14.29, "learning_rate": 2.636192861615529e-06, "loss": 0.3973, "step": 16903, "task_loss": 0.8230830430984497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4544236958026886, "epoch": 14.29, "learning_rate": 2.633061991233563e-06, "loss": 0.386, "step": 16904, "task_loss": 0.9173434972763062 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5352775454521179, "epoch": 14.29, "learning_rate": 2.629931120851597e-06, "loss": 0.5458, "step": 16905, "task_loss": 0.31956344842910767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3606671392917633, "epoch": 14.29, "learning_rate": 2.6268002504696305e-06, "loss": 0.4225, "step": 16906, "task_loss": 0.403218150138855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5469859838485718, "epoch": 14.29, "learning_rate": 2.623669380087665e-06, "loss": 0.4316, "step": 16907, "task_loss": 0.5976096391677856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40991321206092834, "epoch": 14.29, "learning_rate": 2.6205385097056983e-06, "loss": 0.4108, "step": 16908, "task_loss": 0.5744232535362244 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5528317093849182, "epoch": 14.29, "learning_rate": 2.617407639323732e-06, "loss": 0.4103, "step": 16909, "task_loss": 0.28172144293785095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4966018795967102, "epoch": 14.29, "learning_rate": 2.614276768941766e-06, "loss": 0.2914, "step": 16910, "task_loss": 0.21170172095298767 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3967249393463135, "epoch": 14.29, "learning_rate": 2.6111458985597995e-06, "loss": 0.4045, "step": 16911, "task_loss": 1.1174641847610474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36752796173095703, "epoch": 14.3, "learning_rate": 2.6080150281778334e-06, "loss": 0.4732, "step": 16912, "task_loss": 1.5603023767471313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3216938078403473, "epoch": 14.3, "learning_rate": 2.6048841577958673e-06, "loss": 0.5141, "step": 16913, "task_loss": 0.11344992369413376 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7221918106079102, "epoch": 14.3, "learning_rate": 2.6017532874139012e-06, "loss": 0.4798, "step": 16914, "task_loss": 0.7034560441970825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5659767985343933, "epoch": 14.3, "learning_rate": 2.5986224170319347e-06, "loss": 0.4465, "step": 16915, "task_loss": 0.36493560671806335 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3010462820529938, "epoch": 14.3, "learning_rate": 2.595491546649969e-06, "loss": 0.3766, "step": 16916, "task_loss": 0.3652375638484955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5430692434310913, "epoch": 14.3, "learning_rate": 2.5923606762680025e-06, "loss": 0.5085, "step": 16917, "task_loss": 1.2571969032287598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43151992559432983, "epoch": 14.3, "learning_rate": 2.5892298058860364e-06, "loss": 0.3894, "step": 16918, "task_loss": 0.09197002649307251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20767848193645477, "epoch": 14.3, "learning_rate": 2.5860989355040703e-06, "loss": 0.2802, "step": 16919, "task_loss": 0.04735748842358589 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3066610097885132, "epoch": 14.3, "learning_rate": 2.5829680651221042e-06, "loss": 0.3933, "step": 16920, "task_loss": 0.1164565309882164 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49706220626831055, "epoch": 14.3, "learning_rate": 2.5798371947401377e-06, "loss": 0.3759, "step": 16921, "task_loss": 0.7011775374412537 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.770554780960083, "epoch": 14.3, "learning_rate": 2.576706324358172e-06, "loss": 0.6207, "step": 16922, "task_loss": 1.1125162839889526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36150652170181274, "epoch": 14.3, "learning_rate": 2.5735754539762055e-06, "loss": 0.4355, "step": 16923, "task_loss": 0.9214553833007812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39973700046539307, "epoch": 14.31, "learning_rate": 2.5704445835942394e-06, "loss": 0.5276, "step": 16924, "task_loss": 0.35341450572013855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.14942902326583862, "epoch": 14.31, "learning_rate": 2.5673137132122733e-06, "loss": 0.4549, "step": 16925, "task_loss": 0.33860424160957336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6255999803543091, "epoch": 14.31, "learning_rate": 2.564182842830307e-06, "loss": 0.5213, "step": 16926, "task_loss": 0.5427426695823669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2506321370601654, "epoch": 14.31, "learning_rate": 2.5610519724483407e-06, "loss": 0.3997, "step": 16927, "task_loss": 0.43429669737815857 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4168117046356201, "epoch": 14.31, "learning_rate": 2.5579211020663746e-06, "loss": 0.48, "step": 16928, "task_loss": 0.683347225189209 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.285641074180603, "epoch": 14.31, "learning_rate": 2.5547902316844085e-06, "loss": 0.3108, "step": 16929, "task_loss": 0.3767116069793701 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.192729651927948, "epoch": 14.31, "learning_rate": 2.551659361302442e-06, "loss": 0.3644, "step": 16930, "task_loss": 0.24135538935661316 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24821221828460693, "epoch": 14.31, "learning_rate": 2.5485284909204763e-06, "loss": 0.3518, "step": 16931, "task_loss": 0.2761807143688202 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3973490595817566, "epoch": 14.31, "learning_rate": 2.5453976205385097e-06, "loss": 0.4858, "step": 16932, "task_loss": 0.7436056137084961 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3953794836997986, "epoch": 14.31, "learning_rate": 2.5422667501565436e-06, "loss": 0.3311, "step": 16933, "task_loss": 0.6790536046028137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2257581353187561, "epoch": 14.31, "learning_rate": 2.5391358797745775e-06, "loss": 0.3781, "step": 16934, "task_loss": 0.7577512860298157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3755621016025543, "epoch": 14.32, "learning_rate": 2.5360050093926114e-06, "loss": 0.5025, "step": 16935, "task_loss": 0.38519227504730225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.550688624382019, "epoch": 14.32, "learning_rate": 2.532874139010645e-06, "loss": 0.4446, "step": 16936, "task_loss": 0.40538644790649414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.859978199005127, "epoch": 14.32, "learning_rate": 2.5297432686286792e-06, "loss": 0.6814, "step": 16937, "task_loss": 0.8428338170051575 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2833622694015503, "epoch": 14.32, "learning_rate": 2.5266123982467127e-06, "loss": 0.4436, "step": 16938, "task_loss": 0.8469380140304565 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5748258233070374, "epoch": 14.32, "learning_rate": 2.5234815278647466e-06, "loss": 0.4731, "step": 16939, "task_loss": 0.8804970979690552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39345842599868774, "epoch": 14.32, "learning_rate": 2.52035065748278e-06, "loss": 0.429, "step": 16940, "task_loss": 0.9083094596862793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7291097044944763, "epoch": 14.32, "learning_rate": 2.5172197871008144e-06, "loss": 0.3669, "step": 16941, "task_loss": 0.4440445303916931 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2211826890707016, "epoch": 14.32, "learning_rate": 2.514088916718848e-06, "loss": 0.3409, "step": 16942, "task_loss": 1.4219880104064941 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2674461305141449, "epoch": 14.32, "learning_rate": 2.5109580463368818e-06, "loss": 0.4588, "step": 16943, "task_loss": 0.9667202234268188 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33086276054382324, "epoch": 14.32, "learning_rate": 2.5078271759549157e-06, "loss": 0.4308, "step": 16944, "task_loss": 0.41770774126052856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43897926807403564, "epoch": 14.32, "learning_rate": 2.5046963055729496e-06, "loss": 0.4288, "step": 16945, "task_loss": 1.1318669319152832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3584946095943451, "epoch": 14.32, "learning_rate": 2.501565435190983e-06, "loss": 0.5117, "step": 16946, "task_loss": 1.134596586227417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2536831200122833, "epoch": 14.33, "learning_rate": 2.498434564809017e-06, "loss": 0.5121, "step": 16947, "task_loss": 0.38555386662483215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2360806167125702, "epoch": 14.33, "learning_rate": 2.495303694427051e-06, "loss": 0.3961, "step": 16948, "task_loss": 0.6067257523536682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.52390456199646, "epoch": 14.33, "learning_rate": 2.4921728240450843e-06, "loss": 0.431, "step": 16949, "task_loss": 0.8766690492630005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46188926696777344, "epoch": 14.33, "learning_rate": 2.4890419536631186e-06, "loss": 0.397, "step": 16950, "task_loss": 1.375383973121643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29685887694358826, "epoch": 14.33, "learning_rate": 2.485911083281152e-06, "loss": 0.4833, "step": 16951, "task_loss": 0.5362703800201416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3911954462528229, "epoch": 14.33, "learning_rate": 2.482780212899186e-06, "loss": 0.4062, "step": 16952, "task_loss": 0.4345564544200897 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3213005065917969, "epoch": 14.33, "learning_rate": 2.47964934251722e-06, "loss": 0.3981, "step": 16953, "task_loss": 0.6563644409179688 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6857778429985046, "epoch": 14.33, "learning_rate": 2.476518472135254e-06, "loss": 0.4751, "step": 16954, "task_loss": 0.9314416646957397 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18154658377170563, "epoch": 14.33, "learning_rate": 2.4733876017532873e-06, "loss": 0.4153, "step": 16955, "task_loss": 1.0903838872909546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5001688599586487, "epoch": 14.33, "learning_rate": 2.4702567313713216e-06, "loss": 0.5342, "step": 16956, "task_loss": 0.9540747404098511 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29893964529037476, "epoch": 14.33, "learning_rate": 2.467125860989355e-06, "loss": 0.3797, "step": 16957, "task_loss": 0.6282181739807129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3427810072898865, "epoch": 14.33, "learning_rate": 2.463994990607389e-06, "loss": 0.4504, "step": 16958, "task_loss": 0.2690582871437073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44246339797973633, "epoch": 14.34, "learning_rate": 2.460864120225423e-06, "loss": 0.3955, "step": 16959, "task_loss": 0.4824792146682739 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6337159872055054, "epoch": 14.34, "learning_rate": 2.4577332498434568e-06, "loss": 0.45, "step": 16960, "task_loss": 1.3907444477081299 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36312136054039, "epoch": 14.34, "learning_rate": 2.4546023794614903e-06, "loss": 0.4598, "step": 16961, "task_loss": 0.4632100760936737 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4949847161769867, "epoch": 14.34, "learning_rate": 2.451471509079524e-06, "loss": 0.516, "step": 16962, "task_loss": 1.7358111143112183 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4260847270488739, "epoch": 14.34, "learning_rate": 2.448340638697558e-06, "loss": 0.4496, "step": 16963, "task_loss": 1.62985360622406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3400513529777527, "epoch": 14.34, "learning_rate": 2.445209768315592e-06, "loss": 0.3555, "step": 16964, "task_loss": 0.3390430510044098 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5419490337371826, "epoch": 14.34, "learning_rate": 2.442078897933626e-06, "loss": 0.4337, "step": 16965, "task_loss": 1.448744535446167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44007623195648193, "epoch": 14.34, "learning_rate": 2.4389480275516593e-06, "loss": 0.5875, "step": 16966, "task_loss": 1.3985419273376465 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.476474791765213, "epoch": 14.34, "learning_rate": 2.4358171571696932e-06, "loss": 0.5138, "step": 16967, "task_loss": 1.1435050964355469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24502462148666382, "epoch": 14.34, "learning_rate": 2.432686286787727e-06, "loss": 0.4791, "step": 16968, "task_loss": 0.6537520885467529 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40376704931259155, "epoch": 14.34, "learning_rate": 2.429555416405761e-06, "loss": 0.4495, "step": 16969, "task_loss": 0.6741514801979065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3195580542087555, "epoch": 14.34, "learning_rate": 2.4264245460237945e-06, "loss": 0.3825, "step": 16970, "task_loss": 0.7726265788078308 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7048090100288391, "epoch": 14.35, "learning_rate": 2.423293675641829e-06, "loss": 0.5822, "step": 16971, "task_loss": 0.9791837930679321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4710816442966461, "epoch": 14.35, "learning_rate": 2.4201628052598623e-06, "loss": 0.4598, "step": 16972, "task_loss": 0.6541593670845032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25442707538604736, "epoch": 14.35, "learning_rate": 2.417031934877896e-06, "loss": 0.3196, "step": 16973, "task_loss": 0.4482167959213257 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18006816506385803, "epoch": 14.35, "learning_rate": 2.41390106449593e-06, "loss": 0.2951, "step": 16974, "task_loss": 0.09349211305379868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38182422518730164, "epoch": 14.35, "learning_rate": 2.410770194113964e-06, "loss": 0.3794, "step": 16975, "task_loss": 1.0027036666870117 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46744734048843384, "epoch": 14.35, "learning_rate": 2.4076393237319975e-06, "loss": 0.4646, "step": 16976, "task_loss": 0.40447574853897095 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40375226736068726, "epoch": 14.35, "learning_rate": 2.404508453350032e-06, "loss": 0.4712, "step": 16977, "task_loss": 0.8360303640365601 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2627166211605072, "epoch": 14.35, "learning_rate": 2.4013775829680653e-06, "loss": 0.3437, "step": 16978, "task_loss": 0.41338589787483215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7379041314125061, "epoch": 14.35, "learning_rate": 2.398246712586099e-06, "loss": 0.475, "step": 16979, "task_loss": 0.32351577281951904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3998968303203583, "epoch": 14.35, "learning_rate": 2.395115842204133e-06, "loss": 0.4527, "step": 16980, "task_loss": 0.5885001420974731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4916171431541443, "epoch": 14.35, "learning_rate": 2.3919849718221665e-06, "loss": 0.5394, "step": 16981, "task_loss": 0.5033742785453796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4229595363140106, "epoch": 14.35, "learning_rate": 2.3888541014402004e-06, "loss": 0.4983, "step": 16982, "task_loss": 1.2141704559326172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4969249367713928, "epoch": 14.36, "learning_rate": 2.3857232310582343e-06, "loss": 0.5093, "step": 16983, "task_loss": 1.0630598068237305 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46316856145858765, "epoch": 14.36, "learning_rate": 2.3825923606762682e-06, "loss": 0.423, "step": 16984, "task_loss": 0.6580414772033691 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2468324899673462, "epoch": 14.36, "learning_rate": 2.3794614902943017e-06, "loss": 0.4075, "step": 16985, "task_loss": 0.20304061472415924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4632755517959595, "epoch": 14.36, "learning_rate": 2.3763306199123356e-06, "loss": 0.4814, "step": 16986, "task_loss": 0.7388198375701904 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4557124972343445, "epoch": 14.36, "learning_rate": 2.3731997495303695e-06, "loss": 0.3872, "step": 16987, "task_loss": 0.526508092880249 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27184566855430603, "epoch": 14.36, "learning_rate": 2.3700688791484034e-06, "loss": 0.3339, "step": 16988, "task_loss": 0.4505508244037628 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.265281617641449, "epoch": 14.36, "learning_rate": 2.366938008766437e-06, "loss": 0.3103, "step": 16989, "task_loss": 0.48957183957099915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4960874021053314, "epoch": 14.36, "learning_rate": 2.363807138384471e-06, "loss": 0.4687, "step": 16990, "task_loss": 0.9115301370620728 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3451184630393982, "epoch": 14.36, "learning_rate": 2.3606762680025047e-06, "loss": 0.4797, "step": 16991, "task_loss": 0.8047777414321899 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26995527744293213, "epoch": 14.36, "learning_rate": 2.3575453976205386e-06, "loss": 0.333, "step": 16992, "task_loss": 0.7914491891860962 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2461918741464615, "epoch": 14.36, "learning_rate": 2.3544145272385725e-06, "loss": 0.5169, "step": 16993, "task_loss": 0.5036942958831787 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.547360360622406, "epoch": 14.36, "learning_rate": 2.3512836568566064e-06, "loss": 0.4765, "step": 16994, "task_loss": 0.7030032873153687 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5179256200790405, "epoch": 14.37, "learning_rate": 2.34815278647464e-06, "loss": 0.3864, "step": 16995, "task_loss": 0.8099617958068848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4391767084598541, "epoch": 14.37, "learning_rate": 2.345021916092674e-06, "loss": 0.4251, "step": 16996, "task_loss": 0.2637600898742676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3198185861110687, "epoch": 14.37, "learning_rate": 2.3418910457107076e-06, "loss": 0.3185, "step": 16997, "task_loss": 0.2300020009279251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7188162803649902, "epoch": 14.37, "learning_rate": 2.3387601753287415e-06, "loss": 0.5301, "step": 16998, "task_loss": 1.4310415983200073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27125656604766846, "epoch": 14.37, "learning_rate": 2.3356293049467754e-06, "loss": 0.4602, "step": 16999, "task_loss": 0.9330090284347534 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32956308126449585, "epoch": 14.37, "learning_rate": 2.332498434564809e-06, "loss": 0.3264, "step": 17000, "task_loss": 0.08102009445428848 }, { "epoch": 14.37, "eval_accuracy": 0.9133861386138614, "eval_loss": 0.3013368546962738, "eval_runtime": 206.499, "eval_samples_per_second": 122.277, "eval_steps_per_second": 0.959, "step": 17000 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4224097728729248, "epoch": 14.37, "learning_rate": 2.329367564182843e-06, "loss": 0.4423, "step": 17001, "task_loss": 0.6053940057754517 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43341541290283203, "epoch": 14.37, "learning_rate": 2.3262366938008767e-06, "loss": 0.4432, "step": 17002, "task_loss": 0.9978618025779724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26511120796203613, "epoch": 14.37, "learning_rate": 2.3231058234189106e-06, "loss": 0.3964, "step": 17003, "task_loss": 0.17120735347270966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29614952206611633, "epoch": 14.37, "learning_rate": 2.319974953036944e-06, "loss": 0.4291, "step": 17004, "task_loss": 0.386795312166214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3776577413082123, "epoch": 14.37, "learning_rate": 2.3168440826549784e-06, "loss": 0.389, "step": 17005, "task_loss": 0.3832383453845978 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4512473940849304, "epoch": 14.38, "learning_rate": 2.313713212273012e-06, "loss": 0.4569, "step": 17006, "task_loss": 0.2908323407173157 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5078731775283813, "epoch": 14.38, "learning_rate": 2.3105823418910458e-06, "loss": 0.5326, "step": 17007, "task_loss": 0.5618603825569153 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5013499855995178, "epoch": 14.38, "learning_rate": 2.3074514715090797e-06, "loss": 0.5225, "step": 17008, "task_loss": 0.486826628446579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3716428279876709, "epoch": 14.38, "learning_rate": 2.3043206011271136e-06, "loss": 0.4553, "step": 17009, "task_loss": 0.4333835244178772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44317522644996643, "epoch": 14.38, "learning_rate": 2.301189730745147e-06, "loss": 0.3915, "step": 17010, "task_loss": 0.8114355802536011 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3020232021808624, "epoch": 14.38, "learning_rate": 2.2980588603631814e-06, "loss": 0.4188, "step": 17011, "task_loss": 0.1666896641254425 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6019055247306824, "epoch": 14.38, "learning_rate": 2.294927989981215e-06, "loss": 0.6724, "step": 17012, "task_loss": 0.9749858379364014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3760109543800354, "epoch": 14.38, "learning_rate": 2.2917971195992488e-06, "loss": 0.4412, "step": 17013, "task_loss": 0.6557241678237915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5988087058067322, "epoch": 14.38, "learning_rate": 2.2886662492172827e-06, "loss": 0.4657, "step": 17014, "task_loss": 1.0217214822769165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5436655282974243, "epoch": 14.38, "learning_rate": 2.2855353788353166e-06, "loss": 0.51, "step": 17015, "task_loss": 0.46601712703704834 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3836589455604553, "epoch": 14.38, "learning_rate": 2.28240450845335e-06, "loss": 0.4155, "step": 17016, "task_loss": 0.6716861128807068 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3056049644947052, "epoch": 14.38, "learning_rate": 2.279273638071384e-06, "loss": 0.3642, "step": 17017, "task_loss": 0.36967265605926514 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3844407796859741, "epoch": 14.39, "learning_rate": 2.276142767689418e-06, "loss": 0.4531, "step": 17018, "task_loss": 0.5781558156013489 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33329451084136963, "epoch": 14.39, "learning_rate": 2.2730118973074513e-06, "loss": 0.4278, "step": 17019, "task_loss": 0.6403993964195251 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2829543948173523, "epoch": 14.39, "learning_rate": 2.2698810269254856e-06, "loss": 0.4878, "step": 17020, "task_loss": 0.11275778710842133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42371058464050293, "epoch": 14.39, "learning_rate": 2.266750156543519e-06, "loss": 0.3379, "step": 17021, "task_loss": 0.6358447074890137 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37090373039245605, "epoch": 14.39, "learning_rate": 2.263619286161553e-06, "loss": 0.3768, "step": 17022, "task_loss": 0.5261725187301636 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4696246385574341, "epoch": 14.39, "learning_rate": 2.260488415779587e-06, "loss": 0.4002, "step": 17023, "task_loss": 0.5560694336891174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3889421224594116, "epoch": 14.39, "learning_rate": 2.257357545397621e-06, "loss": 0.563, "step": 17024, "task_loss": 0.9493229389190674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.436814546585083, "epoch": 14.39, "learning_rate": 2.2542266750156543e-06, "loss": 0.4278, "step": 17025, "task_loss": 0.8370649218559265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30854129791259766, "epoch": 14.39, "learning_rate": 2.2510958046336886e-06, "loss": 0.495, "step": 17026, "task_loss": 0.7150130271911621 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1429872214794159, "epoch": 14.39, "learning_rate": 2.247964934251722e-06, "loss": 0.371, "step": 17027, "task_loss": 0.06699833273887634 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7891684174537659, "epoch": 14.39, "learning_rate": 2.244834063869756e-06, "loss": 0.4892, "step": 17028, "task_loss": 0.8327986001968384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6311046481132507, "epoch": 14.39, "learning_rate": 2.24170319348779e-06, "loss": 0.4444, "step": 17029, "task_loss": 0.4155322313308716 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5061174631118774, "epoch": 14.4, "learning_rate": 2.2385723231058238e-06, "loss": 0.5137, "step": 17030, "task_loss": 0.262525737285614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.14721810817718506, "epoch": 14.4, "learning_rate": 2.2354414527238572e-06, "loss": 0.4203, "step": 17031, "task_loss": 0.15225312113761902 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5631392002105713, "epoch": 14.4, "learning_rate": 2.232310582341891e-06, "loss": 0.5505, "step": 17032, "task_loss": 0.20860423147678375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40474769473075867, "epoch": 14.4, "learning_rate": 2.229179711959925e-06, "loss": 0.4358, "step": 17033, "task_loss": 1.322075605392456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.50384521484375, "epoch": 14.4, "learning_rate": 2.226048841577959e-06, "loss": 0.4399, "step": 17034, "task_loss": 0.8847029209136963 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5669137835502625, "epoch": 14.4, "learning_rate": 2.2229179711959924e-06, "loss": 0.4219, "step": 17035, "task_loss": 0.7325623035430908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4784701466560364, "epoch": 14.4, "learning_rate": 2.2197871008140263e-06, "loss": 0.4941, "step": 17036, "task_loss": 0.8709529042243958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41373157501220703, "epoch": 14.4, "learning_rate": 2.21665623043206e-06, "loss": 0.4668, "step": 17037, "task_loss": 0.7388058304786682 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34762462973594666, "epoch": 14.4, "learning_rate": 2.2135253600500937e-06, "loss": 0.4131, "step": 17038, "task_loss": 0.10849006474018097 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33532771468162537, "epoch": 14.4, "learning_rate": 2.210394489668128e-06, "loss": 0.4411, "step": 17039, "task_loss": 0.4068177044391632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5850255489349365, "epoch": 14.4, "learning_rate": 2.2072636192861615e-06, "loss": 0.5336, "step": 17040, "task_loss": 1.4384632110595703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7811648845672607, "epoch": 14.4, "learning_rate": 2.2041327489041954e-06, "loss": 0.4967, "step": 17041, "task_loss": 0.6797370910644531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4746514558792114, "epoch": 14.41, "learning_rate": 2.2010018785222293e-06, "loss": 0.5036, "step": 17042, "task_loss": 0.7009791135787964 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2907337546348572, "epoch": 14.41, "learning_rate": 2.197871008140263e-06, "loss": 0.3019, "step": 17043, "task_loss": 0.4674879014492035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2732164263725281, "epoch": 14.41, "learning_rate": 2.1947401377582967e-06, "loss": 0.4702, "step": 17044, "task_loss": 1.4386523962020874 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44329631328582764, "epoch": 14.41, "learning_rate": 2.191609267376331e-06, "loss": 0.3579, "step": 17045, "task_loss": 0.3205452263355255 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30396372079849243, "epoch": 14.41, "learning_rate": 2.1884783969943645e-06, "loss": 0.4185, "step": 17046, "task_loss": 0.6953396797180176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2815837264060974, "epoch": 14.41, "learning_rate": 2.1853475266123984e-06, "loss": 0.429, "step": 17047, "task_loss": 0.6295050382614136 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46043556928634644, "epoch": 14.41, "learning_rate": 2.1822166562304323e-06, "loss": 0.4471, "step": 17048, "task_loss": 0.8123957514762878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3041290044784546, "epoch": 14.41, "learning_rate": 2.179085785848466e-06, "loss": 0.3422, "step": 17049, "task_loss": 1.1939655542373657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31454625725746155, "epoch": 14.41, "learning_rate": 2.1759549154664996e-06, "loss": 0.4323, "step": 17050, "task_loss": 1.0996830463409424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4345274567604065, "epoch": 14.41, "learning_rate": 2.1728240450845335e-06, "loss": 0.402, "step": 17051, "task_loss": 0.48853474855422974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4577874541282654, "epoch": 14.41, "learning_rate": 2.1696931747025674e-06, "loss": 0.4211, "step": 17052, "task_loss": 0.8907535076141357 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35886847972869873, "epoch": 14.41, "learning_rate": 2.1665623043206013e-06, "loss": 0.4073, "step": 17053, "task_loss": 0.4558582305908203 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.504020094871521, "epoch": 14.42, "learning_rate": 2.1634314339386352e-06, "loss": 0.4153, "step": 17054, "task_loss": 0.5848085880279541 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34031933546066284, "epoch": 14.42, "learning_rate": 2.1603005635566687e-06, "loss": 0.3852, "step": 17055, "task_loss": 0.4556264877319336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3586856424808502, "epoch": 14.42, "learning_rate": 2.1571696931747026e-06, "loss": 0.3882, "step": 17056, "task_loss": 0.4623953700065613 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33999255299568176, "epoch": 14.42, "learning_rate": 2.1540388227927365e-06, "loss": 0.3342, "step": 17057, "task_loss": 1.2324965000152588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.513388454914093, "epoch": 14.42, "learning_rate": 2.1509079524107704e-06, "loss": 0.3977, "step": 17058, "task_loss": 0.934048056602478 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49317553639411926, "epoch": 14.42, "learning_rate": 2.147777082028804e-06, "loss": 0.4432, "step": 17059, "task_loss": 0.5496571660041809 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37298110127449036, "epoch": 14.42, "learning_rate": 2.144646211646838e-06, "loss": 0.416, "step": 17060, "task_loss": 0.23669669032096863 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36476579308509827, "epoch": 14.42, "learning_rate": 2.1415153412648717e-06, "loss": 0.4683, "step": 17061, "task_loss": 0.6539167165756226 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4677216112613678, "epoch": 14.42, "learning_rate": 2.1383844708829056e-06, "loss": 0.279, "step": 17062, "task_loss": 0.6168791651725769 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.518626868724823, "epoch": 14.42, "learning_rate": 2.1352536005009395e-06, "loss": 0.5122, "step": 17063, "task_loss": 0.9121850728988647 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47359538078308105, "epoch": 14.42, "learning_rate": 2.1321227301189734e-06, "loss": 0.3231, "step": 17064, "task_loss": 0.9990469217300415 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3886065185070038, "epoch": 14.42, "learning_rate": 2.128991859737007e-06, "loss": 0.4206, "step": 17065, "task_loss": 0.253556489944458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32990631461143494, "epoch": 14.43, "learning_rate": 2.125860989355041e-06, "loss": 0.417, "step": 17066, "task_loss": 0.4137279987335205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3401615023612976, "epoch": 14.43, "learning_rate": 2.1227301189730746e-06, "loss": 0.4204, "step": 17067, "task_loss": 0.8747336864471436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21492800116539001, "epoch": 14.43, "learning_rate": 2.1195992485911085e-06, "loss": 0.3745, "step": 17068, "task_loss": 0.07780715823173523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19542288780212402, "epoch": 14.43, "learning_rate": 2.1164683782091424e-06, "loss": 0.3078, "step": 17069, "task_loss": 0.441522479057312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3404368758201599, "epoch": 14.43, "learning_rate": 2.113337507827176e-06, "loss": 0.4023, "step": 17070, "task_loss": 0.5195527672767639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5529292821884155, "epoch": 14.43, "learning_rate": 2.11020663744521e-06, "loss": 0.538, "step": 17071, "task_loss": 0.4953586161136627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23891079425811768, "epoch": 14.43, "learning_rate": 2.1070757670632437e-06, "loss": 0.2767, "step": 17072, "task_loss": 0.32229769229888916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5608909726142883, "epoch": 14.43, "learning_rate": 2.1039448966812776e-06, "loss": 0.5251, "step": 17073, "task_loss": 0.2852182984352112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31067201495170593, "epoch": 14.43, "learning_rate": 2.100814026299311e-06, "loss": 0.3854, "step": 17074, "task_loss": 0.39424699544906616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48377931118011475, "epoch": 14.43, "learning_rate": 2.0976831559173454e-06, "loss": 0.5126, "step": 17075, "task_loss": 0.8404556512832642 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6903190016746521, "epoch": 14.43, "learning_rate": 2.094552285535379e-06, "loss": 0.579, "step": 17076, "task_loss": 1.341103434562683 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5251779556274414, "epoch": 14.44, "learning_rate": 2.0914214151534128e-06, "loss": 0.4983, "step": 17077, "task_loss": 0.14310218393802643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26847127079963684, "epoch": 14.44, "learning_rate": 2.0882905447714463e-06, "loss": 0.4814, "step": 17078, "task_loss": 0.8884453773498535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4624747931957245, "epoch": 14.44, "learning_rate": 2.0851596743894806e-06, "loss": 0.431, "step": 17079, "task_loss": 1.0339730978012085 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7527148127555847, "epoch": 14.44, "learning_rate": 2.082028804007514e-06, "loss": 0.57, "step": 17080, "task_loss": 0.9098384976387024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3820611834526062, "epoch": 14.44, "learning_rate": 2.078897933625548e-06, "loss": 0.3955, "step": 17081, "task_loss": 0.36976709961891174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1693924218416214, "epoch": 14.44, "learning_rate": 2.075767063243582e-06, "loss": 0.3678, "step": 17082, "task_loss": 0.07108151912689209 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3194323182106018, "epoch": 14.44, "learning_rate": 2.0726361928616157e-06, "loss": 0.5563, "step": 17083, "task_loss": 0.8618040084838867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5273467898368835, "epoch": 14.44, "learning_rate": 2.0695053224796492e-06, "loss": 0.4579, "step": 17084, "task_loss": 0.6510439515113831 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4463006258010864, "epoch": 14.44, "learning_rate": 2.0663744520976835e-06, "loss": 0.4437, "step": 17085, "task_loss": 0.6728331446647644 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4626080393791199, "epoch": 14.44, "learning_rate": 2.063243581715717e-06, "loss": 0.355, "step": 17086, "task_loss": 0.6273505687713623 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4035361409187317, "epoch": 14.44, "learning_rate": 2.060112711333751e-06, "loss": 0.4414, "step": 17087, "task_loss": 0.216421440243721 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4702436625957489, "epoch": 14.44, "learning_rate": 2.056981840951785e-06, "loss": 0.4452, "step": 17088, "task_loss": 0.2622109651565552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38666635751724243, "epoch": 14.45, "learning_rate": 2.0538509705698183e-06, "loss": 0.4329, "step": 17089, "task_loss": 0.08540882170200348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37357574701309204, "epoch": 14.45, "learning_rate": 2.050720100187852e-06, "loss": 0.3731, "step": 17090, "task_loss": 0.32835695147514343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32942062616348267, "epoch": 14.45, "learning_rate": 2.047589229805886e-06, "loss": 0.4521, "step": 17091, "task_loss": 0.6114057898521423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3755573630332947, "epoch": 14.45, "learning_rate": 2.04445835942392e-06, "loss": 0.514, "step": 17092, "task_loss": 0.5717068314552307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4963315427303314, "epoch": 14.45, "learning_rate": 2.0413274890419535e-06, "loss": 0.498, "step": 17093, "task_loss": 0.4628738462924957 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4146181046962738, "epoch": 14.45, "learning_rate": 2.0381966186599878e-06, "loss": 0.3814, "step": 17094, "task_loss": 0.5248845815658569 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44675010442733765, "epoch": 14.45, "learning_rate": 2.0350657482780213e-06, "loss": 0.3645, "step": 17095, "task_loss": 0.39243653416633606 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37641003727912903, "epoch": 14.45, "learning_rate": 2.031934877896055e-06, "loss": 0.4088, "step": 17096, "task_loss": 0.44169825315475464 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20302490890026093, "epoch": 14.45, "learning_rate": 2.028804007514089e-06, "loss": 0.3626, "step": 17097, "task_loss": 0.09893098473548889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5129652619361877, "epoch": 14.45, "learning_rate": 2.025673137132123e-06, "loss": 0.4154, "step": 17098, "task_loss": 0.15456923842430115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37375715374946594, "epoch": 14.45, "learning_rate": 2.0225422667501564e-06, "loss": 0.3438, "step": 17099, "task_loss": 0.6767800450325012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23710153996944427, "epoch": 14.45, "learning_rate": 2.0194113963681908e-06, "loss": 0.3925, "step": 17100, "task_loss": 0.6784706115722656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3963415026664734, "epoch": 14.46, "learning_rate": 2.0162805259862242e-06, "loss": 0.3809, "step": 17101, "task_loss": 1.0974016189575195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27013465762138367, "epoch": 14.46, "learning_rate": 2.013149655604258e-06, "loss": 0.3619, "step": 17102, "task_loss": 0.3456646800041199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32096144556999207, "epoch": 14.46, "learning_rate": 2.010018785222292e-06, "loss": 0.3604, "step": 17103, "task_loss": 0.8245082497596741 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33317047357559204, "epoch": 14.46, "learning_rate": 2.006887914840326e-06, "loss": 0.4236, "step": 17104, "task_loss": 0.20510134100914001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4671610891819, "epoch": 14.46, "learning_rate": 2.0037570444583594e-06, "loss": 0.3713, "step": 17105, "task_loss": 1.302169680595398 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3505662977695465, "epoch": 14.46, "learning_rate": 2.0006261740763933e-06, "loss": 0.4154, "step": 17106, "task_loss": 0.28350159525871277 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1824244260787964, "epoch": 14.46, "learning_rate": 1.997495303694427e-06, "loss": 0.4572, "step": 17107, "task_loss": 0.3434523642063141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25623559951782227, "epoch": 14.46, "learning_rate": 1.9943644333124607e-06, "loss": 0.3574, "step": 17108, "task_loss": 0.10978834331035614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5830317735671997, "epoch": 14.46, "learning_rate": 1.991233562930495e-06, "loss": 0.3877, "step": 17109, "task_loss": 0.663288950920105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5273993015289307, "epoch": 14.46, "learning_rate": 1.9881026925485285e-06, "loss": 0.4051, "step": 17110, "task_loss": 0.35831737518310547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5855656266212463, "epoch": 14.46, "learning_rate": 1.9849718221665624e-06, "loss": 0.4857, "step": 17111, "task_loss": 1.0054036378860474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4674663543701172, "epoch": 14.46, "learning_rate": 1.9818409517845963e-06, "loss": 0.4309, "step": 17112, "task_loss": 0.18503636121749878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2932996153831482, "epoch": 14.47, "learning_rate": 1.97871008140263e-06, "loss": 0.4978, "step": 17113, "task_loss": 0.7442161440849304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3222893476486206, "epoch": 14.47, "learning_rate": 1.9755792110206636e-06, "loss": 0.4192, "step": 17114, "task_loss": 0.7568285465240479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22396814823150635, "epoch": 14.47, "learning_rate": 1.972448340638698e-06, "loss": 0.3434, "step": 17115, "task_loss": 0.6109732389450073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3655206263065338, "epoch": 14.47, "learning_rate": 1.9693174702567314e-06, "loss": 0.4706, "step": 17116, "task_loss": 0.8684963583946228 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6832581758499146, "epoch": 14.47, "learning_rate": 1.9661865998747653e-06, "loss": 0.4615, "step": 17117, "task_loss": 1.083510398864746 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36014050245285034, "epoch": 14.47, "learning_rate": 1.9630557294927992e-06, "loss": 0.3573, "step": 17118, "task_loss": 0.30781424045562744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3714837431907654, "epoch": 14.47, "learning_rate": 1.959924859110833e-06, "loss": 0.3506, "step": 17119, "task_loss": 1.101069688796997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39622974395751953, "epoch": 14.47, "learning_rate": 1.9567939887288666e-06, "loss": 0.3345, "step": 17120, "task_loss": 0.7741361856460571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4194400906562805, "epoch": 14.47, "learning_rate": 1.9536631183469005e-06, "loss": 0.4209, "step": 17121, "task_loss": 0.3098800778388977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24856427311897278, "epoch": 14.47, "learning_rate": 1.9505322479649344e-06, "loss": 0.4003, "step": 17122, "task_loss": 0.46408605575561523 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33665260672569275, "epoch": 14.47, "learning_rate": 1.9474013775829683e-06, "loss": 0.3347, "step": 17123, "task_loss": 0.5319698452949524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5659143924713135, "epoch": 14.47, "learning_rate": 1.9442705072010018e-06, "loss": 0.4905, "step": 17124, "task_loss": 0.7712333798408508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4208787679672241, "epoch": 14.48, "learning_rate": 1.9411396368190357e-06, "loss": 0.4706, "step": 17125, "task_loss": 1.3729037046432495 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26374927163124084, "epoch": 14.48, "learning_rate": 1.9380087664370696e-06, "loss": 0.4103, "step": 17126, "task_loss": 0.10871502012014389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2592189908027649, "epoch": 14.48, "learning_rate": 1.934877896055103e-06, "loss": 0.401, "step": 17127, "task_loss": 0.2634064257144928 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5163629651069641, "epoch": 14.48, "learning_rate": 1.9317470256731374e-06, "loss": 0.3844, "step": 17128, "task_loss": 1.0892045497894287 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2940589487552643, "epoch": 14.48, "learning_rate": 1.928616155291171e-06, "loss": 0.3845, "step": 17129, "task_loss": 0.0971054807305336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3297654092311859, "epoch": 14.48, "learning_rate": 1.9254852849092048e-06, "loss": 0.431, "step": 17130, "task_loss": 1.210829257965088 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4945014715194702, "epoch": 14.48, "learning_rate": 1.9223544145272387e-06, "loss": 0.5262, "step": 17131, "task_loss": 0.4511256814002991 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31258243322372437, "epoch": 14.48, "learning_rate": 1.9192235441452726e-06, "loss": 0.456, "step": 17132, "task_loss": 1.1360478401184082 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41305363178253174, "epoch": 14.48, "learning_rate": 1.916092673763306e-06, "loss": 0.3702, "step": 17133, "task_loss": 0.2268594205379486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5927967429161072, "epoch": 14.48, "learning_rate": 1.9129618033813403e-06, "loss": 0.6354, "step": 17134, "task_loss": 0.776505708694458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6878869533538818, "epoch": 14.48, "learning_rate": 1.909830932999374e-06, "loss": 0.4666, "step": 17135, "task_loss": 0.4707425832748413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5402549505233765, "epoch": 14.48, "learning_rate": 1.9067000626174075e-06, "loss": 0.3974, "step": 17136, "task_loss": 0.5685961246490479 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30798107385635376, "epoch": 14.49, "learning_rate": 1.9035691922354416e-06, "loss": 0.4812, "step": 17137, "task_loss": 0.23045311868190765 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4857250452041626, "epoch": 14.49, "learning_rate": 1.9004383218534753e-06, "loss": 0.5352, "step": 17138, "task_loss": 1.1092054843902588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5569819808006287, "epoch": 14.49, "learning_rate": 1.897307451471509e-06, "loss": 0.3404, "step": 17139, "task_loss": 1.00400710105896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3545370101928711, "epoch": 14.49, "learning_rate": 1.894176581089543e-06, "loss": 0.3603, "step": 17140, "task_loss": 0.1733947992324829 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4760085344314575, "epoch": 14.49, "learning_rate": 1.8910457107075768e-06, "loss": 0.488, "step": 17141, "task_loss": 0.5520528554916382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2508397102355957, "epoch": 14.49, "learning_rate": 1.8879148403256105e-06, "loss": 0.3365, "step": 17142, "task_loss": 0.2727559506893158 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4612678587436676, "epoch": 14.49, "learning_rate": 1.8847839699436446e-06, "loss": 0.4018, "step": 17143, "task_loss": 0.6322349309921265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33001041412353516, "epoch": 14.49, "learning_rate": 1.8816530995616783e-06, "loss": 0.3992, "step": 17144, "task_loss": 0.6406881213188171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30308282375335693, "epoch": 14.49, "learning_rate": 1.878522229179712e-06, "loss": 0.5592, "step": 17145, "task_loss": 1.3972036838531494 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3843546509742737, "epoch": 14.49, "learning_rate": 1.875391358797746e-06, "loss": 0.3329, "step": 17146, "task_loss": 0.7211971879005432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3494929373264313, "epoch": 14.49, "learning_rate": 1.8722604884157798e-06, "loss": 0.4382, "step": 17147, "task_loss": 0.4910690486431122 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4327915608882904, "epoch": 14.5, "learning_rate": 1.8691296180338134e-06, "loss": 0.3976, "step": 17148, "task_loss": 0.543530285358429 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44581884145736694, "epoch": 14.5, "learning_rate": 1.8659987476518473e-06, "loss": 0.3918, "step": 17149, "task_loss": 0.6715652346611023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6174805164337158, "epoch": 14.5, "learning_rate": 1.862867877269881e-06, "loss": 0.4512, "step": 17150, "task_loss": 1.2200100421905518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44085693359375, "epoch": 14.5, "learning_rate": 1.859737006887915e-06, "loss": 0.4335, "step": 17151, "task_loss": 0.8422995209693909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5089803338050842, "epoch": 14.5, "learning_rate": 1.8566061365059488e-06, "loss": 0.4879, "step": 17152, "task_loss": 0.42755523324012756 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3331029713153839, "epoch": 14.5, "learning_rate": 1.8534752661239825e-06, "loss": 0.4475, "step": 17153, "task_loss": 0.3573738634586334 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2687777876853943, "epoch": 14.5, "learning_rate": 1.8503443957420162e-06, "loss": 0.2745, "step": 17154, "task_loss": 0.4204857051372528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4128282964229584, "epoch": 14.5, "learning_rate": 1.8472135253600503e-06, "loss": 0.4586, "step": 17155, "task_loss": 0.4598551392555237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25596052408218384, "epoch": 14.5, "learning_rate": 1.844082654978084e-06, "loss": 0.4257, "step": 17156, "task_loss": 0.515296995639801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3365122675895691, "epoch": 14.5, "learning_rate": 1.8409517845961177e-06, "loss": 0.4396, "step": 17157, "task_loss": 0.8232836723327637 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5818839073181152, "epoch": 14.5, "learning_rate": 1.8378209142141518e-06, "loss": 0.49, "step": 17158, "task_loss": 1.4746359586715698 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.624703586101532, "epoch": 14.5, "learning_rate": 1.8346900438321855e-06, "loss": 0.4433, "step": 17159, "task_loss": 0.6088507175445557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5939818024635315, "epoch": 14.51, "learning_rate": 1.8315591734502192e-06, "loss": 0.4751, "step": 17160, "task_loss": 0.11638196557760239 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32193875312805176, "epoch": 14.51, "learning_rate": 1.8284283030682533e-06, "loss": 0.3962, "step": 17161, "task_loss": 0.5396692752838135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3688032031059265, "epoch": 14.51, "learning_rate": 1.825297432686287e-06, "loss": 0.405, "step": 17162, "task_loss": 0.7146035432815552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34023046493530273, "epoch": 14.51, "learning_rate": 1.8221665623043207e-06, "loss": 0.4062, "step": 17163, "task_loss": 0.5507836937904358 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6317505836486816, "epoch": 14.51, "learning_rate": 1.8190356919223546e-06, "loss": 0.5761, "step": 17164, "task_loss": 0.968604564666748 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26028740406036377, "epoch": 14.51, "learning_rate": 1.8159048215403885e-06, "loss": 0.3625, "step": 17165, "task_loss": 0.1907990276813507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5034575462341309, "epoch": 14.51, "learning_rate": 1.8127739511584221e-06, "loss": 0.3943, "step": 17166, "task_loss": 1.2303072214126587 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3189324140548706, "epoch": 14.51, "learning_rate": 1.809643080776456e-06, "loss": 0.5497, "step": 17167, "task_loss": 0.12679512798786163 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5811895728111267, "epoch": 14.51, "learning_rate": 1.8065122103944897e-06, "loss": 0.5386, "step": 17168, "task_loss": 0.7623231410980225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35094332695007324, "epoch": 14.51, "learning_rate": 1.8033813400125234e-06, "loss": 0.3884, "step": 17169, "task_loss": 0.5511020421981812 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41644835472106934, "epoch": 14.51, "learning_rate": 1.8002504696305573e-06, "loss": 0.4385, "step": 17170, "task_loss": 1.2136207818984985 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4445434808731079, "epoch": 14.51, "learning_rate": 1.7971195992485912e-06, "loss": 0.418, "step": 17171, "task_loss": 1.1115741729736328 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46563416719436646, "epoch": 14.52, "learning_rate": 1.793988728866625e-06, "loss": 0.3417, "step": 17172, "task_loss": 0.7812803387641907 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42502397298812866, "epoch": 14.52, "learning_rate": 1.7908578584846586e-06, "loss": 0.4429, "step": 17173, "task_loss": 0.6695414781570435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6955183744430542, "epoch": 14.52, "learning_rate": 1.7877269881026927e-06, "loss": 0.426, "step": 17174, "task_loss": 1.1654292345046997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23286494612693787, "epoch": 14.52, "learning_rate": 1.7845961177207264e-06, "loss": 0.3558, "step": 17175, "task_loss": 0.24160555005073547 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49450570344924927, "epoch": 14.52, "learning_rate": 1.78146524733876e-06, "loss": 0.5818, "step": 17176, "task_loss": 1.225767970085144 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3347471356391907, "epoch": 14.52, "learning_rate": 1.7783343769567942e-06, "loss": 0.5093, "step": 17177, "task_loss": 0.81629478931427 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43596622347831726, "epoch": 14.52, "learning_rate": 1.7752035065748279e-06, "loss": 0.5388, "step": 17178, "task_loss": 0.5252591967582703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4458302855491638, "epoch": 14.52, "learning_rate": 1.7720726361928616e-06, "loss": 0.5631, "step": 17179, "task_loss": 0.17128820717334747 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4463277757167816, "epoch": 14.52, "learning_rate": 1.7689417658108957e-06, "loss": 0.3589, "step": 17180, "task_loss": 0.8688385486602783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3874890208244324, "epoch": 14.52, "learning_rate": 1.7658108954289294e-06, "loss": 0.3351, "step": 17181, "task_loss": 0.1681387722492218 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2573944926261902, "epoch": 14.52, "learning_rate": 1.762680025046963e-06, "loss": 0.3295, "step": 17182, "task_loss": 0.0813586413860321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33523496985435486, "epoch": 14.52, "learning_rate": 1.7595491546649972e-06, "loss": 0.4247, "step": 17183, "task_loss": 0.20954161882400513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3353843092918396, "epoch": 14.53, "learning_rate": 1.7564182842830308e-06, "loss": 0.451, "step": 17184, "task_loss": 0.3416750133037567 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24771720170974731, "epoch": 14.53, "learning_rate": 1.7532874139010645e-06, "loss": 0.3286, "step": 17185, "task_loss": 0.869066596031189 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1734934002161026, "epoch": 14.53, "learning_rate": 1.7501565435190984e-06, "loss": 0.3207, "step": 17186, "task_loss": 0.3999532461166382 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4610580503940582, "epoch": 14.53, "learning_rate": 1.7470256731371321e-06, "loss": 0.4401, "step": 17187, "task_loss": 0.9649704694747925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8150652647018433, "epoch": 14.53, "learning_rate": 1.7438948027551658e-06, "loss": 0.4996, "step": 17188, "task_loss": 0.41869115829467773 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39124324917793274, "epoch": 14.53, "learning_rate": 1.7407639323732e-06, "loss": 0.3388, "step": 17189, "task_loss": 0.8398354053497314 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5649592876434326, "epoch": 14.53, "learning_rate": 1.7376330619912336e-06, "loss": 0.4813, "step": 17190, "task_loss": 0.8546082377433777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2996498644351959, "epoch": 14.53, "learning_rate": 1.7345021916092673e-06, "loss": 0.4047, "step": 17191, "task_loss": 0.7353487014770508 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35074225068092346, "epoch": 14.53, "learning_rate": 1.7313713212273014e-06, "loss": 0.3285, "step": 17192, "task_loss": 0.06174403056502342 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3170080780982971, "epoch": 14.53, "learning_rate": 1.728240450845335e-06, "loss": 0.431, "step": 17193, "task_loss": 0.7953160405158997 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6341738104820251, "epoch": 14.53, "learning_rate": 1.7251095804633688e-06, "loss": 0.4163, "step": 17194, "task_loss": 1.128311038017273 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4974920153617859, "epoch": 14.53, "learning_rate": 1.7219787100814029e-06, "loss": 0.4072, "step": 17195, "task_loss": 0.40026476979255676 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2541795074939728, "epoch": 14.54, "learning_rate": 1.7188478396994366e-06, "loss": 0.617, "step": 17196, "task_loss": 1.1928843259811401 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23028606176376343, "epoch": 14.54, "learning_rate": 1.7157169693174703e-06, "loss": 0.3363, "step": 17197, "task_loss": 0.5183702707290649 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23577705025672913, "epoch": 14.54, "learning_rate": 1.7125860989355044e-06, "loss": 0.3573, "step": 17198, "task_loss": 0.28214848041534424 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5232810378074646, "epoch": 14.54, "learning_rate": 1.709455228553538e-06, "loss": 0.4981, "step": 17199, "task_loss": 0.4145178496837616 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37670832872390747, "epoch": 14.54, "learning_rate": 1.7063243581715717e-06, "loss": 0.4411, "step": 17200, "task_loss": 0.35126960277557373 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6056106090545654, "epoch": 14.54, "learning_rate": 1.7031934877896056e-06, "loss": 0.4467, "step": 17201, "task_loss": 0.5828391313552856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.310541570186615, "epoch": 14.54, "learning_rate": 1.7000626174076395e-06, "loss": 0.3043, "step": 17202, "task_loss": 0.33325737714767456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5862204432487488, "epoch": 14.54, "learning_rate": 1.6969317470256732e-06, "loss": 0.5235, "step": 17203, "task_loss": 0.4860534965991974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6971747875213623, "epoch": 14.54, "learning_rate": 1.6938008766437071e-06, "loss": 0.6157, "step": 17204, "task_loss": 0.758511483669281 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3226625621318817, "epoch": 14.54, "learning_rate": 1.6906700062617408e-06, "loss": 0.4864, "step": 17205, "task_loss": 0.23236438632011414 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32875359058380127, "epoch": 14.54, "learning_rate": 1.6875391358797745e-06, "loss": 0.5218, "step": 17206, "task_loss": 1.018942952156067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6945748925209045, "epoch": 14.54, "learning_rate": 1.6844082654978086e-06, "loss": 0.5539, "step": 17207, "task_loss": 1.2604470252990723 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3931906819343567, "epoch": 14.55, "learning_rate": 1.6812773951158423e-06, "loss": 0.4116, "step": 17208, "task_loss": 0.3568468689918518 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49545061588287354, "epoch": 14.55, "learning_rate": 1.678146524733876e-06, "loss": 0.3723, "step": 17209, "task_loss": 0.3244306743144989 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3989715874195099, "epoch": 14.55, "learning_rate": 1.67501565435191e-06, "loss": 0.5803, "step": 17210, "task_loss": 0.2940036654472351 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26443877816200256, "epoch": 14.55, "learning_rate": 1.6718847839699438e-06, "loss": 0.3887, "step": 17211, "task_loss": 0.2641715109348297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2366946041584015, "epoch": 14.55, "learning_rate": 1.6687539135879775e-06, "loss": 0.3777, "step": 17212, "task_loss": 0.21904629468917847 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23915889859199524, "epoch": 14.55, "learning_rate": 1.6656230432060116e-06, "loss": 0.3311, "step": 17213, "task_loss": 0.48719826340675354 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3997771739959717, "epoch": 14.55, "learning_rate": 1.6624921728240453e-06, "loss": 0.4296, "step": 17214, "task_loss": 0.37938666343688965 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3508864641189575, "epoch": 14.55, "learning_rate": 1.659361302442079e-06, "loss": 0.4152, "step": 17215, "task_loss": 1.053418755531311 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42499208450317383, "epoch": 14.55, "learning_rate": 1.6562304320601126e-06, "loss": 0.4675, "step": 17216, "task_loss": 0.6221652626991272 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42762434482574463, "epoch": 14.55, "learning_rate": 1.6530995616781467e-06, "loss": 0.383, "step": 17217, "task_loss": 0.7209107875823975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1759936362504959, "epoch": 14.55, "learning_rate": 1.6499686912961804e-06, "loss": 0.31, "step": 17218, "task_loss": 0.02143777348101139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32599878311157227, "epoch": 14.56, "learning_rate": 1.6468378209142141e-06, "loss": 0.3883, "step": 17219, "task_loss": 0.7349538803100586 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3515596389770508, "epoch": 14.56, "learning_rate": 1.643706950532248e-06, "loss": 0.3706, "step": 17220, "task_loss": 0.21460795402526855 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4390218257904053, "epoch": 14.56, "learning_rate": 1.640576080150282e-06, "loss": 0.51, "step": 17221, "task_loss": 1.0591636896133423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4569816589355469, "epoch": 14.56, "learning_rate": 1.6374452097683156e-06, "loss": 0.5575, "step": 17222, "task_loss": 1.3406659364700317 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6836150288581848, "epoch": 14.56, "learning_rate": 1.6343143393863495e-06, "loss": 0.5776, "step": 17223, "task_loss": 1.5863791704177856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30748996138572693, "epoch": 14.56, "learning_rate": 1.6311834690043832e-06, "loss": 0.4324, "step": 17224, "task_loss": 0.20071610808372498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45699822902679443, "epoch": 14.56, "learning_rate": 1.6280525986224169e-06, "loss": 0.4688, "step": 17225, "task_loss": 1.3305816650390625 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3302941918373108, "epoch": 14.56, "learning_rate": 1.624921728240451e-06, "loss": 0.4475, "step": 17226, "task_loss": 0.8264939188957214 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3029214143753052, "epoch": 14.56, "learning_rate": 1.6217908578584847e-06, "loss": 0.5776, "step": 17227, "task_loss": 0.5357025861740112 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5746191740036011, "epoch": 14.56, "learning_rate": 1.6186599874765184e-06, "loss": 0.4967, "step": 17228, "task_loss": 1.407219409942627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5744656920433044, "epoch": 14.56, "learning_rate": 1.6155291170945525e-06, "loss": 0.4828, "step": 17229, "task_loss": 1.3388798236846924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3589078187942505, "epoch": 14.56, "learning_rate": 1.6123982467125862e-06, "loss": 0.4991, "step": 17230, "task_loss": 0.40696027874946594 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1679442822933197, "epoch": 14.57, "learning_rate": 1.6092673763306198e-06, "loss": 0.4005, "step": 17231, "task_loss": 0.14925263822078705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5106877088546753, "epoch": 14.57, "learning_rate": 1.606136505948654e-06, "loss": 0.447, "step": 17232, "task_loss": 0.8910394310951233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4541136920452118, "epoch": 14.57, "learning_rate": 1.6030056355666876e-06, "loss": 0.4363, "step": 17233, "task_loss": 0.5315420031547546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45758163928985596, "epoch": 14.57, "learning_rate": 1.5998747651847213e-06, "loss": 0.4115, "step": 17234, "task_loss": 0.5779350996017456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3459309935569763, "epoch": 14.57, "learning_rate": 1.5967438948027554e-06, "loss": 0.4796, "step": 17235, "task_loss": 0.45665818452835083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3580375909805298, "epoch": 14.57, "learning_rate": 1.5936130244207891e-06, "loss": 0.4072, "step": 17236, "task_loss": 0.9093347191810608 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5252633094787598, "epoch": 14.57, "learning_rate": 1.5904821540388228e-06, "loss": 0.4174, "step": 17237, "task_loss": 0.8439884185791016 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5348396301269531, "epoch": 14.57, "learning_rate": 1.5873512836568567e-06, "loss": 0.5759, "step": 17238, "task_loss": 0.9875627160072327 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6612797975540161, "epoch": 14.57, "learning_rate": 1.5842204132748904e-06, "loss": 0.4851, "step": 17239, "task_loss": 0.3518453538417816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3970454931259155, "epoch": 14.57, "learning_rate": 1.5810895428929243e-06, "loss": 0.5387, "step": 17240, "task_loss": 0.6235886812210083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47881972789764404, "epoch": 14.57, "learning_rate": 1.5779586725109582e-06, "loss": 0.4952, "step": 17241, "task_loss": 1.0482873916625977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30198046565055847, "epoch": 14.57, "learning_rate": 1.5748278021289919e-06, "loss": 0.4056, "step": 17242, "task_loss": 0.05242134630680084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4714646339416504, "epoch": 14.58, "learning_rate": 1.5716969317470256e-06, "loss": 0.4147, "step": 17243, "task_loss": 0.5463011264801025 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2727894186973572, "epoch": 14.58, "learning_rate": 1.5685660613650597e-06, "loss": 0.4421, "step": 17244, "task_loss": 0.9505635499954224 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5830545425415039, "epoch": 14.58, "learning_rate": 1.5654351909830934e-06, "loss": 0.4016, "step": 17245, "task_loss": 1.1720823049545288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5475051999092102, "epoch": 14.58, "learning_rate": 1.5623043206011273e-06, "loss": 0.5234, "step": 17246, "task_loss": 0.614881694316864 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3637557625770569, "epoch": 14.58, "learning_rate": 1.559173450219161e-06, "loss": 0.4011, "step": 17247, "task_loss": 0.561176598072052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42638325691223145, "epoch": 14.58, "learning_rate": 1.5560425798371949e-06, "loss": 0.4295, "step": 17248, "task_loss": 0.7968301773071289 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5601773262023926, "epoch": 14.58, "learning_rate": 1.5529117094552285e-06, "loss": 0.4841, "step": 17249, "task_loss": 0.6022218465805054 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3087989091873169, "epoch": 14.58, "learning_rate": 1.5497808390732624e-06, "loss": 0.3729, "step": 17250, "task_loss": 0.6413635611534119 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4698597192764282, "epoch": 14.58, "learning_rate": 1.5466499686912963e-06, "loss": 0.417, "step": 17251, "task_loss": 0.6083942651748657 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5878416895866394, "epoch": 14.58, "learning_rate": 1.54351909830933e-06, "loss": 0.4146, "step": 17252, "task_loss": 1.0150325298309326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3796367049217224, "epoch": 14.58, "learning_rate": 1.540388227927364e-06, "loss": 0.4923, "step": 17253, "task_loss": 0.7415944337844849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4676128625869751, "epoch": 14.58, "learning_rate": 1.5372573575453978e-06, "loss": 0.3949, "step": 17254, "task_loss": 0.9034557342529297 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45033371448516846, "epoch": 14.59, "learning_rate": 1.5341264871634315e-06, "loss": 0.5445, "step": 17255, "task_loss": 1.0362704992294312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27280083298683167, "epoch": 14.59, "learning_rate": 1.5309956167814654e-06, "loss": 0.4749, "step": 17256, "task_loss": 1.136778473854065 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45280537009239197, "epoch": 14.59, "learning_rate": 1.527864746399499e-06, "loss": 0.4788, "step": 17257, "task_loss": 0.9951274394989014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26841744780540466, "epoch": 14.59, "learning_rate": 1.5247338760175328e-06, "loss": 0.419, "step": 17258, "task_loss": 0.4445120394229889 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2763978838920593, "epoch": 14.59, "learning_rate": 1.5216030056355667e-06, "loss": 0.286, "step": 17259, "task_loss": 0.7184416055679321 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2963075637817383, "epoch": 14.59, "learning_rate": 1.5184721352536006e-06, "loss": 0.3339, "step": 17260, "task_loss": 0.17656856775283813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2531379759311676, "epoch": 14.59, "learning_rate": 1.5153412648716343e-06, "loss": 0.3941, "step": 17261, "task_loss": 0.3306145966053009 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33183592557907104, "epoch": 14.59, "learning_rate": 1.5122103944896682e-06, "loss": 0.3327, "step": 17262, "task_loss": 1.0159673690795898 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3940688371658325, "epoch": 14.59, "learning_rate": 1.509079524107702e-06, "loss": 0.4577, "step": 17263, "task_loss": 0.9605355262756348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28974857926368713, "epoch": 14.59, "learning_rate": 1.5059486537257358e-06, "loss": 0.3525, "step": 17264, "task_loss": 0.03501519933342934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3745652139186859, "epoch": 14.59, "learning_rate": 1.5028177833437697e-06, "loss": 0.577, "step": 17265, "task_loss": 0.1273271143436432 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37303176522254944, "epoch": 14.59, "learning_rate": 1.4996869129618036e-06, "loss": 0.3651, "step": 17266, "task_loss": 0.40447163581848145 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.860044538974762, "epoch": 14.6, "learning_rate": 1.4965560425798372e-06, "loss": 0.5916, "step": 17267, "task_loss": 0.5383151173591614 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2233278900384903, "epoch": 14.6, "learning_rate": 1.4934251721978711e-06, "loss": 0.318, "step": 17268, "task_loss": 0.0745527520775795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20887090265750885, "epoch": 14.6, "learning_rate": 1.490294301815905e-06, "loss": 0.2685, "step": 17269, "task_loss": 0.265988290309906 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37814944982528687, "epoch": 14.6, "learning_rate": 1.4871634314339387e-06, "loss": 0.428, "step": 17270, "task_loss": 0.6687100529670715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3034237027168274, "epoch": 14.6, "learning_rate": 1.4840325610519726e-06, "loss": 0.4598, "step": 17271, "task_loss": 0.6815308332443237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48559582233428955, "epoch": 14.6, "learning_rate": 1.4809016906700063e-06, "loss": 0.4193, "step": 17272, "task_loss": 1.1546450853347778 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3958321809768677, "epoch": 14.6, "learning_rate": 1.4777708202880402e-06, "loss": 0.3847, "step": 17273, "task_loss": 0.5302948355674744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.580419659614563, "epoch": 14.6, "learning_rate": 1.474639949906074e-06, "loss": 0.5044, "step": 17274, "task_loss": 1.3623911142349243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5084754228591919, "epoch": 14.6, "learning_rate": 1.4715090795241078e-06, "loss": 0.4033, "step": 17275, "task_loss": 0.792866051197052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5393266677856445, "epoch": 14.6, "learning_rate": 1.4683782091421415e-06, "loss": 0.4836, "step": 17276, "task_loss": 0.5101063251495361 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2739027142524719, "epoch": 14.6, "learning_rate": 1.4652473387601754e-06, "loss": 0.3873, "step": 17277, "task_loss": 0.3916671872138977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32769453525543213, "epoch": 14.6, "learning_rate": 1.462116468378209e-06, "loss": 0.3751, "step": 17278, "task_loss": 1.182735562324524 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3917797803878784, "epoch": 14.61, "learning_rate": 1.458985597996243e-06, "loss": 0.3567, "step": 17279, "task_loss": 0.19637514650821686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5470305681228638, "epoch": 14.61, "learning_rate": 1.4558547276142769e-06, "loss": 0.3494, "step": 17280, "task_loss": 0.9769011735916138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36930564045906067, "epoch": 14.61, "learning_rate": 1.4527238572323106e-06, "loss": 0.3386, "step": 17281, "task_loss": 0.2685878276824951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6520836353302002, "epoch": 14.61, "learning_rate": 1.4495929868503445e-06, "loss": 0.5598, "step": 17282, "task_loss": 0.8928559422492981 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6259812116622925, "epoch": 14.61, "learning_rate": 1.4464621164683784e-06, "loss": 0.4141, "step": 17283, "task_loss": 1.0065593719482422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5967378616333008, "epoch": 14.61, "learning_rate": 1.443331246086412e-06, "loss": 0.4944, "step": 17284, "task_loss": 0.6277262568473816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4341997504234314, "epoch": 14.61, "learning_rate": 1.440200375704446e-06, "loss": 0.3934, "step": 17285, "task_loss": 0.965166449546814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4914790093898773, "epoch": 14.61, "learning_rate": 1.4370695053224798e-06, "loss": 0.4996, "step": 17286, "task_loss": 1.4441951513290405 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23579122126102448, "epoch": 14.61, "learning_rate": 1.4339386349405135e-06, "loss": 0.3826, "step": 17287, "task_loss": 0.07365141808986664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3960835933685303, "epoch": 14.61, "learning_rate": 1.4308077645585474e-06, "loss": 0.4272, "step": 17288, "task_loss": 0.42647862434387207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5296005010604858, "epoch": 14.61, "learning_rate": 1.4276768941765813e-06, "loss": 0.4535, "step": 17289, "task_loss": 0.19638118147850037 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4988381564617157, "epoch": 14.61, "learning_rate": 1.424546023794615e-06, "loss": 0.431, "step": 17290, "task_loss": 0.6012593507766724 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4685893654823303, "epoch": 14.62, "learning_rate": 1.421415153412649e-06, "loss": 0.4694, "step": 17291, "task_loss": 0.5627483129501343 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4197808504104614, "epoch": 14.62, "learning_rate": 1.4182842830306826e-06, "loss": 0.3968, "step": 17292, "task_loss": 0.8268234729766846 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6513392925262451, "epoch": 14.62, "learning_rate": 1.4151534126487163e-06, "loss": 0.5, "step": 17293, "task_loss": 0.5154443979263306 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5747188925743103, "epoch": 14.62, "learning_rate": 1.4120225422667502e-06, "loss": 0.4371, "step": 17294, "task_loss": 1.051249384880066 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3696342706680298, "epoch": 14.62, "learning_rate": 1.4088916718847839e-06, "loss": 0.2927, "step": 17295, "task_loss": 0.12619948387145996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2519896328449249, "epoch": 14.62, "learning_rate": 1.4057608015028178e-06, "loss": 0.4181, "step": 17296, "task_loss": 0.7974138855934143 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20119225978851318, "epoch": 14.62, "learning_rate": 1.4026299311208517e-06, "loss": 0.2917, "step": 17297, "task_loss": 0.260373592376709 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45740705728530884, "epoch": 14.62, "learning_rate": 1.3994990607388854e-06, "loss": 0.3982, "step": 17298, "task_loss": 0.2627262473106384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.379129558801651, "epoch": 14.62, "learning_rate": 1.3963681903569193e-06, "loss": 0.3549, "step": 17299, "task_loss": 0.3021126687526703 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21881410479545593, "epoch": 14.62, "learning_rate": 1.3932373199749531e-06, "loss": 0.3969, "step": 17300, "task_loss": 0.6385721564292908 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35031658411026, "epoch": 14.62, "learning_rate": 1.3901064495929868e-06, "loss": 0.4585, "step": 17301, "task_loss": 1.2311146259307861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3911670446395874, "epoch": 14.63, "learning_rate": 1.3869755792110207e-06, "loss": 0.3901, "step": 17302, "task_loss": 0.891785740852356 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24634358286857605, "epoch": 14.63, "learning_rate": 1.3838447088290546e-06, "loss": 0.4279, "step": 17303, "task_loss": 0.3305288851261139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22739079594612122, "epoch": 14.63, "learning_rate": 1.3807138384470883e-06, "loss": 0.4081, "step": 17304, "task_loss": 0.23333978652954102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2509737014770508, "epoch": 14.63, "learning_rate": 1.3775829680651222e-06, "loss": 0.5666, "step": 17305, "task_loss": 0.8968561291694641 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3248009979724884, "epoch": 14.63, "learning_rate": 1.3744520976831561e-06, "loss": 0.4029, "step": 17306, "task_loss": 0.4645647406578064 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3371959328651428, "epoch": 14.63, "learning_rate": 1.3713212273011898e-06, "loss": 0.4657, "step": 17307, "task_loss": 0.6260866522789001 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22023454308509827, "epoch": 14.63, "learning_rate": 1.3681903569192237e-06, "loss": 0.3779, "step": 17308, "task_loss": 0.9183065891265869 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5598596930503845, "epoch": 14.63, "learning_rate": 1.3650594865372574e-06, "loss": 0.4527, "step": 17309, "task_loss": 0.4833635985851288 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5848575830459595, "epoch": 14.63, "learning_rate": 1.3619286161552913e-06, "loss": 0.4894, "step": 17310, "task_loss": 0.12340700626373291 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3043230175971985, "epoch": 14.63, "learning_rate": 1.358797745773325e-06, "loss": 0.3666, "step": 17311, "task_loss": 0.28896433115005493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6778768301010132, "epoch": 14.63, "learning_rate": 1.3556668753913589e-06, "loss": 0.4348, "step": 17312, "task_loss": 0.3596498668193817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33063891530036926, "epoch": 14.63, "learning_rate": 1.3525360050093926e-06, "loss": 0.4262, "step": 17313, "task_loss": 0.46729838848114014 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2666628360748291, "epoch": 14.64, "learning_rate": 1.3494051346274265e-06, "loss": 0.3828, "step": 17314, "task_loss": 0.3897934556007385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37514349818229675, "epoch": 14.64, "learning_rate": 1.3462742642454604e-06, "loss": 0.3996, "step": 17315, "task_loss": 0.3655327260494232 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3465350568294525, "epoch": 14.64, "learning_rate": 1.343143393863494e-06, "loss": 0.4306, "step": 17316, "task_loss": 0.8971902132034302 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44819894433021545, "epoch": 14.64, "learning_rate": 1.340012523481528e-06, "loss": 0.3959, "step": 17317, "task_loss": 0.889652669429779 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35851773619651794, "epoch": 14.64, "learning_rate": 1.3368816530995618e-06, "loss": 0.3855, "step": 17318, "task_loss": 0.1520441323518753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2131715565919876, "epoch": 14.64, "learning_rate": 1.3337507827175955e-06, "loss": 0.385, "step": 17319, "task_loss": 0.6853291988372803 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2293245643377304, "epoch": 14.64, "learning_rate": 1.3306199123356294e-06, "loss": 0.4229, "step": 17320, "task_loss": 0.24237465858459473 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26593366265296936, "epoch": 14.64, "learning_rate": 1.3274890419536631e-06, "loss": 0.383, "step": 17321, "task_loss": 0.13682004809379578 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38023146986961365, "epoch": 14.64, "learning_rate": 1.324358171571697e-06, "loss": 0.3263, "step": 17322, "task_loss": 0.6493584513664246 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3394341468811035, "epoch": 14.64, "learning_rate": 1.321227301189731e-06, "loss": 0.3428, "step": 17323, "task_loss": 0.9056917428970337 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4006498456001282, "epoch": 14.64, "learning_rate": 1.3180964308077646e-06, "loss": 0.4714, "step": 17324, "task_loss": 0.6913872361183167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39156514406204224, "epoch": 14.64, "learning_rate": 1.3149655604257985e-06, "loss": 0.3509, "step": 17325, "task_loss": 0.682421863079071 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5162192583084106, "epoch": 14.65, "learning_rate": 1.3118346900438324e-06, "loss": 0.4333, "step": 17326, "task_loss": 1.0378249883651733 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34804651141166687, "epoch": 14.65, "learning_rate": 1.308703819661866e-06, "loss": 0.4198, "step": 17327, "task_loss": 0.28543534874916077 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4196206331253052, "epoch": 14.65, "learning_rate": 1.3055729492798998e-06, "loss": 0.3214, "step": 17328, "task_loss": 0.47839874029159546 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23498983681201935, "epoch": 14.65, "learning_rate": 1.3024420788979337e-06, "loss": 0.3362, "step": 17329, "task_loss": 0.053035132586956024 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30291736125946045, "epoch": 14.65, "learning_rate": 1.2993112085159674e-06, "loss": 0.3664, "step": 17330, "task_loss": 0.0866217091679573 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39516544342041016, "epoch": 14.65, "learning_rate": 1.2961803381340013e-06, "loss": 0.3211, "step": 17331, "task_loss": 0.32983914017677307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43149691820144653, "epoch": 14.65, "learning_rate": 1.2930494677520352e-06, "loss": 0.4812, "step": 17332, "task_loss": 0.4982120394706726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37537384033203125, "epoch": 14.65, "learning_rate": 1.2899185973700688e-06, "loss": 0.4215, "step": 17333, "task_loss": 0.8161500096321106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2348639965057373, "epoch": 14.65, "learning_rate": 1.2867877269881027e-06, "loss": 0.3724, "step": 17334, "task_loss": 0.050869494676589966 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3106135129928589, "epoch": 14.65, "learning_rate": 1.2836568566061366e-06, "loss": 0.3454, "step": 17335, "task_loss": 0.20367896556854248 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4267975687980652, "epoch": 14.65, "learning_rate": 1.2805259862241703e-06, "loss": 0.347, "step": 17336, "task_loss": 0.5396055579185486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5422550439834595, "epoch": 14.65, "learning_rate": 1.2773951158422042e-06, "loss": 0.5408, "step": 17337, "task_loss": 0.7073110938072205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4812384843826294, "epoch": 14.66, "learning_rate": 1.2742642454602381e-06, "loss": 0.3293, "step": 17338, "task_loss": 1.171248197555542 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5541421175003052, "epoch": 14.66, "learning_rate": 1.2711333750782718e-06, "loss": 0.3957, "step": 17339, "task_loss": 0.9048150777816772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2813737988471985, "epoch": 14.66, "learning_rate": 1.2680025046963057e-06, "loss": 0.3664, "step": 17340, "task_loss": 0.42456522583961487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39900636672973633, "epoch": 14.66, "learning_rate": 1.2648716343143396e-06, "loss": 0.4364, "step": 17341, "task_loss": 0.3720570504665375 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3787494897842407, "epoch": 14.66, "learning_rate": 1.2617407639323733e-06, "loss": 0.4323, "step": 17342, "task_loss": 0.7346801161766052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3247024714946747, "epoch": 14.66, "learning_rate": 1.2586098935504072e-06, "loss": 0.4424, "step": 17343, "task_loss": 0.5584707856178284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38433927297592163, "epoch": 14.66, "learning_rate": 1.2554790231684409e-06, "loss": 0.373, "step": 17344, "task_loss": 0.19019602239131927 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5202157497406006, "epoch": 14.66, "learning_rate": 1.2523481527864748e-06, "loss": 0.4585, "step": 17345, "task_loss": 1.3662285804748535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26578718423843384, "epoch": 14.66, "learning_rate": 1.2492172824045085e-06, "loss": 0.4264, "step": 17346, "task_loss": 0.18300405144691467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4707932472229004, "epoch": 14.66, "learning_rate": 1.2460864120225422e-06, "loss": 0.5369, "step": 17347, "task_loss": 0.6052685976028442 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29120200872421265, "epoch": 14.66, "learning_rate": 1.242955541640576e-06, "loss": 0.3963, "step": 17348, "task_loss": 0.5371758937835693 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5255059003829956, "epoch": 14.66, "learning_rate": 1.23982467125861e-06, "loss": 0.4466, "step": 17349, "task_loss": 0.45712339878082275 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40821969509124756, "epoch": 14.67, "learning_rate": 1.2366938008766436e-06, "loss": 0.4004, "step": 17350, "task_loss": 0.5345869660377502 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2725977897644043, "epoch": 14.67, "learning_rate": 1.2335629304946775e-06, "loss": 0.3883, "step": 17351, "task_loss": 0.4434405267238617 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28785955905914307, "epoch": 14.67, "learning_rate": 1.2304320601127114e-06, "loss": 0.4715, "step": 17352, "task_loss": 0.6909152269363403 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5244784355163574, "epoch": 14.67, "learning_rate": 1.2273011897307451e-06, "loss": 0.4769, "step": 17353, "task_loss": 0.5924908518791199 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2312956154346466, "epoch": 14.67, "learning_rate": 1.224170319348779e-06, "loss": 0.4093, "step": 17354, "task_loss": 0.25616440176963806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22343634068965912, "epoch": 14.67, "learning_rate": 1.221039448966813e-06, "loss": 0.4333, "step": 17355, "task_loss": 0.2728278338909149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8343144655227661, "epoch": 14.67, "learning_rate": 1.2179085785848466e-06, "loss": 0.48, "step": 17356, "task_loss": 0.3452765941619873 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5060250163078308, "epoch": 14.67, "learning_rate": 1.2147777082028805e-06, "loss": 0.4382, "step": 17357, "task_loss": 0.4810221493244171 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3893536329269409, "epoch": 14.67, "learning_rate": 1.2116468378209144e-06, "loss": 0.5207, "step": 17358, "task_loss": 0.7343451976776123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2749096751213074, "epoch": 14.67, "learning_rate": 1.208515967438948e-06, "loss": 0.351, "step": 17359, "task_loss": 0.6341217160224915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33660343289375305, "epoch": 14.67, "learning_rate": 1.205385097056982e-06, "loss": 0.4, "step": 17360, "task_loss": 0.8653087615966797 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37508708238601685, "epoch": 14.67, "learning_rate": 1.202254226675016e-06, "loss": 0.4525, "step": 17361, "task_loss": 1.0624148845672607 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3612176179885864, "epoch": 14.68, "learning_rate": 1.1991233562930496e-06, "loss": 0.5289, "step": 17362, "task_loss": 1.0991487503051758 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25758111476898193, "epoch": 14.68, "learning_rate": 1.1959924859110833e-06, "loss": 0.3172, "step": 17363, "task_loss": 0.32445183396339417 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4308975636959076, "epoch": 14.68, "learning_rate": 1.1928616155291172e-06, "loss": 0.3682, "step": 17364, "task_loss": 0.32317912578582764 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1985877901315689, "epoch": 14.68, "learning_rate": 1.1897307451471509e-06, "loss": 0.3949, "step": 17365, "task_loss": 0.24320802092552185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3549959659576416, "epoch": 14.68, "learning_rate": 1.1865998747651848e-06, "loss": 0.4008, "step": 17366, "task_loss": 0.30011215806007385 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2536064088344574, "epoch": 14.68, "learning_rate": 1.1834690043832184e-06, "loss": 0.4361, "step": 17367, "task_loss": 0.592095136642456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42818060517311096, "epoch": 14.68, "learning_rate": 1.1803381340012523e-06, "loss": 0.3571, "step": 17368, "task_loss": 0.24397972226142883 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3769943118095398, "epoch": 14.68, "learning_rate": 1.1772072636192862e-06, "loss": 0.4642, "step": 17369, "task_loss": 0.21095257997512817 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2481415867805481, "epoch": 14.68, "learning_rate": 1.17407639323732e-06, "loss": 0.3922, "step": 17370, "task_loss": 0.10882478952407837 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4800458550453186, "epoch": 14.68, "learning_rate": 1.1709455228553538e-06, "loss": 0.4952, "step": 17371, "task_loss": 0.5917230248451233 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4422609508037567, "epoch": 14.68, "learning_rate": 1.1678146524733877e-06, "loss": 0.4287, "step": 17372, "task_loss": 0.9385318756103516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3874049484729767, "epoch": 14.69, "learning_rate": 1.1646837820914214e-06, "loss": 0.36, "step": 17373, "task_loss": 0.25232720375061035 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.376666396856308, "epoch": 14.69, "learning_rate": 1.1615529117094553e-06, "loss": 0.3961, "step": 17374, "task_loss": 0.29091933369636536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3603006601333618, "epoch": 14.69, "learning_rate": 1.1584220413274892e-06, "loss": 0.3844, "step": 17375, "task_loss": 0.6357316970825195 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5914411544799805, "epoch": 14.69, "learning_rate": 1.1552911709455229e-06, "loss": 0.4495, "step": 17376, "task_loss": 0.9139946699142456 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29654550552368164, "epoch": 14.69, "learning_rate": 1.1521603005635568e-06, "loss": 0.3222, "step": 17377, "task_loss": 0.32494306564331055 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.415818989276886, "epoch": 14.69, "learning_rate": 1.1490294301815907e-06, "loss": 0.5121, "step": 17378, "task_loss": 0.7932120561599731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.460249125957489, "epoch": 14.69, "learning_rate": 1.1458985597996244e-06, "loss": 0.3762, "step": 17379, "task_loss": 1.0070444345474243 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.639504075050354, "epoch": 14.69, "learning_rate": 1.1427676894176583e-06, "loss": 0.5678, "step": 17380, "task_loss": 1.273556113243103 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6360897421836853, "epoch": 14.69, "learning_rate": 1.139636819035692e-06, "loss": 0.4435, "step": 17381, "task_loss": 1.1172157526016235 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34216105937957764, "epoch": 14.69, "learning_rate": 1.1365059486537257e-06, "loss": 0.3918, "step": 17382, "task_loss": 0.671303391456604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5708301067352295, "epoch": 14.69, "learning_rate": 1.1333750782717596e-06, "loss": 0.5533, "step": 17383, "task_loss": 0.8994846940040588 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5834185481071472, "epoch": 14.69, "learning_rate": 1.1302442078897934e-06, "loss": 0.5211, "step": 17384, "task_loss": 0.7600444555282593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42027348279953003, "epoch": 14.7, "learning_rate": 1.1271133375078271e-06, "loss": 0.4344, "step": 17385, "task_loss": 0.5078406929969788 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7834392786026001, "epoch": 14.7, "learning_rate": 1.123982467125861e-06, "loss": 0.5621, "step": 17386, "task_loss": 0.41875359416007996 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34334659576416016, "epoch": 14.7, "learning_rate": 1.120851596743895e-06, "loss": 0.4132, "step": 17387, "task_loss": 0.30398181080818176 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6912457942962646, "epoch": 14.7, "learning_rate": 1.1177207263619286e-06, "loss": 0.4895, "step": 17388, "task_loss": 1.6065850257873535 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7028560638427734, "epoch": 14.7, "learning_rate": 1.1145898559799625e-06, "loss": 0.4504, "step": 17389, "task_loss": 0.3876378536224365 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.13954076170921326, "epoch": 14.7, "learning_rate": 1.1114589855979962e-06, "loss": 0.4218, "step": 17390, "task_loss": 0.1344662606716156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4366316795349121, "epoch": 14.7, "learning_rate": 1.10832811521603e-06, "loss": 0.4222, "step": 17391, "task_loss": 0.6768348217010498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.407121866941452, "epoch": 14.7, "learning_rate": 1.105197244834064e-06, "loss": 0.4499, "step": 17392, "task_loss": 0.15005278587341309 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2684584856033325, "epoch": 14.7, "learning_rate": 1.1020663744520977e-06, "loss": 0.3569, "step": 17393, "task_loss": 0.3548615276813507 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3194260597229004, "epoch": 14.7, "learning_rate": 1.0989355040701316e-06, "loss": 0.3383, "step": 17394, "task_loss": 0.1356050670146942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4407597780227661, "epoch": 14.7, "learning_rate": 1.0958046336881655e-06, "loss": 0.4217, "step": 17395, "task_loss": 0.9829456806182861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29224491119384766, "epoch": 14.7, "learning_rate": 1.0926737633061992e-06, "loss": 0.4157, "step": 17396, "task_loss": 0.6412857174873352 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20881730318069458, "epoch": 14.71, "learning_rate": 1.089542892924233e-06, "loss": 0.3567, "step": 17397, "task_loss": 0.254591703414917 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33709707856178284, "epoch": 14.71, "learning_rate": 1.0864120225422668e-06, "loss": 0.5665, "step": 17398, "task_loss": 0.8532932996749878 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4143275022506714, "epoch": 14.71, "learning_rate": 1.0832811521603007e-06, "loss": 0.4538, "step": 17399, "task_loss": 1.2819515466690063 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4600514769554138, "epoch": 14.71, "learning_rate": 1.0801502817783343e-06, "loss": 0.3806, "step": 17400, "task_loss": 0.6199791431427002 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34286630153656006, "epoch": 14.71, "learning_rate": 1.0770194113963682e-06, "loss": 0.4145, "step": 17401, "task_loss": 0.809653639793396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39348769187927246, "epoch": 14.71, "learning_rate": 1.073888541014402e-06, "loss": 0.4032, "step": 17402, "task_loss": 0.6886075139045715 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43215420842170715, "epoch": 14.71, "learning_rate": 1.0707576706324358e-06, "loss": 0.361, "step": 17403, "task_loss": 0.39189571142196655 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40688496828079224, "epoch": 14.71, "learning_rate": 1.0676268002504697e-06, "loss": 0.368, "step": 17404, "task_loss": 0.35265645384788513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3959280252456665, "epoch": 14.71, "learning_rate": 1.0644959298685034e-06, "loss": 0.4171, "step": 17405, "task_loss": 0.4814203381538391 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31471648812294006, "epoch": 14.71, "learning_rate": 1.0613650594865373e-06, "loss": 0.309, "step": 17406, "task_loss": 0.8572899103164673 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30698534846305847, "epoch": 14.71, "learning_rate": 1.0582341891045712e-06, "loss": 0.5049, "step": 17407, "task_loss": 0.3886871635913849 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7809884548187256, "epoch": 14.71, "learning_rate": 1.055103318722605e-06, "loss": 0.5281, "step": 17408, "task_loss": 1.0888711214065552 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3002720773220062, "epoch": 14.72, "learning_rate": 1.0519724483406388e-06, "loss": 0.4001, "step": 17409, "task_loss": 0.7748699188232422 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49653035402297974, "epoch": 14.72, "learning_rate": 1.0488415779586727e-06, "loss": 0.6249, "step": 17410, "task_loss": 0.5444968342781067 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31930750608444214, "epoch": 14.72, "learning_rate": 1.0457107075767064e-06, "loss": 0.3162, "step": 17411, "task_loss": 0.16732609272003174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34647712111473083, "epoch": 14.72, "learning_rate": 1.0425798371947403e-06, "loss": 0.4434, "step": 17412, "task_loss": 0.6444751620292664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23883169889450073, "epoch": 14.72, "learning_rate": 1.039448966812774e-06, "loss": 0.4472, "step": 17413, "task_loss": 1.1905170679092407 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31375470757484436, "epoch": 14.72, "learning_rate": 1.0363180964308079e-06, "loss": 0.3518, "step": 17414, "task_loss": 0.453307569026947 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5161420106887817, "epoch": 14.72, "learning_rate": 1.0331872260488418e-06, "loss": 0.384, "step": 17415, "task_loss": 0.6035832762718201 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.505784809589386, "epoch": 14.72, "learning_rate": 1.0300563556668755e-06, "loss": 0.325, "step": 17416, "task_loss": 0.4957762062549591 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16352707147598267, "epoch": 14.72, "learning_rate": 1.0269254852849091e-06, "loss": 0.3376, "step": 17417, "task_loss": 0.16791246831417084 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3442937135696411, "epoch": 14.72, "learning_rate": 1.023794614902943e-06, "loss": 0.3803, "step": 17418, "task_loss": 0.7513871192932129 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23505069315433502, "epoch": 14.72, "learning_rate": 1.0206637445209767e-06, "loss": 0.4604, "step": 17419, "task_loss": 0.0573924221098423 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5975992679595947, "epoch": 14.72, "learning_rate": 1.0175328741390106e-06, "loss": 0.4332, "step": 17420, "task_loss": 1.0179466009140015 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3659694492816925, "epoch": 14.73, "learning_rate": 1.0144020037570445e-06, "loss": 0.3809, "step": 17421, "task_loss": 1.4169472455978394 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.560700535774231, "epoch": 14.73, "learning_rate": 1.0112711333750782e-06, "loss": 0.4484, "step": 17422, "task_loss": 0.8913619518280029 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49463433027267456, "epoch": 14.73, "learning_rate": 1.0081402629931121e-06, "loss": 0.4759, "step": 17423, "task_loss": 1.0891544818878174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2712947726249695, "epoch": 14.73, "learning_rate": 1.005009392611146e-06, "loss": 0.3221, "step": 17424, "task_loss": 0.637656033039093 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4720640182495117, "epoch": 14.73, "learning_rate": 1.0018785222291797e-06, "loss": 0.3757, "step": 17425, "task_loss": 0.9228683114051819 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3467295169830322, "epoch": 14.73, "learning_rate": 9.987476518472136e-07, "loss": 0.4439, "step": 17426, "task_loss": 0.8807796239852905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34752440452575684, "epoch": 14.73, "learning_rate": 9.956167814652475e-07, "loss": 0.4187, "step": 17427, "task_loss": 0.04205408692359924 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3150089979171753, "epoch": 14.73, "learning_rate": 9.924859110832812e-07, "loss": 0.4544, "step": 17428, "task_loss": 0.3654564917087555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4225042164325714, "epoch": 14.73, "learning_rate": 9.89355040701315e-07, "loss": 0.3925, "step": 17429, "task_loss": 0.982194185256958 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39625781774520874, "epoch": 14.73, "learning_rate": 9.86224170319349e-07, "loss": 0.3319, "step": 17430, "task_loss": 0.05219617113471031 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.252630352973938, "epoch": 14.73, "learning_rate": 9.830932999373827e-07, "loss": 0.4429, "step": 17431, "task_loss": 0.18696680665016174 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4000331163406372, "epoch": 14.73, "learning_rate": 9.799624295554166e-07, "loss": 0.4659, "step": 17432, "task_loss": 0.3604182302951813 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43774551153182983, "epoch": 14.74, "learning_rate": 9.768315591734503e-07, "loss": 0.4623, "step": 17433, "task_loss": 0.7056487202644348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.320538729429245, "epoch": 14.74, "learning_rate": 9.737006887914842e-07, "loss": 0.4072, "step": 17434, "task_loss": 0.3710910677909851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26145827770233154, "epoch": 14.74, "learning_rate": 9.705698184095178e-07, "loss": 0.4368, "step": 17435, "task_loss": 0.26749199628829956 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2123226821422577, "epoch": 14.74, "learning_rate": 9.674389480275515e-07, "loss": 0.4299, "step": 17436, "task_loss": 0.19484879076480865 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6983475685119629, "epoch": 14.74, "learning_rate": 9.643080776455854e-07, "loss": 0.4502, "step": 17437, "task_loss": 0.3637765049934387 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39821961522102356, "epoch": 14.74, "learning_rate": 9.611772072636193e-07, "loss": 0.3823, "step": 17438, "task_loss": 0.7282067537307739 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5302053093910217, "epoch": 14.74, "learning_rate": 9.58046336881653e-07, "loss": 0.5213, "step": 17439, "task_loss": 0.2823602855205536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33287501335144043, "epoch": 14.74, "learning_rate": 9.54915466499687e-07, "loss": 0.4575, "step": 17440, "task_loss": 1.3701931238174438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.16663630306720734, "epoch": 14.74, "learning_rate": 9.517845961177208e-07, "loss": 0.3661, "step": 17441, "task_loss": 0.013462650589644909 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39852631092071533, "epoch": 14.74, "learning_rate": 9.486537257357545e-07, "loss": 0.3821, "step": 17442, "task_loss": 1.2780202627182007 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42874374985694885, "epoch": 14.74, "learning_rate": 9.455228553537884e-07, "loss": 0.4517, "step": 17443, "task_loss": 0.5594418048858643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3272033929824829, "epoch": 14.75, "learning_rate": 9.423919849718223e-07, "loss": 0.4178, "step": 17444, "task_loss": 0.5640364289283752 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4368686079978943, "epoch": 14.75, "learning_rate": 9.39261114589856e-07, "loss": 0.5698, "step": 17445, "task_loss": 0.2570543885231018 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3082779049873352, "epoch": 14.75, "learning_rate": 9.361302442078899e-07, "loss": 0.4276, "step": 17446, "task_loss": 0.31510087847709656 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.23514336347579956, "epoch": 14.75, "learning_rate": 9.329993738259237e-07, "loss": 0.3427, "step": 17447, "task_loss": 0.21492718160152435 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3259336054325104, "epoch": 14.75, "learning_rate": 9.298685034439575e-07, "loss": 0.4189, "step": 17448, "task_loss": 0.7075896263122559 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3067833483219147, "epoch": 14.75, "learning_rate": 9.267376330619913e-07, "loss": 0.3505, "step": 17449, "task_loss": 0.1717100292444229 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5601735711097717, "epoch": 14.75, "learning_rate": 9.236067626800252e-07, "loss": 0.5265, "step": 17450, "task_loss": 0.9018524885177612 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3329622745513916, "epoch": 14.75, "learning_rate": 9.204758922980588e-07, "loss": 0.4976, "step": 17451, "task_loss": 0.5651460289955139 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4696974456310272, "epoch": 14.75, "learning_rate": 9.173450219160927e-07, "loss": 0.5509, "step": 17452, "task_loss": 1.4400758743286133 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6467139720916748, "epoch": 14.75, "learning_rate": 9.142141515341266e-07, "loss": 0.5047, "step": 17453, "task_loss": 1.5452544689178467 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8448483943939209, "epoch": 14.75, "learning_rate": 9.110832811521603e-07, "loss": 0.4616, "step": 17454, "task_loss": 0.27767035365104675 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4567837715148926, "epoch": 14.75, "learning_rate": 9.079524107701942e-07, "loss": 0.4324, "step": 17455, "task_loss": 0.42696890234947205 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26188087463378906, "epoch": 14.76, "learning_rate": 9.04821540388228e-07, "loss": 0.4912, "step": 17456, "task_loss": 0.1051701158285141 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.343291699886322, "epoch": 14.76, "learning_rate": 9.016906700062617e-07, "loss": 0.5523, "step": 17457, "task_loss": 0.5831454992294312 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3836226463317871, "epoch": 14.76, "learning_rate": 8.985597996242956e-07, "loss": 0.503, "step": 17458, "task_loss": 0.13538476824760437 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3833887577056885, "epoch": 14.76, "learning_rate": 8.954289292423293e-07, "loss": 0.4067, "step": 17459, "task_loss": 0.7514165043830872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48567795753479004, "epoch": 14.76, "learning_rate": 8.922980588603632e-07, "loss": 0.4322, "step": 17460, "task_loss": 0.702591061592102 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41265633702278137, "epoch": 14.76, "learning_rate": 8.891671884783971e-07, "loss": 0.3855, "step": 17461, "task_loss": 0.9043744802474976 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30036476254463196, "epoch": 14.76, "learning_rate": 8.860363180964308e-07, "loss": 0.4103, "step": 17462, "task_loss": 0.8510677814483643 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3822983503341675, "epoch": 14.76, "learning_rate": 8.829054477144647e-07, "loss": 0.4527, "step": 17463, "task_loss": 0.6902025938034058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5090236067771912, "epoch": 14.76, "learning_rate": 8.797745773324986e-07, "loss": 0.4017, "step": 17464, "task_loss": 0.21932515501976013 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3839564919471741, "epoch": 14.76, "learning_rate": 8.766437069505323e-07, "loss": 0.4784, "step": 17465, "task_loss": 0.7465022802352905 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.35774463415145874, "epoch": 14.76, "learning_rate": 8.735128365685661e-07, "loss": 0.3798, "step": 17466, "task_loss": 0.4019920825958252 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2913977801799774, "epoch": 14.76, "learning_rate": 8.703819661866e-07, "loss": 0.2911, "step": 17467, "task_loss": 0.5612772107124329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4239044785499573, "epoch": 14.77, "learning_rate": 8.672510958046336e-07, "loss": 0.4932, "step": 17468, "task_loss": 1.431950330734253 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3280411958694458, "epoch": 14.77, "learning_rate": 8.641202254226675e-07, "loss": 0.3995, "step": 17469, "task_loss": 0.6154451966285706 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5969080924987793, "epoch": 14.77, "learning_rate": 8.609893550407014e-07, "loss": 0.4401, "step": 17470, "task_loss": 0.732383668422699 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4905722737312317, "epoch": 14.77, "learning_rate": 8.578584846587351e-07, "loss": 0.506, "step": 17471, "task_loss": 0.7851163148880005 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34914064407348633, "epoch": 14.77, "learning_rate": 8.54727614276769e-07, "loss": 0.2991, "step": 17472, "task_loss": 0.22110235691070557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6537275314331055, "epoch": 14.77, "learning_rate": 8.515967438948028e-07, "loss": 0.4077, "step": 17473, "task_loss": 0.7838630080223083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40424567461013794, "epoch": 14.77, "learning_rate": 8.484658735128366e-07, "loss": 0.4139, "step": 17474, "task_loss": 0.5364176034927368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27456700801849365, "epoch": 14.77, "learning_rate": 8.453350031308704e-07, "loss": 0.4089, "step": 17475, "task_loss": 0.6102344989776611 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7104566693305969, "epoch": 14.77, "learning_rate": 8.422041327489043e-07, "loss": 0.44, "step": 17476, "task_loss": 0.9411097764968872 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2729089856147766, "epoch": 14.77, "learning_rate": 8.39073262366938e-07, "loss": 0.349, "step": 17477, "task_loss": 0.501157283782959 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42582279443740845, "epoch": 14.77, "learning_rate": 8.359423919849719e-07, "loss": 0.4453, "step": 17478, "task_loss": 0.5360393524169922 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34667107462882996, "epoch": 14.77, "learning_rate": 8.328115216030058e-07, "loss": 0.2978, "step": 17479, "task_loss": 0.3711074888706207 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4704796075820923, "epoch": 14.78, "learning_rate": 8.296806512210395e-07, "loss": 0.4074, "step": 17480, "task_loss": 0.22302082180976868 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.587498664855957, "epoch": 14.78, "learning_rate": 8.265497808390734e-07, "loss": 0.5165, "step": 17481, "task_loss": 0.2699899971485138 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33280861377716064, "epoch": 14.78, "learning_rate": 8.234189104571071e-07, "loss": 0.3306, "step": 17482, "task_loss": 0.2420700639486313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5269851684570312, "epoch": 14.78, "learning_rate": 8.20288040075141e-07, "loss": 0.309, "step": 17483, "task_loss": 0.8841294646263123 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8096013069152832, "epoch": 14.78, "learning_rate": 8.171571696931748e-07, "loss": 0.4249, "step": 17484, "task_loss": 1.155704379081726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36395174264907837, "epoch": 14.78, "learning_rate": 8.140262993112084e-07, "loss": 0.4649, "step": 17485, "task_loss": 0.3848752975463867 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20805776119232178, "epoch": 14.78, "learning_rate": 8.108954289292423e-07, "loss": 0.4177, "step": 17486, "task_loss": 0.17736120522022247 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5451095104217529, "epoch": 14.78, "learning_rate": 8.077645585472762e-07, "loss": 0.5641, "step": 17487, "task_loss": 0.5736080408096313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.30495142936706543, "epoch": 14.78, "learning_rate": 8.046336881653099e-07, "loss": 0.3108, "step": 17488, "task_loss": 0.30009031295776367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41035521030426025, "epoch": 14.78, "learning_rate": 8.015028177833438e-07, "loss": 0.3932, "step": 17489, "task_loss": 0.8912566304206848 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2616909146308899, "epoch": 14.78, "learning_rate": 7.983719474013777e-07, "loss": 0.458, "step": 17490, "task_loss": 0.5497511029243469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2604526877403259, "epoch": 14.78, "learning_rate": 7.952410770194114e-07, "loss": 0.4477, "step": 17491, "task_loss": 0.25668787956237793 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3615787923336029, "epoch": 14.79, "learning_rate": 7.921102066374452e-07, "loss": 0.4801, "step": 17492, "task_loss": 0.9310441017150879 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.9154462218284607, "epoch": 14.79, "learning_rate": 7.889793362554791e-07, "loss": 0.6573, "step": 17493, "task_loss": 0.5186396241188049 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3194252848625183, "epoch": 14.79, "learning_rate": 7.858484658735128e-07, "loss": 0.3876, "step": 17494, "task_loss": 0.7270980477333069 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5203757286071777, "epoch": 14.79, "learning_rate": 7.827175954915467e-07, "loss": 0.5252, "step": 17495, "task_loss": 0.8221213221549988 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3201165199279785, "epoch": 14.79, "learning_rate": 7.795867251095805e-07, "loss": 0.3535, "step": 17496, "task_loss": 0.7729270458221436 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42389512062072754, "epoch": 14.79, "learning_rate": 7.764558547276143e-07, "loss": 0.563, "step": 17497, "task_loss": 0.7029582858085632 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7272272109985352, "epoch": 14.79, "learning_rate": 7.733249843456482e-07, "loss": 0.5191, "step": 17498, "task_loss": 0.43924814462661743 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46120187640190125, "epoch": 14.79, "learning_rate": 7.70194113963682e-07, "loss": 0.3977, "step": 17499, "task_loss": 0.5990526080131531 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40943384170532227, "epoch": 14.79, "learning_rate": 7.670632435817158e-07, "loss": 0.4136, "step": 17500, "task_loss": 0.5921612977981567 }, { "epoch": 14.79, "eval_accuracy": 0.9140990099009901, "eval_loss": 0.2976529002189636, "eval_runtime": 206.8137, "eval_samples_per_second": 122.091, "eval_steps_per_second": 0.957, "step": 17500 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2598341405391693, "epoch": 14.79, "learning_rate": 7.639323731997496e-07, "loss": 0.3664, "step": 17501, "task_loss": 0.025721464306116104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1979258805513382, "epoch": 14.79, "learning_rate": 7.608015028177833e-07, "loss": 0.3513, "step": 17502, "task_loss": 0.08259792625904083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4841429889202118, "epoch": 14.79, "learning_rate": 7.576706324358171e-07, "loss": 0.4214, "step": 17503, "task_loss": 0.577531099319458 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.76414555311203, "epoch": 14.8, "learning_rate": 7.54539762053851e-07, "loss": 0.6161, "step": 17504, "task_loss": 2.748385429382324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5579216480255127, "epoch": 14.8, "learning_rate": 7.514088916718848e-07, "loss": 0.48, "step": 17505, "task_loss": 0.8262220025062561 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5748134255409241, "epoch": 14.8, "learning_rate": 7.482780212899186e-07, "loss": 0.4692, "step": 17506, "task_loss": 0.48197776079177856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3231087625026703, "epoch": 14.8, "learning_rate": 7.451471509079525e-07, "loss": 0.3773, "step": 17507, "task_loss": 1.1516470909118652 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7705633044242859, "epoch": 14.8, "learning_rate": 7.420162805259863e-07, "loss": 0.4414, "step": 17508, "task_loss": 0.5439612865447998 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2558196485042572, "epoch": 14.8, "learning_rate": 7.388854101440201e-07, "loss": 0.3254, "step": 17509, "task_loss": 1.0462419986724854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46708303689956665, "epoch": 14.8, "learning_rate": 7.357545397620539e-07, "loss": 0.3844, "step": 17510, "task_loss": 0.5954985022544861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.42407864332199097, "epoch": 14.8, "learning_rate": 7.326236693800877e-07, "loss": 0.378, "step": 17511, "task_loss": 0.945533037185669 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3543528914451599, "epoch": 14.8, "learning_rate": 7.294927989981215e-07, "loss": 0.3777, "step": 17512, "task_loss": 0.3572937846183777 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5372596383094788, "epoch": 14.8, "learning_rate": 7.263619286161553e-07, "loss": 0.4151, "step": 17513, "task_loss": 0.5137104392051697 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.8001260161399841, "epoch": 14.8, "learning_rate": 7.232310582341892e-07, "loss": 0.4848, "step": 17514, "task_loss": 0.7138250470161438 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43612268567085266, "epoch": 14.81, "learning_rate": 7.20100187852223e-07, "loss": 0.3718, "step": 17515, "task_loss": 0.31604647636413574 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3975498676300049, "epoch": 14.81, "learning_rate": 7.169693174702568e-07, "loss": 0.4253, "step": 17516, "task_loss": 1.1441733837127686 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38566941022872925, "epoch": 14.81, "learning_rate": 7.138384470882907e-07, "loss": 0.4213, "step": 17517, "task_loss": 0.049607452005147934 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.197536438703537, "epoch": 14.81, "learning_rate": 7.107075767063245e-07, "loss": 0.363, "step": 17518, "task_loss": 0.09107402712106705 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3784664273262024, "epoch": 14.81, "learning_rate": 7.075767063243581e-07, "loss": 0.3426, "step": 17519, "task_loss": 1.3022325038909912 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4547731280326843, "epoch": 14.81, "learning_rate": 7.044458359423919e-07, "loss": 0.5645, "step": 17520, "task_loss": 0.7327596545219421 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4261188507080078, "epoch": 14.81, "learning_rate": 7.013149655604258e-07, "loss": 0.3802, "step": 17521, "task_loss": 0.9268866181373596 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40910184383392334, "epoch": 14.81, "learning_rate": 6.981840951784596e-07, "loss": 0.3579, "step": 17522, "task_loss": 1.0435246229171753 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.831295371055603, "epoch": 14.81, "learning_rate": 6.950532247964934e-07, "loss": 0.6164, "step": 17523, "task_loss": 0.5343356132507324 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6697612404823303, "epoch": 14.81, "learning_rate": 6.919223544145273e-07, "loss": 0.4787, "step": 17524, "task_loss": 0.8470689058303833 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40081679821014404, "epoch": 14.81, "learning_rate": 6.887914840325611e-07, "loss": 0.3513, "step": 17525, "task_loss": 0.5980989933013916 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3950958251953125, "epoch": 14.81, "learning_rate": 6.856606136505949e-07, "loss": 0.3481, "step": 17526, "task_loss": 0.9271990060806274 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3536144196987152, "epoch": 14.82, "learning_rate": 6.825297432686287e-07, "loss": 0.4374, "step": 17527, "task_loss": 0.39516815543174744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2720435857772827, "epoch": 14.82, "learning_rate": 6.793988728866625e-07, "loss": 0.3892, "step": 17528, "task_loss": 0.1648320108652115 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39514821767807007, "epoch": 14.82, "learning_rate": 6.762680025046963e-07, "loss": 0.2902, "step": 17529, "task_loss": 0.3208186626434326 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5034984350204468, "epoch": 14.82, "learning_rate": 6.731371321227302e-07, "loss": 0.428, "step": 17530, "task_loss": 0.8440741896629333 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.505757212638855, "epoch": 14.82, "learning_rate": 6.70006261740764e-07, "loss": 0.4813, "step": 17531, "task_loss": 0.7865477204322815 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4364761412143707, "epoch": 14.82, "learning_rate": 6.668753913587978e-07, "loss": 0.4035, "step": 17532, "task_loss": 0.9838846325874329 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4231335520744324, "epoch": 14.82, "learning_rate": 6.637445209768316e-07, "loss": 0.4172, "step": 17533, "task_loss": 0.4857509136199951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.48028767108917236, "epoch": 14.82, "learning_rate": 6.606136505948655e-07, "loss": 0.3843, "step": 17534, "task_loss": 1.2388051748275757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5117086172103882, "epoch": 14.82, "learning_rate": 6.574827802128993e-07, "loss": 0.4853, "step": 17535, "task_loss": 0.7916340231895447 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6138275861740112, "epoch": 14.82, "learning_rate": 6.54351909830933e-07, "loss": 0.5149, "step": 17536, "task_loss": 1.1724854707717896 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3642282485961914, "epoch": 14.82, "learning_rate": 6.512210394489668e-07, "loss": 0.3558, "step": 17537, "task_loss": 0.32095590233802795 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4202703535556793, "epoch": 14.82, "learning_rate": 6.480901690670006e-07, "loss": 0.3946, "step": 17538, "task_loss": 0.6058179140090942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34916841983795166, "epoch": 14.83, "learning_rate": 6.449592986850344e-07, "loss": 0.3996, "step": 17539, "task_loss": 0.10230612009763718 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28134337067604065, "epoch": 14.83, "learning_rate": 6.418284283030683e-07, "loss": 0.392, "step": 17540, "task_loss": 0.5215031504631042 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5221632719039917, "epoch": 14.83, "learning_rate": 6.386975579211021e-07, "loss": 0.5479, "step": 17541, "task_loss": 1.4279117584228516 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4339410066604614, "epoch": 14.83, "learning_rate": 6.355666875391359e-07, "loss": 0.3485, "step": 17542, "task_loss": 1.6459170579910278 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.33682870864868164, "epoch": 14.83, "learning_rate": 6.324358171571698e-07, "loss": 0.4242, "step": 17543, "task_loss": 0.32505449652671814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37038639187812805, "epoch": 14.83, "learning_rate": 6.293049467752036e-07, "loss": 0.3847, "step": 17544, "task_loss": 0.5257224440574646 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2917725145816803, "epoch": 14.83, "learning_rate": 6.261740763932374e-07, "loss": 0.3892, "step": 17545, "task_loss": 1.35243558883667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3776872754096985, "epoch": 14.83, "learning_rate": 6.230432060112711e-07, "loss": 0.4617, "step": 17546, "task_loss": 0.5930166840553284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49372828006744385, "epoch": 14.83, "learning_rate": 6.19912335629305e-07, "loss": 0.4772, "step": 17547, "task_loss": 0.8281850218772888 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18115711212158203, "epoch": 14.83, "learning_rate": 6.167814652473388e-07, "loss": 0.3798, "step": 17548, "task_loss": 0.18855668604373932 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41381436586380005, "epoch": 14.83, "learning_rate": 6.136505948653726e-07, "loss": 0.5021, "step": 17549, "task_loss": 1.2749016284942627 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36895155906677246, "epoch": 14.83, "learning_rate": 6.105197244834065e-07, "loss": 0.4032, "step": 17550, "task_loss": 0.5066489577293396 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6858806014060974, "epoch": 14.84, "learning_rate": 6.073888541014403e-07, "loss": 0.4899, "step": 17551, "task_loss": 1.055842399597168 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24971146881580353, "epoch": 14.84, "learning_rate": 6.04257983719474e-07, "loss": 0.4572, "step": 17552, "task_loss": 0.21455754339694977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3308858871459961, "epoch": 14.84, "learning_rate": 6.01127113337508e-07, "loss": 0.5633, "step": 17553, "task_loss": 0.5308948755264282 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40589678287506104, "epoch": 14.84, "learning_rate": 5.979962429555416e-07, "loss": 0.4619, "step": 17554, "task_loss": 0.35316145420074463 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4050978422164917, "epoch": 14.84, "learning_rate": 5.948653725735754e-07, "loss": 0.4544, "step": 17555, "task_loss": 0.684256374835968 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3541520833969116, "epoch": 14.84, "learning_rate": 5.917345021916092e-07, "loss": 0.363, "step": 17556, "task_loss": 0.44787025451660156 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3399156630039215, "epoch": 14.84, "learning_rate": 5.886036318096431e-07, "loss": 0.3764, "step": 17557, "task_loss": 0.2588905990123749 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49765264987945557, "epoch": 14.84, "learning_rate": 5.854727614276769e-07, "loss": 0.4766, "step": 17558, "task_loss": 0.4553956389427185 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3220192492008209, "epoch": 14.84, "learning_rate": 5.823418910457107e-07, "loss": 0.3247, "step": 17559, "task_loss": 1.2170994281768799 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36056309938430786, "epoch": 14.84, "learning_rate": 5.792110206637446e-07, "loss": 0.4233, "step": 17560, "task_loss": 0.5571791529655457 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.306240975856781, "epoch": 14.84, "learning_rate": 5.760801502817784e-07, "loss": 0.3784, "step": 17561, "task_loss": 0.522879958152771 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4721112847328186, "epoch": 14.84, "learning_rate": 5.729492798998122e-07, "loss": 0.464, "step": 17562, "task_loss": 0.6090897917747498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3265133202075958, "epoch": 14.85, "learning_rate": 5.69818409517846e-07, "loss": 0.5156, "step": 17563, "task_loss": 0.4393616020679474 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49956047534942627, "epoch": 14.85, "learning_rate": 5.666875391358798e-07, "loss": 0.4116, "step": 17564, "task_loss": 1.3780517578125 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6288386583328247, "epoch": 14.85, "learning_rate": 5.635566687539136e-07, "loss": 0.5507, "step": 17565, "task_loss": 0.928476870059967 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29654383659362793, "epoch": 14.85, "learning_rate": 5.604257983719475e-07, "loss": 0.4465, "step": 17566, "task_loss": 1.0601093769073486 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6197415590286255, "epoch": 14.85, "learning_rate": 5.572949279899813e-07, "loss": 0.5841, "step": 17567, "task_loss": 1.2889174222946167 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4269367456436157, "epoch": 14.85, "learning_rate": 5.54164057608015e-07, "loss": 0.4386, "step": 17568, "task_loss": 0.3141123652458191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5048753619194031, "epoch": 14.85, "learning_rate": 5.510331872260488e-07, "loss": 0.4548, "step": 17569, "task_loss": 1.103257656097412 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3075945973396301, "epoch": 14.85, "learning_rate": 5.479023168440827e-07, "loss": 0.4447, "step": 17570, "task_loss": 0.858813464641571 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3894437551498413, "epoch": 14.85, "learning_rate": 5.447714464621165e-07, "loss": 0.4565, "step": 17571, "task_loss": 0.5231536030769348 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24991455674171448, "epoch": 14.85, "learning_rate": 5.416405760801503e-07, "loss": 0.4259, "step": 17572, "task_loss": 0.10049643367528915 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5623128414154053, "epoch": 14.85, "learning_rate": 5.385097056981841e-07, "loss": 0.4755, "step": 17573, "task_loss": 0.33129605650901794 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3157581090927124, "epoch": 14.85, "learning_rate": 5.353788353162179e-07, "loss": 0.3701, "step": 17574, "task_loss": 0.4209228754043579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41167280077934265, "epoch": 14.86, "learning_rate": 5.322479649342517e-07, "loss": 0.4702, "step": 17575, "task_loss": 0.21744078397750854 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4107395112514496, "epoch": 14.86, "learning_rate": 5.291170945522856e-07, "loss": 0.5268, "step": 17576, "task_loss": 0.2712561786174774 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4754185974597931, "epoch": 14.86, "learning_rate": 5.259862241703194e-07, "loss": 0.4596, "step": 17577, "task_loss": 0.7322368621826172 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3179304301738739, "epoch": 14.86, "learning_rate": 5.228553537883532e-07, "loss": 0.4072, "step": 17578, "task_loss": 0.20879985392093658 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5530340671539307, "epoch": 14.86, "learning_rate": 5.19724483406387e-07, "loss": 0.4383, "step": 17579, "task_loss": 0.7900775074958801 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22078537940979004, "epoch": 14.86, "learning_rate": 5.165936130244209e-07, "loss": 0.4432, "step": 17580, "task_loss": 0.5249754190444946 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5464311838150024, "epoch": 14.86, "learning_rate": 5.134627426424546e-07, "loss": 0.3981, "step": 17581, "task_loss": 1.3972012996673584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17836017906665802, "epoch": 14.86, "learning_rate": 5.103318722604884e-07, "loss": 0.3898, "step": 17582, "task_loss": 0.018086811527609825 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6300501823425293, "epoch": 14.86, "learning_rate": 5.072010018785223e-07, "loss": 0.4812, "step": 17583, "task_loss": 1.1363258361816406 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.47281235456466675, "epoch": 14.86, "learning_rate": 5.040701314965561e-07, "loss": 0.4743, "step": 17584, "task_loss": 0.5606404542922974 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.21208234131336212, "epoch": 14.86, "learning_rate": 5.009392611145899e-07, "loss": 0.3985, "step": 17585, "task_loss": 0.5515217781066895 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34449225664138794, "epoch": 14.87, "learning_rate": 4.978083907326237e-07, "loss": 0.3971, "step": 17586, "task_loss": 1.8762232065200806 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37055498361587524, "epoch": 14.87, "learning_rate": 4.946775203506575e-07, "loss": 0.38, "step": 17587, "task_loss": 0.7306036353111267 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3581199049949646, "epoch": 14.87, "learning_rate": 4.915466499686913e-07, "loss": 0.484, "step": 17588, "task_loss": 0.5135178565979004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2960324287414551, "epoch": 14.87, "learning_rate": 4.884157795867251e-07, "loss": 0.3304, "step": 17589, "task_loss": 0.14335878193378448 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3835577964782715, "epoch": 14.87, "learning_rate": 4.852849092047589e-07, "loss": 0.4825, "step": 17590, "task_loss": 0.6124812960624695 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40307706594467163, "epoch": 14.87, "learning_rate": 4.821540388227927e-07, "loss": 0.4985, "step": 17591, "task_loss": 0.9361086487770081 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5105020999908447, "epoch": 14.87, "learning_rate": 4.790231684408265e-07, "loss": 0.4527, "step": 17592, "task_loss": 0.5390802025794983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3137331008911133, "epoch": 14.87, "learning_rate": 4.758922980588604e-07, "loss": 0.3717, "step": 17593, "task_loss": 0.23397018015384674 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5652321577072144, "epoch": 14.87, "learning_rate": 4.727614276768942e-07, "loss": 0.3968, "step": 17594, "task_loss": 0.5541998744010925 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2251037210226059, "epoch": 14.87, "learning_rate": 4.69630557294928e-07, "loss": 0.3843, "step": 17595, "task_loss": 0.6369791626930237 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2663033604621887, "epoch": 14.87, "learning_rate": 4.6649968691296184e-07, "loss": 0.4792, "step": 17596, "task_loss": 0.11835335940122604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3358374536037445, "epoch": 14.87, "learning_rate": 4.6336881653099563e-07, "loss": 0.2809, "step": 17597, "task_loss": 0.8153470754623413 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25809627771377563, "epoch": 14.88, "learning_rate": 4.602379461490294e-07, "loss": 0.3003, "step": 17598, "task_loss": 0.2918861508369446 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2904086709022522, "epoch": 14.88, "learning_rate": 4.571070757670633e-07, "loss": 0.3751, "step": 17599, "task_loss": 0.3954797089099884 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43943268060684204, "epoch": 14.88, "learning_rate": 4.539762053850971e-07, "loss": 0.5454, "step": 17600, "task_loss": 0.23605696856975555 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4396343231201172, "epoch": 14.88, "learning_rate": 4.5084533500313085e-07, "loss": 0.5032, "step": 17601, "task_loss": 0.03237997740507126 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20469655096530914, "epoch": 14.88, "learning_rate": 4.4771446462116465e-07, "loss": 0.3348, "step": 17602, "task_loss": 0.25800418853759766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.44377076625823975, "epoch": 14.88, "learning_rate": 4.4458359423919855e-07, "loss": 0.4258, "step": 17603, "task_loss": 0.6069327592849731 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2625168263912201, "epoch": 14.88, "learning_rate": 4.4145272385723234e-07, "loss": 0.3542, "step": 17604, "task_loss": 0.4182102382183075 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3382393419742584, "epoch": 14.88, "learning_rate": 4.3832185347526613e-07, "loss": 0.4722, "step": 17605, "task_loss": 0.28137171268463135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3344721496105194, "epoch": 14.88, "learning_rate": 4.351909830933e-07, "loss": 0.3694, "step": 17606, "task_loss": 0.27721765637397766 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3416728079319, "epoch": 14.88, "learning_rate": 4.3206011271133377e-07, "loss": 0.3935, "step": 17607, "task_loss": 1.5352779626846313 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4634465277194977, "epoch": 14.88, "learning_rate": 4.2892924232936756e-07, "loss": 0.4282, "step": 17608, "task_loss": 0.43033042550086975 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3956453502178192, "epoch": 14.88, "learning_rate": 4.257983719474014e-07, "loss": 0.3738, "step": 17609, "task_loss": 0.7599621415138245 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39760684967041016, "epoch": 14.89, "learning_rate": 4.226675015654352e-07, "loss": 0.4372, "step": 17610, "task_loss": 0.28803345561027527 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17225588858127594, "epoch": 14.89, "learning_rate": 4.19536631183469e-07, "loss": 0.3616, "step": 17611, "task_loss": 0.07472171634435654 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4491419792175293, "epoch": 14.89, "learning_rate": 4.164057608015029e-07, "loss": 0.3904, "step": 17612, "task_loss": 0.5757243037223816 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.284164696931839, "epoch": 14.89, "learning_rate": 4.132748904195367e-07, "loss": 0.3669, "step": 17613, "task_loss": 0.3568648099899292 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4325053095817566, "epoch": 14.89, "learning_rate": 4.101440200375705e-07, "loss": 0.3356, "step": 17614, "task_loss": 0.9708911776542664 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4102494716644287, "epoch": 14.89, "learning_rate": 4.070131496556042e-07, "loss": 0.4197, "step": 17615, "task_loss": 1.0905511379241943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31898927688598633, "epoch": 14.89, "learning_rate": 4.038822792736381e-07, "loss": 0.3821, "step": 17616, "task_loss": 0.12836097180843353 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3192773461341858, "epoch": 14.89, "learning_rate": 4.007514088916719e-07, "loss": 0.3998, "step": 17617, "task_loss": 0.3166307806968689 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2971268892288208, "epoch": 14.89, "learning_rate": 3.976205385097057e-07, "loss": 0.4231, "step": 17618, "task_loss": 0.1345132738351822 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34349939227104187, "epoch": 14.89, "learning_rate": 3.9448966812773955e-07, "loss": 0.3349, "step": 17619, "task_loss": 0.21473681926727295 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2855689227581024, "epoch": 14.89, "learning_rate": 3.9135879774577334e-07, "loss": 0.4952, "step": 17620, "task_loss": 1.1192454099655151 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5287128686904907, "epoch": 14.89, "learning_rate": 3.8822792736380714e-07, "loss": 0.4184, "step": 17621, "task_loss": 1.2449426651000977 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3846719264984131, "epoch": 14.9, "learning_rate": 3.85097056981841e-07, "loss": 0.5279, "step": 17622, "task_loss": 0.2531341314315796 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20635715126991272, "epoch": 14.9, "learning_rate": 3.819661865998748e-07, "loss": 0.3537, "step": 17623, "task_loss": 0.48051387071609497 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.19678714871406555, "epoch": 14.9, "learning_rate": 3.7883531621790857e-07, "loss": 0.3831, "step": 17624, "task_loss": 0.5995179414749146 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3667167127132416, "epoch": 14.9, "learning_rate": 3.757044458359424e-07, "loss": 0.3717, "step": 17625, "task_loss": 0.7488047480583191 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5668405294418335, "epoch": 14.9, "learning_rate": 3.7257357545397626e-07, "loss": 0.4703, "step": 17626, "task_loss": 0.4997750222682953 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.22011932730674744, "epoch": 14.9, "learning_rate": 3.6944270507201005e-07, "loss": 0.3207, "step": 17627, "task_loss": 0.3595038950443268 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2270529568195343, "epoch": 14.9, "learning_rate": 3.6631183469004385e-07, "loss": 0.4266, "step": 17628, "task_loss": 0.5726651549339294 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2128572016954422, "epoch": 14.9, "learning_rate": 3.6318096430807764e-07, "loss": 0.438, "step": 17629, "task_loss": 0.2950975000858307 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5031152367591858, "epoch": 14.9, "learning_rate": 3.600500939261115e-07, "loss": 0.3766, "step": 17630, "task_loss": 0.7206988334655762 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3759785294532776, "epoch": 14.9, "learning_rate": 3.5691922354414533e-07, "loss": 0.4701, "step": 17631, "task_loss": 0.8151062726974487 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.284443199634552, "epoch": 14.9, "learning_rate": 3.5378835316217907e-07, "loss": 0.4094, "step": 17632, "task_loss": 0.5158631801605225 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.43754443526268005, "epoch": 14.9, "learning_rate": 3.506574827802129e-07, "loss": 0.42, "step": 17633, "task_loss": 0.917314887046814 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29839521646499634, "epoch": 14.91, "learning_rate": 3.475266123982467e-07, "loss": 0.2594, "step": 17634, "task_loss": 0.5234326720237732 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26470157504081726, "epoch": 14.91, "learning_rate": 3.4439574201628056e-07, "loss": 0.3845, "step": 17635, "task_loss": 0.7388529181480408 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.262781023979187, "epoch": 14.91, "learning_rate": 3.4126487163431435e-07, "loss": 0.3405, "step": 17636, "task_loss": 0.859865665435791 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5761114954948425, "epoch": 14.91, "learning_rate": 3.3813400125234814e-07, "loss": 0.4632, "step": 17637, "task_loss": 0.9745572209358215 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3674583435058594, "epoch": 14.91, "learning_rate": 3.35003130870382e-07, "loss": 0.458, "step": 17638, "task_loss": 0.6946890950202942 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4425508379936218, "epoch": 14.91, "learning_rate": 3.318722604884158e-07, "loss": 0.3762, "step": 17639, "task_loss": 0.8066801428794861 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27836722135543823, "epoch": 14.91, "learning_rate": 3.287413901064496e-07, "loss": 0.4019, "step": 17640, "task_loss": 0.19153276085853577 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3243670165538788, "epoch": 14.91, "learning_rate": 3.256105197244834e-07, "loss": 0.469, "step": 17641, "task_loss": 0.28611400723457336 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6165218353271484, "epoch": 14.91, "learning_rate": 3.224796493425172e-07, "loss": 0.4719, "step": 17642, "task_loss": 0.6709120869636536 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3201737403869629, "epoch": 14.91, "learning_rate": 3.1934877896055106e-07, "loss": 0.4949, "step": 17643, "task_loss": 0.21267947554588318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.20068439841270447, "epoch": 14.91, "learning_rate": 3.162179085785849e-07, "loss": 0.2511, "step": 17644, "task_loss": 0.05503396317362785 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4900583028793335, "epoch": 14.91, "learning_rate": 3.130870381966187e-07, "loss": 0.351, "step": 17645, "task_loss": 1.112764835357666 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5228042006492615, "epoch": 14.92, "learning_rate": 3.099561678146525e-07, "loss": 0.5099, "step": 17646, "task_loss": 0.7502408623695374 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38042783737182617, "epoch": 14.92, "learning_rate": 3.068252974326863e-07, "loss": 0.4406, "step": 17647, "task_loss": 1.6200642585754395 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.424304723739624, "epoch": 14.92, "learning_rate": 3.0369442705072013e-07, "loss": 0.3872, "step": 17648, "task_loss": 0.29938241839408875 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24452419579029083, "epoch": 14.92, "learning_rate": 3.00563556668754e-07, "loss": 0.4435, "step": 17649, "task_loss": 0.3204987049102783 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6119600534439087, "epoch": 14.92, "learning_rate": 2.974326862867877e-07, "loss": 0.4946, "step": 17650, "task_loss": 1.0370515584945679 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32277339696884155, "epoch": 14.92, "learning_rate": 2.9430181590482156e-07, "loss": 0.3057, "step": 17651, "task_loss": 0.1638922244310379 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.372586727142334, "epoch": 14.92, "learning_rate": 2.9117094552285535e-07, "loss": 0.4255, "step": 17652, "task_loss": 0.779973566532135 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2662366032600403, "epoch": 14.92, "learning_rate": 2.880400751408892e-07, "loss": 0.3166, "step": 17653, "task_loss": 0.29961293935775757 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2663744390010834, "epoch": 14.92, "learning_rate": 2.84909204758923e-07, "loss": 0.3607, "step": 17654, "task_loss": 0.3134545087814331 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34819838404655457, "epoch": 14.92, "learning_rate": 2.817783343769568e-07, "loss": 0.3942, "step": 17655, "task_loss": 0.17172372341156006 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3003309369087219, "epoch": 14.92, "learning_rate": 2.7864746399499063e-07, "loss": 0.4249, "step": 17656, "task_loss": 0.5041530728340149 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5178362727165222, "epoch": 14.93, "learning_rate": 2.755165936130244e-07, "loss": 0.3988, "step": 17657, "task_loss": 0.33355239033699036 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5845444798469543, "epoch": 14.93, "learning_rate": 2.7238572323105827e-07, "loss": 0.3575, "step": 17658, "task_loss": 0.9513549208641052 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4537816643714905, "epoch": 14.93, "learning_rate": 2.6925485284909206e-07, "loss": 0.3547, "step": 17659, "task_loss": 0.5008144974708557 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3117698132991791, "epoch": 14.93, "learning_rate": 2.6612398246712585e-07, "loss": 0.389, "step": 17660, "task_loss": 0.6406015753746033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.309421569108963, "epoch": 14.93, "learning_rate": 2.629931120851597e-07, "loss": 0.4529, "step": 17661, "task_loss": 0.3461102247238159 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3088131844997406, "epoch": 14.93, "learning_rate": 2.598622417031935e-07, "loss": 0.4368, "step": 17662, "task_loss": 0.3051970303058624 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4856882095336914, "epoch": 14.93, "learning_rate": 2.567313713212273e-07, "loss": 0.5728, "step": 17663, "task_loss": 0.49107104539871216 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3714711666107178, "epoch": 14.93, "learning_rate": 2.5360050093926113e-07, "loss": 0.3414, "step": 17664, "task_loss": 0.2670043706893921 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.279105544090271, "epoch": 14.93, "learning_rate": 2.504696305572949e-07, "loss": 0.3767, "step": 17665, "task_loss": 0.2608501613140106 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4319036304950714, "epoch": 14.93, "learning_rate": 2.4733876017532877e-07, "loss": 0.3281, "step": 17666, "task_loss": 0.2595122456550598 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2723411023616791, "epoch": 14.93, "learning_rate": 2.4420788979336256e-07, "loss": 0.4384, "step": 17667, "task_loss": 0.7524128556251526 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5171504020690918, "epoch": 14.93, "learning_rate": 2.4107701941139636e-07, "loss": 0.4155, "step": 17668, "task_loss": 0.6023274064064026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.24918577075004578, "epoch": 14.94, "learning_rate": 2.379461490294302e-07, "loss": 0.3922, "step": 17669, "task_loss": 0.5215128064155579 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5472376942634583, "epoch": 14.94, "learning_rate": 2.34815278647464e-07, "loss": 0.5556, "step": 17670, "task_loss": 0.5129343271255493 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.28533339500427246, "epoch": 14.94, "learning_rate": 2.3168440826549781e-07, "loss": 0.2976, "step": 17671, "task_loss": 0.5082351565361023 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4810675382614136, "epoch": 14.94, "learning_rate": 2.2855353788353166e-07, "loss": 0.538, "step": 17672, "task_loss": 0.3577061891555786 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.25956273078918457, "epoch": 14.94, "learning_rate": 2.2542266750156543e-07, "loss": 0.334, "step": 17673, "task_loss": 0.047693341970443726 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3272782564163208, "epoch": 14.94, "learning_rate": 2.2229179711959927e-07, "loss": 0.3212, "step": 17674, "task_loss": 0.7277323007583618 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5272454619407654, "epoch": 14.94, "learning_rate": 2.1916092673763307e-07, "loss": 0.5016, "step": 17675, "task_loss": 1.7146008014678955 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5584010481834412, "epoch": 14.94, "learning_rate": 2.1603005635566689e-07, "loss": 0.5199, "step": 17676, "task_loss": 0.48655539751052856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46018683910369873, "epoch": 14.94, "learning_rate": 2.128991859737007e-07, "loss": 0.3822, "step": 17677, "task_loss": 0.469916969537735 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3087535798549652, "epoch": 14.94, "learning_rate": 2.097683155917345e-07, "loss": 0.3573, "step": 17678, "task_loss": 0.10643457621335983 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5647298097610474, "epoch": 14.94, "learning_rate": 2.0663744520976834e-07, "loss": 0.3788, "step": 17679, "task_loss": 0.6593875288963318 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7062994241714478, "epoch": 14.94, "learning_rate": 2.035065748278021e-07, "loss": 0.5581, "step": 17680, "task_loss": 1.197843074798584 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.34617745876312256, "epoch": 14.95, "learning_rate": 2.0037570444583596e-07, "loss": 0.3957, "step": 17681, "task_loss": 0.36241739988327026 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6047194004058838, "epoch": 14.95, "learning_rate": 1.9724483406386978e-07, "loss": 0.429, "step": 17682, "task_loss": 0.11992556601762772 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41939419507980347, "epoch": 14.95, "learning_rate": 1.9411396368190357e-07, "loss": 0.4941, "step": 17683, "task_loss": 0.5133957862854004 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2541331648826599, "epoch": 14.95, "learning_rate": 1.909830932999374e-07, "loss": 0.3356, "step": 17684, "task_loss": 0.26775655150413513 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27744925022125244, "epoch": 14.95, "learning_rate": 1.878522229179712e-07, "loss": 0.31, "step": 17685, "task_loss": 0.2689451277256012 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.18323099613189697, "epoch": 14.95, "learning_rate": 1.8472135253600503e-07, "loss": 0.3176, "step": 17686, "task_loss": 0.15737859904766083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.17710766196250916, "epoch": 14.95, "learning_rate": 1.8159048215403882e-07, "loss": 0.3905, "step": 17687, "task_loss": 0.020725395530462265 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4390680491924286, "epoch": 14.95, "learning_rate": 1.7845961177207267e-07, "loss": 0.4626, "step": 17688, "task_loss": 1.110566258430481 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.32032376527786255, "epoch": 14.95, "learning_rate": 1.7532874139010646e-07, "loss": 0.4299, "step": 17689, "task_loss": 0.09557007998228073 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39051729440689087, "epoch": 14.95, "learning_rate": 1.7219787100814028e-07, "loss": 0.5228, "step": 17690, "task_loss": 0.6805740594863892 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4595537483692169, "epoch": 14.95, "learning_rate": 1.6906700062617407e-07, "loss": 0.3918, "step": 17691, "task_loss": 0.33104637265205383 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.375398725271225, "epoch": 14.95, "learning_rate": 1.659361302442079e-07, "loss": 0.3532, "step": 17692, "task_loss": 0.4346473217010498 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.27125802636146545, "epoch": 14.96, "learning_rate": 1.628052598622417e-07, "loss": 0.4653, "step": 17693, "task_loss": 0.26340827345848083 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5297833681106567, "epoch": 14.96, "learning_rate": 1.5967438948027553e-07, "loss": 0.56, "step": 17694, "task_loss": 0.41308343410491943 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29006141424179077, "epoch": 14.96, "learning_rate": 1.5654351909830935e-07, "loss": 0.2663, "step": 17695, "task_loss": 0.2736333906650543 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4184926152229309, "epoch": 14.96, "learning_rate": 1.5341264871634314e-07, "loss": 0.3493, "step": 17696, "task_loss": 0.7829653024673462 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.384692907333374, "epoch": 14.96, "learning_rate": 1.50281778334377e-07, "loss": 0.4681, "step": 17697, "task_loss": 0.6243430972099304 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3380065858364105, "epoch": 14.96, "learning_rate": 1.4715090795241078e-07, "loss": 0.4102, "step": 17698, "task_loss": 0.7630986571311951 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.31436991691589355, "epoch": 14.96, "learning_rate": 1.440200375704446e-07, "loss": 0.3699, "step": 17699, "task_loss": 0.3381306529045105 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5032584071159363, "epoch": 14.96, "learning_rate": 1.408891671884784e-07, "loss": 0.4149, "step": 17700, "task_loss": 0.6535839438438416 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5881023406982422, "epoch": 14.96, "learning_rate": 1.377582968065122e-07, "loss": 0.5044, "step": 17701, "task_loss": 0.368975967168808 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3048665523529053, "epoch": 14.96, "learning_rate": 1.3462742642454603e-07, "loss": 0.4193, "step": 17702, "task_loss": 0.626523494720459 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.4814201891422272, "epoch": 14.96, "learning_rate": 1.3149655604257985e-07, "loss": 0.3988, "step": 17703, "task_loss": 0.29614001512527466 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41782140731811523, "epoch": 14.96, "learning_rate": 1.2836568566061364e-07, "loss": 0.336, "step": 17704, "task_loss": 0.588909387588501 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6296596527099609, "epoch": 14.97, "learning_rate": 1.2523481527864746e-07, "loss": 0.4192, "step": 17705, "task_loss": 1.821213960647583 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2432197630405426, "epoch": 14.97, "learning_rate": 1.2210394489668128e-07, "loss": 0.332, "step": 17706, "task_loss": 0.5642341375350952 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2715016007423401, "epoch": 14.97, "learning_rate": 1.189730745147151e-07, "loss": 0.4343, "step": 17707, "task_loss": 0.8865232467651367 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.40872323513031006, "epoch": 14.97, "learning_rate": 1.1584220413274891e-07, "loss": 0.4647, "step": 17708, "task_loss": 0.6493148803710938 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1956721395254135, "epoch": 14.97, "learning_rate": 1.1271133375078271e-07, "loss": 0.2819, "step": 17709, "task_loss": 0.17170341312885284 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.38174229860305786, "epoch": 14.97, "learning_rate": 1.0958046336881653e-07, "loss": 0.488, "step": 17710, "task_loss": 0.6999618411064148 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2841135561466217, "epoch": 14.97, "learning_rate": 1.0644959298685035e-07, "loss": 0.3803, "step": 17711, "task_loss": 0.30578944087028503 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5259791016578674, "epoch": 14.97, "learning_rate": 1.0331872260488417e-07, "loss": 0.3932, "step": 17712, "task_loss": 0.7537779211997986 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6430478096008301, "epoch": 14.97, "learning_rate": 1.0018785222291798e-07, "loss": 0.5307, "step": 17713, "task_loss": 0.7224963903427124 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.37145647406578064, "epoch": 14.97, "learning_rate": 9.705698184095178e-08, "loss": 0.4161, "step": 17714, "task_loss": 1.2658225297927856 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7267906665802002, "epoch": 14.97, "learning_rate": 9.39261114589856e-08, "loss": 0.5041, "step": 17715, "task_loss": 0.7551901340484619 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2146683931350708, "epoch": 14.97, "learning_rate": 9.079524107701941e-08, "loss": 0.3704, "step": 17716, "task_loss": 0.2300504893064499 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.49091100692749023, "epoch": 14.98, "learning_rate": 8.766437069505323e-08, "loss": 0.407, "step": 17717, "task_loss": 0.6440739631652832 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.41270673274993896, "epoch": 14.98, "learning_rate": 8.453350031308704e-08, "loss": 0.3425, "step": 17718, "task_loss": 0.6616591215133667 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3102077543735504, "epoch": 14.98, "learning_rate": 8.140262993112085e-08, "loss": 0.3757, "step": 17719, "task_loss": 0.3418375551700592 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.39616596698760986, "epoch": 14.98, "learning_rate": 7.827175954915467e-08, "loss": 0.3568, "step": 17720, "task_loss": 0.3934228718280792 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3543587625026703, "epoch": 14.98, "learning_rate": 7.51408891671885e-08, "loss": 0.4369, "step": 17721, "task_loss": 1.143181324005127 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.2768803536891937, "epoch": 14.98, "learning_rate": 7.20100187852223e-08, "loss": 0.4273, "step": 17722, "task_loss": 1.2091238498687744 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6343824863433838, "epoch": 14.98, "learning_rate": 6.88791484032561e-08, "loss": 0.498, "step": 17723, "task_loss": 0.13667918741703033 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6028141975402832, "epoch": 14.98, "learning_rate": 6.574827802128993e-08, "loss": 0.4411, "step": 17724, "task_loss": 2.0306191444396973 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3386038541793823, "epoch": 14.98, "learning_rate": 6.261740763932373e-08, "loss": 0.4958, "step": 17725, "task_loss": 0.7755573391914368 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3860078454017639, "epoch": 14.98, "learning_rate": 5.948653725735755e-08, "loss": 0.3348, "step": 17726, "task_loss": 0.7015226483345032 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5790970921516418, "epoch": 14.98, "learning_rate": 5.635566687539136e-08, "loss": 0.4937, "step": 17727, "task_loss": 0.3056791126728058 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3524666428565979, "epoch": 14.99, "learning_rate": 5.3224796493425176e-08, "loss": 0.3947, "step": 17728, "task_loss": 1.5284773111343384 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29941999912261963, "epoch": 14.99, "learning_rate": 5.009392611145899e-08, "loss": 0.4895, "step": 17729, "task_loss": 0.30867427587509155 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.29999321699142456, "epoch": 14.99, "learning_rate": 4.69630557294928e-08, "loss": 0.2893, "step": 17730, "task_loss": 0.24930277466773987 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.36976271867752075, "epoch": 14.99, "learning_rate": 4.3832185347526615e-08, "loss": 0.3958, "step": 17731, "task_loss": 0.3603629767894745 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.7327448129653931, "epoch": 14.99, "learning_rate": 4.070131496556043e-08, "loss": 0.5276, "step": 17732, "task_loss": 1.042410135269165 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3764733672142029, "epoch": 14.99, "learning_rate": 3.757044458359425e-08, "loss": 0.43, "step": 17733, "task_loss": 0.8457286357879639 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.45506879687309265, "epoch": 14.99, "learning_rate": 3.443957420162805e-08, "loss": 0.421, "step": 17734, "task_loss": 0.4726117253303528 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5280320048332214, "epoch": 14.99, "learning_rate": 3.1308703819661866e-08, "loss": 0.4135, "step": 17735, "task_loss": 0.677895188331604 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3642042577266693, "epoch": 14.99, "learning_rate": 2.817783343769568e-08, "loss": 0.3653, "step": 17736, "task_loss": 0.3693331778049469 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3142012357711792, "epoch": 14.99, "learning_rate": 2.5046963055729494e-08, "loss": 0.4261, "step": 17737, "task_loss": 1.6189566850662231 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3734574019908905, "epoch": 14.99, "learning_rate": 2.1916092673763307e-08, "loss": 0.4688, "step": 17738, "task_loss": 0.7763621807098389 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.1641080677509308, "epoch": 14.99, "learning_rate": 1.8785222291797123e-08, "loss": 0.2917, "step": 17739, "task_loss": 0.015387102961540222 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.3648202419281006, "epoch": 15.0, "learning_rate": 1.5654351909830933e-08, "loss": 0.4145, "step": 17740, "task_loss": 0.5035169124603271 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5697094202041626, "epoch": 15.0, "learning_rate": 1.2523481527864747e-08, "loss": 0.3603, "step": 17741, "task_loss": 0.6707842350006104 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.46914881467819214, "epoch": 15.0, "learning_rate": 9.392611145898562e-09, "loss": 0.3854, "step": 17742, "task_loss": 0.4619447588920593 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.5434233546257019, "epoch": 15.0, "learning_rate": 6.261740763932374e-09, "loss": 0.4922, "step": 17743, "task_loss": 0.9814426302909851 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.26151156425476074, "epoch": 15.0, "learning_rate": 3.130870381966187e-09, "loss": 0.42, "step": 17744, "task_loss": 1.0141013860702515 }, { "compression/movement_sparsity/importance_regularization_factor": 1.0, "compression/movement_sparsity/importance_threshold": 0.0, "compression/movement_sparsity/linear_layer_sparsity": 0.43105931586566787, "compression/movement_sparsity/model_sparsity": 0.4162511022721184, "compression_loss": 0.0, "distillation_loss": 0.6076030731201172, "epoch": 15.0, "learning_rate": 0.0, "loss": 0.4777, "step": 17745, "task_loss": 0.30477121472358704 }, { "epoch": 15.0, "step": 17745, "total_flos": 8.94984447373277e+19, "train_loss": 17.32710409305085, "train_runtime": 64003.0137, "train_samples_per_second": 17.753, "train_steps_per_second": 0.277 } ], "max_steps": 17745, "num_train_epochs": 15, "total_flos": 8.94984447373277e+19, "trial_name": null, "trial_params": null }